2002-01-19 14:57:47 +00:00
|
|
|
|
/* GStreamer
|
|
|
|
|
* Copyright (C) <1999> Erik Walthinsen <omega@temple-baptist.com>
|
2001-12-22 23:22:30 +00:00
|
|
|
|
*
|
|
|
|
|
* This library is free software; you can redistribute it and/or
|
|
|
|
|
* modify it under the terms of the GNU Library General Public
|
|
|
|
|
* License as published by the Free Software Foundation; either
|
|
|
|
|
* version 2 of the License, or (at your option) any later version.
|
|
|
|
|
*
|
|
|
|
|
* This library is distributed in the hope that it will be useful,
|
|
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
|
|
|
* Library General Public License for more details.
|
|
|
|
|
*
|
|
|
|
|
* You should have received a copy of the GNU Library General Public
|
|
|
|
|
* License along with this library; if not, write to the
|
|
|
|
|
* Free Software Foundation, Inc., 59 Temple Place - Suite 330,
|
|
|
|
|
* Boston, MA 02111-1307, USA.
|
|
|
|
|
*/
|
2004-01-27 09:00:01 +00:00
|
|
|
|
/* Element-Checklist-Version: 5 */
|
2001-12-22 23:22:30 +00:00
|
|
|
|
|
2003-06-29 19:46:12 +00:00
|
|
|
|
#ifdef HAVE_CONFIG_H
|
|
|
|
|
#include "config.h"
|
|
|
|
|
#endif
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
2001-12-22 23:22:30 +00:00
|
|
|
|
#include <string.h>
|
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
#include "gst/riff/riff-media.h"
|
2001-12-22 23:22:30 +00:00
|
|
|
|
#include "gstavidemux.h"
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
#include "avi-ids.h"
|
2005-11-11 16:48:58 +00:00
|
|
|
|
#include <gst/gst-i18n-plugin.h>
|
2001-12-22 23:22:30 +00:00
|
|
|
|
|
2003-11-24 23:11:42 +00:00
|
|
|
|
GST_DEBUG_CATEGORY_STATIC (avidemux_debug);
|
|
|
|
|
#define GST_CAT_DEFAULT avidemux_debug
|
|
|
|
|
|
2005-01-13 17:11:08 +00:00
|
|
|
|
GST_DEBUG_CATEGORY_EXTERN (GST_CAT_EVENT);
|
|
|
|
|
|
2004-03-14 22:34:33 +00:00
|
|
|
|
static GstStaticPadTemplate sink_templ = GST_STATIC_PAD_TEMPLATE ("sink",
|
|
|
|
|
GST_PAD_SINK,
|
|
|
|
|
GST_PAD_ALWAYS,
|
|
|
|
|
GST_STATIC_CAPS ("video/x-msvideo")
|
|
|
|
|
);
|
2003-07-06 20:49:52 +00:00
|
|
|
|
|
2004-03-14 22:34:33 +00:00
|
|
|
|
static void gst_avi_demux_base_init (GstAviDemuxClass * klass);
|
|
|
|
|
static void gst_avi_demux_class_init (GstAviDemuxClass * klass);
|
|
|
|
|
static void gst_avi_demux_init (GstAviDemux * avi);
|
2003-07-06 20:49:52 +00:00
|
|
|
|
|
2004-03-14 22:34:33 +00:00
|
|
|
|
static void gst_avi_demux_reset (GstAviDemux * avi);
|
2003-07-06 20:49:52 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
#if 0
|
2004-03-14 22:34:33 +00:00
|
|
|
|
static const GstEventMask *gst_avi_demux_get_event_mask (GstPad * pad);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
#endif
|
2004-03-14 22:34:33 +00:00
|
|
|
|
static gboolean gst_avi_demux_handle_src_event (GstPad * pad, GstEvent * event);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
|
|
|
|
|
#if 0
|
2004-03-14 22:34:33 +00:00
|
|
|
|
static const GstFormat *gst_avi_demux_get_src_formats (GstPad * pad);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
#endif
|
2004-03-14 22:34:33 +00:00
|
|
|
|
static const GstQueryType *gst_avi_demux_get_src_query_types (GstPad * pad);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
static gboolean gst_avi_demux_handle_src_query (GstPad * pad, GstQuery * query);
|
2004-03-14 22:34:33 +00:00
|
|
|
|
static gboolean gst_avi_demux_src_convert (GstPad * pad,
|
|
|
|
|
GstFormat src_format,
|
|
|
|
|
gint64 src_value, GstFormat * dest_format, gint64 * dest_value);
|
2002-05-26 21:59:22 +00:00
|
|
|
|
|
2005-11-11 16:48:58 +00:00
|
|
|
|
static gboolean gst_avi_demux_handle_seek (GstAviDemux * avi, gboolean update);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
static void gst_avi_demux_loop (GstPad * pad);
|
|
|
|
|
static gboolean gst_avi_demux_sink_activate (GstPad * sinkpad);
|
|
|
|
|
static gboolean gst_avi_demux_sink_activate_pull (GstPad * sinkpad,
|
|
|
|
|
gboolean active);
|
2005-09-02 15:44:50 +00:00
|
|
|
|
static GstStateChangeReturn gst_avi_demux_change_state (GstElement * element,
|
|
|
|
|
GstStateChange transition);
|
2001-12-22 23:22:30 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
static GstElementClass *parent_class = NULL;
|
2004-03-14 22:34:33 +00:00
|
|
|
|
|
2001-12-22 23:22:30 +00:00
|
|
|
|
GType
|
2004-03-14 22:34:33 +00:00
|
|
|
|
gst_avi_demux_get_type (void)
|
2001-12-22 23:22:30 +00:00
|
|
|
|
{
|
|
|
|
|
static GType avi_demux_type = 0;
|
|
|
|
|
|
|
|
|
|
if (!avi_demux_type) {
|
|
|
|
|
static const GTypeInfo avi_demux_info = {
|
2004-03-14 22:34:33 +00:00
|
|
|
|
sizeof (GstAviDemuxClass),
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
(GBaseInitFunc) gst_avi_demux_base_init,
|
2001-12-22 23:22:30 +00:00
|
|
|
|
NULL,
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
(GClassInitFunc) gst_avi_demux_class_init,
|
2001-12-22 23:22:30 +00:00
|
|
|
|
NULL,
|
|
|
|
|
NULL,
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
sizeof (GstAviDemux),
|
2001-12-22 23:22:30 +00:00
|
|
|
|
0,
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
(GInstanceInitFunc) gst_avi_demux_init,
|
2001-12-22 23:22:30 +00:00
|
|
|
|
};
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
|
|
|
|
avi_demux_type =
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
g_type_register_static (GST_TYPE_ELEMENT,
|
2004-03-15 19:32:27 +00:00
|
|
|
|
"GstAviDemux", &avi_demux_info, 0);
|
2001-12-22 23:22:30 +00:00
|
|
|
|
}
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
2001-12-22 23:22:30 +00:00
|
|
|
|
return avi_demux_type;
|
|
|
|
|
}
|
|
|
|
|
|
2003-11-01 02:07:52 +00:00
|
|
|
|
static void
|
2004-03-14 22:34:33 +00:00
|
|
|
|
gst_avi_demux_base_init (GstAviDemuxClass * klass)
|
2003-11-01 02:07:52 +00:00
|
|
|
|
{
|
2004-03-14 22:34:33 +00:00
|
|
|
|
static GstElementDetails gst_avi_demux_details =
|
|
|
|
|
GST_ELEMENT_DETAILS ("Avi demuxer",
|
|
|
|
|
"Codec/Demuxer",
|
|
|
|
|
"Demultiplex an avi file into audio and video",
|
|
|
|
|
"Erik Walthinsen <omega@cse.ogi.edu>\n"
|
|
|
|
|
"Wim Taymans <wim.taymans@chello.be>\n"
|
|
|
|
|
"Ronald Bultje <rbultje@ronald.bitfreak.net>");
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
GstElementClass *element_class = GST_ELEMENT_CLASS (klass);
|
|
|
|
|
GstPadTemplate *videosrctempl, *audiosrctempl;
|
|
|
|
|
GstCaps *audcaps, *vidcaps;
|
2003-11-01 02:07:52 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
audcaps = gst_riff_create_audio_template_caps ();
|
2005-09-26 13:06:27 +00:00
|
|
|
|
gst_caps_append (audcaps, gst_caps_new_simple ("audio/x-avi-unknown", NULL));
|
2003-11-01 02:07:52 +00:00
|
|
|
|
audiosrctempl = gst_pad_template_new ("audio_%02d",
|
2004-03-14 22:34:33 +00:00
|
|
|
|
GST_PAD_SRC, GST_PAD_SOMETIMES, audcaps);
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
2003-12-22 01:47:09 +00:00
|
|
|
|
vidcaps = gst_riff_create_video_template_caps ();
|
|
|
|
|
gst_caps_append (vidcaps, gst_riff_create_iavs_template_caps ());
|
2005-09-26 13:06:27 +00:00
|
|
|
|
gst_caps_append (vidcaps, gst_caps_new_simple ("video/x-avi-unknown", NULL));
|
2003-11-01 02:07:52 +00:00
|
|
|
|
videosrctempl = gst_pad_template_new ("video_%02d",
|
2004-03-14 22:34:33 +00:00
|
|
|
|
GST_PAD_SRC, GST_PAD_SOMETIMES, vidcaps);
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
2003-11-01 02:07:52 +00:00
|
|
|
|
gst_element_class_add_pad_template (element_class, audiosrctempl);
|
|
|
|
|
gst_element_class_add_pad_template (element_class, videosrctempl);
|
|
|
|
|
gst_element_class_add_pad_template (element_class,
|
2003-12-22 01:47:09 +00:00
|
|
|
|
gst_static_pad_template_get (&sink_templ));
|
2003-11-01 02:07:52 +00:00
|
|
|
|
gst_element_class_set_details (element_class, &gst_avi_demux_details);
|
|
|
|
|
}
|
|
|
|
|
|
2001-12-22 23:22:30 +00:00
|
|
|
|
static void
|
2004-03-14 22:34:33 +00:00
|
|
|
|
gst_avi_demux_class_init (GstAviDemuxClass * klass)
|
2001-12-22 23:22:30 +00:00
|
|
|
|
{
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GstElementClass *gstelement_class = GST_ELEMENT_CLASS (klass);
|
2001-12-22 23:22:30 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
GST_DEBUG_CATEGORY_INIT (avidemux_debug, "avidemux",
|
2004-03-14 22:34:33 +00:00
|
|
|
|
0, "Demuxer for AVI streams");
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
2005-09-26 13:06:27 +00:00
|
|
|
|
parent_class = g_type_class_peek_parent (klass);
|
2004-03-14 22:34:33 +00:00
|
|
|
|
|
2001-12-23 22:14:25 +00:00
|
|
|
|
gstelement_class->change_state = gst_avi_demux_change_state;
|
2001-12-22 23:22:30 +00:00
|
|
|
|
}
|
|
|
|
|
|
2004-03-14 22:34:33 +00:00
|
|
|
|
static void
|
|
|
|
|
gst_avi_demux_init (GstAviDemux * avi)
|
2003-01-15 00:41:40 +00:00
|
|
|
|
{
|
2004-03-14 22:34:33 +00:00
|
|
|
|
avi->sinkpad =
|
|
|
|
|
gst_pad_new_from_template (gst_static_pad_template_get (&sink_templ),
|
|
|
|
|
"sink");
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_pad_set_activate_function (avi->sinkpad, gst_avi_demux_sink_activate);
|
|
|
|
|
gst_pad_set_activatepull_function (avi->sinkpad,
|
|
|
|
|
gst_avi_demux_sink_activate_pull);
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
gst_element_add_pad (GST_ELEMENT (avi), avi->sinkpad);
|
2003-01-21 22:23:32 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
gst_avi_demux_reset (avi);
|
2003-01-21 22:23:32 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
avi->index_entries = NULL;
|
|
|
|
|
memset (&avi->stream, 0, sizeof (avi->stream));
|
2003-01-21 22:23:32 +00:00
|
|
|
|
}
|
|
|
|
|
|
2003-01-31 19:37:26 +00:00
|
|
|
|
static void
|
2004-03-14 22:34:33 +00:00
|
|
|
|
gst_avi_demux_reset (GstAviDemux * avi)
|
2003-01-31 19:37:26 +00:00
|
|
|
|
{
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
gint i;
|
2003-02-02 23:48:30 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
for (i = 0; i < avi->num_streams; i++) {
|
|
|
|
|
g_free (avi->stream[i].strh);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
g_free (avi->stream[i].strf.data);
|
2005-08-04 18:31:12 +00:00
|
|
|
|
if (avi->stream[i].name)
|
|
|
|
|
g_free (avi->stream[i].name);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (avi->stream[i].initdata)
|
|
|
|
|
gst_buffer_unref (avi->stream[i].initdata);
|
|
|
|
|
if (avi->stream[i].extradata)
|
|
|
|
|
gst_buffer_unref (avi->stream[i].extradata);
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
if (avi->stream[i].pad)
|
|
|
|
|
gst_element_remove_pad (GST_ELEMENT (avi), avi->stream[i].pad);
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
|
|
|
|
memset (&avi->stream, 0, sizeof (avi->stream));
|
2003-01-31 19:37:26 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
avi->num_streams = 0;
|
|
|
|
|
avi->num_v_streams = 0;
|
|
|
|
|
avi->num_a_streams = 0;
|
2003-01-31 19:37:26 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
avi->state = GST_AVI_DEMUX_START;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
avi->offset = 0;
|
2003-01-31 19:37:26 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
g_free (avi->index_entries);
|
|
|
|
|
avi->index_entries = NULL;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
avi->index_size = 0;
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
avi->index_offset = 0;
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
avi->current_entry = 0;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
g_free (avi->avih);
|
|
|
|
|
avi->avih = NULL;
|
2003-01-31 19:37:26 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (avi->seek_event)
|
|
|
|
|
gst_event_unref (avi->seek_event);
|
|
|
|
|
avi->seek_event = NULL;
|
2005-11-11 16:48:58 +00:00
|
|
|
|
|
|
|
|
|
avi->segment_rate = 1.0;
|
|
|
|
|
avi->segment_flags = 0;
|
|
|
|
|
avi->segment_start = -1;
|
|
|
|
|
avi->segment_stop = -1;
|
2003-01-31 19:37:26 +00:00
|
|
|
|
}
|
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
static gst_avi_index_entry *
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_index_next (GstAviDemux * avi, gint stream_nr, gint start)
|
2001-12-22 23:22:30 +00:00
|
|
|
|
{
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
gint i;
|
|
|
|
|
gst_avi_index_entry *entry = NULL;
|
2003-07-06 20:49:52 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
for (i = start; i < avi->index_size; i++) {
|
|
|
|
|
entry = &avi->index_entries[i];
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (entry->stream_nr == stream_nr)
|
2002-01-31 22:22:42 +00:00
|
|
|
|
break;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
2003-07-06 20:49:52 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
return entry;
|
|
|
|
|
}
|
2003-07-06 20:49:52 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
static gst_avi_index_entry *
|
2004-03-14 22:34:33 +00:00
|
|
|
|
gst_avi_demux_index_entry_for_time (GstAviDemux * avi,
|
|
|
|
|
gint stream_nr, guint64 time, guint32 flags)
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
{
|
|
|
|
|
gst_avi_index_entry *entry = NULL, *last_entry = NULL;
|
|
|
|
|
gint i;
|
2003-07-06 20:49:52 +00:00
|
|
|
|
|
2005-11-11 16:48:58 +00:00
|
|
|
|
GST_LOG_OBJECT (avi, "stream_nr:%d , time:%" GST_TIME_FORMAT " flags:%d",
|
|
|
|
|
stream_nr, GST_TIME_ARGS (time), flags);
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
i = -1;
|
|
|
|
|
do {
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
entry = gst_avi_demux_index_next (avi, stream_nr, i + 1);
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
if (!entry)
|
|
|
|
|
return NULL;
|
2003-07-06 20:49:52 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
i = entry->index_nr;
|
2003-07-06 20:49:52 +00:00
|
|
|
|
|
2005-11-11 16:48:58 +00:00
|
|
|
|
GST_LOG_OBJECT (avi,
|
|
|
|
|
"looking at entry %d / ts:%" GST_TIME_FORMAT " / dur:%" GST_TIME_FORMAT
|
|
|
|
|
" flags:%d", i, GST_TIME_ARGS (entry->ts), GST_TIME_ARGS (entry->dur),
|
|
|
|
|
entry->flags);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (entry->ts <= time && (entry->flags & flags) == flags)
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
last_entry = entry;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
} while (entry->ts < time);
|
2003-07-06 20:49:52 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
return last_entry;
|
|
|
|
|
}
|
2003-07-06 20:49:52 +00:00
|
|
|
|
|
2001-12-22 23:22:30 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
#if 0
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
static const GstFormat *
|
2004-03-14 22:34:33 +00:00
|
|
|
|
gst_avi_demux_get_src_formats (GstPad * pad)
|
2001-12-22 23:22:30 +00:00
|
|
|
|
{
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
avi_stream_context *stream = gst_pad_get_element_private (pad);
|
2001-12-22 23:22:30 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
static const GstFormat src_a_formats[] = {
|
|
|
|
|
GST_FORMAT_TIME,
|
|
|
|
|
GST_FORMAT_BYTES,
|
|
|
|
|
GST_FORMAT_DEFAULT,
|
|
|
|
|
0
|
|
|
|
|
};
|
|
|
|
|
static const GstFormat src_v_formats[] = {
|
|
|
|
|
GST_FORMAT_TIME,
|
|
|
|
|
GST_FORMAT_DEFAULT,
|
|
|
|
|
0
|
|
|
|
|
};
|
2001-12-22 23:22:30 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
return (stream->strh->type == GST_RIFF_FCC_auds ?
|
2004-03-14 22:34:33 +00:00
|
|
|
|
src_a_formats : src_v_formats);
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
#endif
|
2003-07-06 20:49:52 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
static gboolean
|
2004-03-14 22:34:33 +00:00
|
|
|
|
gst_avi_demux_src_convert (GstPad * pad,
|
|
|
|
|
GstFormat src_format,
|
|
|
|
|
gint64 src_value, GstFormat * dest_format, gint64 * dest_value)
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
{
|
|
|
|
|
gboolean res = TRUE;
|
2005-11-11 16:48:58 +00:00
|
|
|
|
GstAviDemux *avidemux = GST_AVI_DEMUX (gst_pad_get_parent (pad));
|
2004-03-14 22:34:33 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
avi_stream_context *stream = gst_pad_get_element_private (pad);
|
2003-07-06 20:49:52 +00:00
|
|
|
|
|
2005-11-11 16:48:58 +00:00
|
|
|
|
GST_LOG_OBJECT (avidemux,
|
|
|
|
|
"Received src_format:%d, src_value:%lld, dest_format:%d", src_format,
|
|
|
|
|
src_value, *dest_format);
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (src_format == *dest_format) {
|
|
|
|
|
*dest_value = src_value;
|
2005-11-11 16:48:58 +00:00
|
|
|
|
goto done;
|
|
|
|
|
}
|
|
|
|
|
if (!stream->strh || !stream->strf.data) {
|
|
|
|
|
res = FALSE;
|
|
|
|
|
goto done;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
}
|
ext/dirac/: Do something. Don't actually know if this works because I don't have a demuxer yet.
Original commit message from CVS:
* ext/dirac/Makefile.am:
* ext/dirac/gstdirac.cc:
* ext/dirac/gstdiracdec.cc:
* ext/dirac/gstdiracdec.h:
Do something. Don't actually know if this works because I don't
have a demuxer yet.
* ext/gsm/gstgsmdec.c: (gst_gsmdec_getcaps):
Add channels=1 to caps returned from _getcaps().
* ext/ogg/gstogmparse.c: (gst_ogm_audio_parse_get_type),
(gst_ogm_video_parse_get_type), (gst_ogm_audio_parse_base_init),
(gst_ogm_video_parse_base_init), (gst_ogm_parse_init),
(gst_ogm_audio_parse_init), (gst_ogm_video_parse_init),
(gst_ogm_parse_sink_convert), (gst_ogm_parse_chain),
(gst_ogm_parse_change_state):
Separate between audio/video so ogmaudioparse actually uses the
audio pad templates. Both audio and video work now, including
autoplugging. Also use sometimes-srcpad hack.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_seek):
Handle events better. Don't hang on infinite loops.
* gst/avi/gstavidemux.c: (gst_avi_demux_class_init),
(gst_avi_demux_init), (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_stream_header), (gst_avi_demux_stream_data),
(gst_avi_demux_change_state):
* gst/avi/gstavidemux.h:
Improve A/V sync. Still not perfect.
* gst/matroska/ebml-read.c: (gst_ebml_read_seek),
(gst_ebml_read_skip):
Handle events better.
* gst/qtdemux/qtdemux.c: (gst_qtdemux_handle_sink_event),
(gst_qtdemux_loop_header), (qtdemux_parse_trak),
(qtdemux_audio_caps):
Add IMA4. Improve event handling. Save offset after a seek when
the headers are at the end of the file so that we don't end up in
an infinite loop.
* gst/typefind/gsttypefindfunctions.c: (qt_type_find):
Add low-priority typefind support for files with no length.
2004-09-23 14:59:22 +00:00
|
|
|
|
if (stream->strh->type == GST_RIFF_FCC_vids &&
|
2005-11-11 16:48:58 +00:00
|
|
|
|
(src_format == GST_FORMAT_BYTES || *dest_format == GST_FORMAT_BYTES)) {
|
|
|
|
|
res = FALSE;
|
|
|
|
|
goto done;
|
|
|
|
|
}
|
2003-07-06 20:49:52 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
switch (src_format) {
|
|
|
|
|
case GST_FORMAT_TIME:
|
|
|
|
|
switch (*dest_format) {
|
2004-03-15 19:32:27 +00:00
|
|
|
|
case GST_FORMAT_BYTES:
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
*dest_value = src_value * stream->strf.auds->av_bps / GST_SECOND;
|
2004-03-15 19:32:27 +00:00
|
|
|
|
break;
|
|
|
|
|
case GST_FORMAT_DEFAULT:
|
|
|
|
|
*dest_value = src_value * stream->strh->rate /
|
|
|
|
|
(stream->strh->scale * GST_SECOND);
|
|
|
|
|
break;
|
|
|
|
|
default:
|
|
|
|
|
res = FALSE;
|
|
|
|
|
break;
|
2003-07-06 20:49:52 +00:00
|
|
|
|
}
|
2002-01-31 22:22:42 +00:00
|
|
|
|
break;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
case GST_FORMAT_BYTES:
|
|
|
|
|
switch (*dest_format) {
|
2004-03-15 19:32:27 +00:00
|
|
|
|
case GST_FORMAT_TIME:
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (stream->strf.auds->av_bps != 0) {
|
|
|
|
|
*dest_value = ((gfloat) src_value) * GST_SECOND /
|
|
|
|
|
stream->strf.auds->av_bps;
|
|
|
|
|
} else
|
|
|
|
|
res = FALSE;
|
2004-03-15 19:32:27 +00:00
|
|
|
|
break;
|
|
|
|
|
default:
|
|
|
|
|
res = FALSE;
|
|
|
|
|
break;
|
2003-07-06 20:49:52 +00:00
|
|
|
|
}
|
|
|
|
|
break;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
case GST_FORMAT_DEFAULT:
|
|
|
|
|
switch (*dest_format) {
|
2004-03-15 19:32:27 +00:00
|
|
|
|
case GST_FORMAT_TIME:
|
|
|
|
|
*dest_value = ((((gfloat) src_value) * stream->strh->scale) /
|
|
|
|
|
stream->strh->rate) * GST_SECOND;
|
|
|
|
|
break;
|
|
|
|
|
default:
|
|
|
|
|
res = FALSE;
|
|
|
|
|
break;
|
2003-07-06 20:49:52 +00:00
|
|
|
|
}
|
2003-01-18 16:13:29 +00:00
|
|
|
|
break;
|
|
|
|
|
default:
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
res = FALSE;
|
2003-07-06 20:49:52 +00:00
|
|
|
|
}
|
2001-12-22 23:22:30 +00:00
|
|
|
|
|
2005-11-11 16:48:58 +00:00
|
|
|
|
done:
|
|
|
|
|
GST_LOG_OBJECT (avidemux, "Returning res:%d dest_format:%d dest_value:%lld",
|
|
|
|
|
res, *dest_format, *dest_value);
|
|
|
|
|
gst_object_unref (avidemux);
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
return res;
|
2003-07-06 20:49:52 +00:00
|
|
|
|
}
|
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
static const GstQueryType *
|
2004-03-14 22:34:33 +00:00
|
|
|
|
gst_avi_demux_get_src_query_types (GstPad * pad)
|
2001-12-22 23:22:30 +00:00
|
|
|
|
{
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
static const GstQueryType src_types[] = {
|
|
|
|
|
GST_QUERY_POSITION,
|
2005-10-21 16:15:57 +00:00
|
|
|
|
GST_QUERY_DURATION,
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
0
|
|
|
|
|
};
|
2001-12-22 23:22:30 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
return src_types;
|
2002-05-26 21:59:22 +00:00
|
|
|
|
}
|
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
static gboolean
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_handle_src_query (GstPad * pad, GstQuery * query)
|
2001-12-22 23:22:30 +00:00
|
|
|
|
{
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
gboolean res = TRUE;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GstAviDemux *demux = GST_AVI_DEMUX (GST_PAD_PARENT (pad));
|
2004-03-14 22:34:33 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
avi_stream_context *stream = gst_pad_get_element_private (pad);
|
2002-05-26 21:59:22 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (!stream->strh || !stream->strf.data)
|
|
|
|
|
return FALSE;
|
|
|
|
|
|
|
|
|
|
switch (GST_QUERY_TYPE (query)) {
|
|
|
|
|
case GST_QUERY_POSITION:{
|
2005-10-19 15:57:04 +00:00
|
|
|
|
gint64 pos = 0;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
|
|
|
|
|
if (stream->strh->type == GST_RIFF_FCC_auds) {
|
|
|
|
|
if (!stream->strh->samplesize) {
|
|
|
|
|
pos = GST_SECOND * stream->current_frame *
|
|
|
|
|
stream->strh->scale / stream->strh->rate;
|
|
|
|
|
} else if (stream->strf.auds->av_bps != 0) {
|
|
|
|
|
pos = ((gfloat) stream->current_byte) * GST_SECOND /
|
|
|
|
|
stream->strf.auds->av_bps;
|
|
|
|
|
} else if (stream->total_frames != 0 && stream->total_bytes != 0) {
|
|
|
|
|
/* calculate timestamps based on video size */
|
|
|
|
|
guint64 xlen = demux->avih->us_frame *
|
|
|
|
|
demux->avih->tot_frames * GST_USECOND;
|
|
|
|
|
|
|
|
|
|
if (!stream->strh->samplesize)
|
|
|
|
|
pos = xlen * stream->current_frame / stream->total_frames;
|
2004-03-15 19:32:27 +00:00
|
|
|
|
else
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
pos = xlen * stream->current_byte / stream->total_bytes;
|
|
|
|
|
} else {
|
2004-03-15 19:32:27 +00:00
|
|
|
|
res = FALSE;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
}
|
|
|
|
|
} else {
|
|
|
|
|
if (stream->strh->rate != 0) {
|
|
|
|
|
pos = ((gfloat) stream->current_frame * stream->strh->scale *
|
|
|
|
|
GST_SECOND / stream->strh->rate);
|
|
|
|
|
} else {
|
|
|
|
|
pos = stream->current_frame * demux->avih->us_frame * GST_USECOND;
|
|
|
|
|
}
|
2002-05-26 21:59:22 +00:00
|
|
|
|
}
|
2005-06-30 14:28:14 +00:00
|
|
|
|
if (res)
|
2005-10-19 15:57:04 +00:00
|
|
|
|
gst_query_set_position (query, GST_FORMAT_TIME, pos);
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
case GST_QUERY_DURATION:
|
|
|
|
|
{
|
|
|
|
|
gint64 len;
|
|
|
|
|
|
|
|
|
|
len = (((gfloat) stream->strh->scale) * stream->strh->length /
|
|
|
|
|
stream->strh->rate) * GST_SECOND;
|
|
|
|
|
gst_query_set_duration (query, GST_FORMAT_TIME, len);
|
2002-05-26 21:59:22 +00:00
|
|
|
|
break;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
}
|
2002-05-26 21:59:22 +00:00
|
|
|
|
default:
|
2006-02-06 12:18:45 +00:00
|
|
|
|
res = gst_pad_query_default (pad, query);
|
2002-05-26 21:59:22 +00:00
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return res;
|
|
|
|
|
}
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
#if 0
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
static const GstEventMask *
|
2004-03-14 22:34:33 +00:00
|
|
|
|
gst_avi_demux_get_event_mask (GstPad * pad)
|
2002-07-24 21:41:38 +00:00
|
|
|
|
{
|
|
|
|
|
static const GstEventMask masks[] = {
|
2004-03-14 22:34:33 +00:00
|
|
|
|
{GST_EVENT_SEEK, GST_SEEK_METHOD_SET | GST_SEEK_FLAG_KEY_UNIT},
|
|
|
|
|
{0,}
|
2002-07-24 21:41:38 +00:00
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
return masks;
|
|
|
|
|
}
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
#endif
|
2004-03-14 22:34:33 +00:00
|
|
|
|
|
2002-05-26 21:59:22 +00:00
|
|
|
|
static gboolean
|
2004-03-14 22:34:33 +00:00
|
|
|
|
gst_avi_demux_handle_src_event (GstPad * pad, GstEvent * event)
|
2002-05-26 21:59:22 +00:00
|
|
|
|
{
|
|
|
|
|
gboolean res = TRUE;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GstAviDemux *avi = GST_AVI_DEMUX (GST_PAD_PARENT (pad));
|
2002-05-26 21:59:22 +00:00
|
|
|
|
avi_stream_context *stream;
|
2004-03-14 22:34:33 +00:00
|
|
|
|
|
2005-01-13 17:11:08 +00:00
|
|
|
|
GST_CAT_DEBUG_OBJECT (GST_CAT_EVENT, avi,
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
"have event type %s: %p on src pad", GST_EVENT_TYPE_NAME (event), event);
|
2005-01-13 17:11:08 +00:00
|
|
|
|
if (!avi->index_entries) {
|
|
|
|
|
GST_CAT_DEBUG_OBJECT (GST_CAT_EVENT, avi, "no index entries, returning");
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
return FALSE;
|
2005-01-13 17:11:08 +00:00
|
|
|
|
}
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
|
2002-05-26 21:59:22 +00:00
|
|
|
|
stream = gst_pad_get_element_private (pad);
|
|
|
|
|
|
|
|
|
|
switch (GST_EVENT_TYPE (event)) {
|
|
|
|
|
case GST_EVENT_SEEK:
|
Various event updates and cleanups.
Original commit message from CVS:
* ext/amrnb/amrnbparse.c: (gst_amrnbparse_event),
(gst_amrnbparse_loop):
* ext/dv/gstdvdec.c: (gst_dvdec_handle_sink_event),
(gst_dvdec_handle_src_event), (gst_dvdec_decode_frame):
* ext/mad/gstid3tag.c: (gst_id3_tag_src_event),
(gst_id3_tag_sink_event), (gst_id3_tag_chain):
* ext/mad/gstmad.c: (gst_mad_src_query), (index_seek),
(normal_seek), (gst_mad_sink_event), (gst_mad_chain):
* ext/mpeg2dec/gstmpeg2dec.c:
* ext/shout2/gstshout2.c: (gst_shout2send_event):
* ext/sidplay/gstsiddec.cc:
* gst/avi/gstavidemux.c: (gst_avi_demux_handle_src_event),
(gst_avi_demux_send_event), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry):
* gst/goom/gstgoom.c: (gst_goom_event):
* gst/realmedia/rmdemux.c: (gst_rmdemux_sink_event),
(gst_rmdemux_chain), (gst_rmdemux_send_event),
(gst_rmdemux_add_stream):
* gst/wavparse/gstwavparse.c: (gst_wavparse_handle_seek),
(gst_wavparse_stream_headers), (gst_wavparse_stream_data),
(gst_wavparse_loop), (gst_wavparse_srcpad_event):
Various event updates and cleanups.
2005-07-27 18:37:25 +00:00
|
|
|
|
{
|
2006-02-15 10:15:47 +00:00
|
|
|
|
/* FIXME, this seeking code is not correct, look at wavparse for
|
|
|
|
|
* a better example */
|
Various event updates and cleanups.
Original commit message from CVS:
* ext/amrnb/amrnbparse.c: (gst_amrnbparse_event),
(gst_amrnbparse_loop):
* ext/dv/gstdvdec.c: (gst_dvdec_handle_sink_event),
(gst_dvdec_handle_src_event), (gst_dvdec_decode_frame):
* ext/mad/gstid3tag.c: (gst_id3_tag_src_event),
(gst_id3_tag_sink_event), (gst_id3_tag_chain):
* ext/mad/gstmad.c: (gst_mad_src_query), (index_seek),
(normal_seek), (gst_mad_sink_event), (gst_mad_chain):
* ext/mpeg2dec/gstmpeg2dec.c:
* ext/shout2/gstshout2.c: (gst_shout2send_event):
* ext/sidplay/gstsiddec.cc:
* gst/avi/gstavidemux.c: (gst_avi_demux_handle_src_event),
(gst_avi_demux_send_event), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry):
* gst/goom/gstgoom.c: (gst_goom_event):
* gst/realmedia/rmdemux.c: (gst_rmdemux_sink_event),
(gst_rmdemux_chain), (gst_rmdemux_send_event),
(gst_rmdemux_add_stream):
* gst/wavparse/gstwavparse.c: (gst_wavparse_handle_seek),
(gst_wavparse_stream_headers), (gst_wavparse_stream_data),
(gst_wavparse_loop), (gst_wavparse_srcpad_event):
Various event updates and cleanups.
2005-07-27 18:37:25 +00:00
|
|
|
|
GstFormat format;
|
|
|
|
|
GstSeekFlags flags;
|
2005-11-11 16:48:58 +00:00
|
|
|
|
gdouble rate;
|
|
|
|
|
gint64 start, stop;
|
|
|
|
|
gint64 tstart, tstop;
|
|
|
|
|
gint64 duration;
|
|
|
|
|
GstFormat tformat = GST_FORMAT_TIME;
|
|
|
|
|
GstSeekType start_type, stop_type;
|
|
|
|
|
gboolean update_start = TRUE;
|
|
|
|
|
gboolean update_stop = TRUE;
|
|
|
|
|
|
|
|
|
|
gst_event_parse_seek (event, &rate, &format, &flags, &start_type, &start,
|
|
|
|
|
&stop_type, &stop);
|
|
|
|
|
|
2005-12-07 11:40:46 +00:00
|
|
|
|
GST_DEBUG_OBJECT (avi,
|
|
|
|
|
"seek format %d, flags:%d, %08x, start:%lld, stop:%lld", format,
|
|
|
|
|
flags, stream->strh->type, start, stop);
|
2005-11-11 16:48:58 +00:00
|
|
|
|
|
|
|
|
|
if (format != GST_FORMAT_TIME) {
|
|
|
|
|
res &=
|
|
|
|
|
gst_avi_demux_src_convert (pad, format, start, &tformat, &tstart);
|
|
|
|
|
res &= gst_avi_demux_src_convert (pad, format, stop, &tformat, &tstop);
|
|
|
|
|
if (!res)
|
|
|
|
|
goto done;
|
|
|
|
|
} else {
|
|
|
|
|
tstart = start;
|
|
|
|
|
tstop = stop;
|
|
|
|
|
}
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
2005-11-11 16:48:58 +00:00
|
|
|
|
duration = (((gfloat) stream->strh->scale) * stream->strh->length /
|
|
|
|
|
stream->strh->rate) * GST_SECOND;
|
2004-03-15 19:32:27 +00:00
|
|
|
|
|
2005-11-11 16:48:58 +00:00
|
|
|
|
switch (start_type) {
|
|
|
|
|
case GST_SEEK_TYPE_CUR:
|
|
|
|
|
tstart = avi->segment_start + tstart;
|
|
|
|
|
break;
|
|
|
|
|
case GST_SEEK_TYPE_END:
|
|
|
|
|
tstart = duration + tstart;
|
|
|
|
|
break;
|
|
|
|
|
case GST_SEEK_TYPE_NONE:
|
|
|
|
|
tstart = avi->segment_start;
|
|
|
|
|
update_start = FALSE;
|
|
|
|
|
break;
|
|
|
|
|
case GST_SEEK_TYPE_SET:
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
tstart = CLAMP (tstart, 0, duration);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
|
2005-11-11 16:48:58 +00:00
|
|
|
|
switch (stop_type) {
|
|
|
|
|
case GST_SEEK_TYPE_CUR:
|
|
|
|
|
tstop = avi->segment_stop + tstop;
|
2004-03-15 19:32:27 +00:00
|
|
|
|
break;
|
2005-11-11 16:48:58 +00:00
|
|
|
|
case GST_SEEK_TYPE_END:
|
|
|
|
|
tstop = duration + tstop;
|
|
|
|
|
break;
|
|
|
|
|
case GST_SEEK_TYPE_NONE:
|
|
|
|
|
tstop = avi->segment_stop;
|
|
|
|
|
update_stop = FALSE;
|
|
|
|
|
break;
|
|
|
|
|
case GST_SEEK_TYPE_SET:
|
2004-03-15 19:32:27 +00:00
|
|
|
|
break;
|
2002-05-26 21:59:22 +00:00
|
|
|
|
}
|
2005-11-11 16:48:58 +00:00
|
|
|
|
tstop = CLAMP (tstop, 0, duration);
|
|
|
|
|
|
|
|
|
|
/* now store the values */
|
|
|
|
|
avi->segment_rate = rate;
|
|
|
|
|
avi->segment_flags = flags;
|
|
|
|
|
avi->segment_start = tstart;
|
|
|
|
|
avi->segment_stop = tstop;
|
|
|
|
|
|
|
|
|
|
gst_avi_demux_handle_seek (avi, update_start || update_stop);
|
2002-05-26 21:59:22 +00:00
|
|
|
|
break;
|
2005-11-11 16:48:58 +00:00
|
|
|
|
|
Various event updates and cleanups.
Original commit message from CVS:
* ext/amrnb/amrnbparse.c: (gst_amrnbparse_event),
(gst_amrnbparse_loop):
* ext/dv/gstdvdec.c: (gst_dvdec_handle_sink_event),
(gst_dvdec_handle_src_event), (gst_dvdec_decode_frame):
* ext/mad/gstid3tag.c: (gst_id3_tag_src_event),
(gst_id3_tag_sink_event), (gst_id3_tag_chain):
* ext/mad/gstmad.c: (gst_mad_src_query), (index_seek),
(normal_seek), (gst_mad_sink_event), (gst_mad_chain):
* ext/mpeg2dec/gstmpeg2dec.c:
* ext/shout2/gstshout2.c: (gst_shout2send_event):
* ext/sidplay/gstsiddec.cc:
* gst/avi/gstavidemux.c: (gst_avi_demux_handle_src_event),
(gst_avi_demux_send_event), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry):
* gst/goom/gstgoom.c: (gst_goom_event):
* gst/realmedia/rmdemux.c: (gst_rmdemux_sink_event),
(gst_rmdemux_chain), (gst_rmdemux_send_event),
(gst_rmdemux_add_stream):
* gst/wavparse/gstwavparse.c: (gst_wavparse_handle_seek),
(gst_wavparse_stream_headers), (gst_wavparse_stream_data),
(gst_wavparse_loop), (gst_wavparse_srcpad_event):
Various event updates and cleanups.
2005-07-27 18:37:25 +00:00
|
|
|
|
}
|
2002-05-26 21:59:22 +00:00
|
|
|
|
default:
|
|
|
|
|
res = FALSE;
|
|
|
|
|
break;
|
|
|
|
|
}
|
2002-07-08 19:40:37 +00:00
|
|
|
|
|
2005-11-11 16:48:58 +00:00
|
|
|
|
done:
|
2002-07-08 19:40:37 +00:00
|
|
|
|
gst_event_unref (event);
|
|
|
|
|
|
2002-05-26 21:59:22 +00:00
|
|
|
|
return res;
|
|
|
|
|
}
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
/**
|
|
|
|
|
* gst_avi_demux_parse_file_header:
|
|
|
|
|
* @element: caller element (used for errors/debug).
|
|
|
|
|
* @buf: input data to be used for parsing.
|
|
|
|
|
*
|
|
|
|
|
* "Open" a RIFF/AVI file. The buffer should be at least 12
|
|
|
|
|
* bytes long. Discards buffer after use.
|
|
|
|
|
*
|
|
|
|
|
* Returns: TRUE if the file is a RIFF/AVI file, FALSE otherwise.
|
|
|
|
|
* Throws an error, caller should error out (fatal).
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
*/
|
|
|
|
|
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
static gboolean
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_parse_file_header (GstElement * element, GstBuffer * buf)
|
2001-12-22 23:22:30 +00:00
|
|
|
|
{
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
guint32 doctype;
|
2001-12-22 23:22:30 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (!gst_riff_parse_file_header (element, buf, &doctype))
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
return FALSE;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
if (doctype != GST_RIFF_RIFF_AVI) {
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_ELEMENT_ERROR (element, STREAM, WRONG_TYPE, (NULL),
|
2005-11-21 13:38:24 +00:00
|
|
|
|
("File is not an AVI file: %" GST_FOURCC_FORMAT,
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_FOURCC_ARGS (doctype)));
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
return FALSE;
|
|
|
|
|
}
|
2001-12-22 23:22:30 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
return TRUE;
|
|
|
|
|
}
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
static GstFlowReturn
|
|
|
|
|
gst_avi_demux_stream_init (GstAviDemux * avi)
|
|
|
|
|
{
|
|
|
|
|
GstFlowReturn res;
|
|
|
|
|
GstBuffer *buf = NULL;
|
|
|
|
|
|
|
|
|
|
if ((res = gst_pad_pull_range (avi->sinkpad,
|
|
|
|
|
avi->offset, 12, &buf)) != GST_FLOW_OK)
|
|
|
|
|
return res;
|
|
|
|
|
else if (!gst_avi_demux_parse_file_header (GST_ELEMENT (avi), buf))
|
|
|
|
|
return GST_FLOW_ERROR;
|
|
|
|
|
|
|
|
|
|
avi->offset += 12;
|
|
|
|
|
|
|
|
|
|
return GST_FLOW_OK;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
|
* gst_avi_demux_parse_avih:
|
|
|
|
|
* @element: caller element (used for errors/debug).
|
|
|
|
|
* @buf: input data to be used for parsing.
|
|
|
|
|
* @avih: pointer to structure (filled in by function) containing
|
|
|
|
|
* stream information (such as flags, number of streams, etc.).
|
|
|
|
|
*
|
|
|
|
|
* Read 'avih' header. Discards buffer after use.
|
|
|
|
|
*
|
|
|
|
|
* Returns: TRUE on success, FALSE otherwise. Throws an error if
|
|
|
|
|
* the header is invalid. The caller should error out
|
|
|
|
|
* (fatal).
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
*/
|
|
|
|
|
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
static gboolean
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_parse_avih (GstElement * element,
|
|
|
|
|
GstBuffer * buf, gst_riff_avih ** _avih)
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
{
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_riff_avih *avih;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (!buf || GST_BUFFER_SIZE (buf) < sizeof (gst_riff_avih)) {
|
|
|
|
|
GST_ELEMENT_ERROR (element, STREAM, DEMUX, (NULL),
|
|
|
|
|
("Too small avih (%d available, %d needed)",
|
|
|
|
|
GST_BUFFER_SIZE (buf), (int) sizeof (gst_riff_avih)));
|
|
|
|
|
if (buf)
|
|
|
|
|
gst_buffer_unref (buf);
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
return FALSE;
|
|
|
|
|
}
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
avih = g_memdup (GST_BUFFER_DATA (buf), GST_BUFFER_SIZE (buf));
|
|
|
|
|
|
|
|
|
|
#if (G_BYTE_ORDER == G_BIG_ENDIAN)
|
|
|
|
|
avih->us_frame = GUINT32_FROM_LE (avih->us_frame);
|
|
|
|
|
avih->max_bps = GUINT32_FROM_LE (avih->max_bps);
|
|
|
|
|
avih->pad_gran = GUINT32_FROM_LE (avih->pad_gran);
|
|
|
|
|
avih->flags = GUINT32_FROM_LE (avih->flags);
|
|
|
|
|
avih->tot_frames = GUINT32_FROM_LE (avih->tot_frames);
|
|
|
|
|
avih->init_frames = GUINT32_FROM_LE (avih->init_frames);
|
|
|
|
|
avih->streams = GUINT32_FROM_LE (avih->streams);
|
|
|
|
|
avih->bufsize = GUINT32_FROM_LE (avih->bufsize);
|
|
|
|
|
avih->width = GUINT32_FROM_LE (avih->width);
|
|
|
|
|
avih->height = GUINT32_FROM_LE (avih->height);
|
|
|
|
|
avih->scale = GUINT32_FROM_LE (avih->scale);
|
|
|
|
|
avih->rate = GUINT32_FROM_LE (avih->rate);
|
|
|
|
|
avih->start = GUINT32_FROM_LE (avih->start);
|
|
|
|
|
avih->length = GUINT32_FROM_LE (avih->length);
|
|
|
|
|
#endif
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
|
|
|
|
/* debug stuff */
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_INFO_OBJECT (element, "avih tag found:");
|
|
|
|
|
GST_INFO_OBJECT (element, " us_frame %u", avih->us_frame);
|
|
|
|
|
GST_INFO_OBJECT (element, " max_bps %u", avih->max_bps);
|
|
|
|
|
GST_INFO_OBJECT (element, " pad_gran %u", avih->pad_gran);
|
|
|
|
|
GST_INFO_OBJECT (element, " flags 0x%08x", avih->flags);
|
|
|
|
|
GST_INFO_OBJECT (element, " tot_frames %u", avih->tot_frames);
|
|
|
|
|
GST_INFO_OBJECT (element, " init_frames %u", avih->init_frames);
|
|
|
|
|
GST_INFO_OBJECT (element, " streams %u", avih->streams);
|
|
|
|
|
GST_INFO_OBJECT (element, " bufsize %u", avih->bufsize);
|
|
|
|
|
GST_INFO_OBJECT (element, " width %u", avih->width);
|
|
|
|
|
GST_INFO_OBJECT (element, " height %u", avih->height);
|
|
|
|
|
GST_INFO_OBJECT (element, " scale %u", avih->scale);
|
|
|
|
|
GST_INFO_OBJECT (element, " rate %u", avih->rate);
|
|
|
|
|
GST_INFO_OBJECT (element, " start %u", avih->start);
|
|
|
|
|
GST_INFO_OBJECT (element, " length %u", avih->length);
|
|
|
|
|
|
|
|
|
|
*_avih = avih;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
gst_buffer_unref (buf);
|
|
|
|
|
|
|
|
|
|
return TRUE;
|
|
|
|
|
}
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
/**
|
|
|
|
|
* gst_avi_demux_parse_superindex:
|
|
|
|
|
* @element: caller element (used for debugging/errors).
|
|
|
|
|
* @buf: input data to use for parsing.
|
|
|
|
|
* @locations: locations in the file (byte-offsets) that contain
|
|
|
|
|
* the actual indexes (see get_avi_demux_parse_subindex()).
|
|
|
|
|
* The array ends with GST_BUFFER_OFFSET_NONE.
|
|
|
|
|
*
|
|
|
|
|
* Reads superindex (openDML-2 spec stuff) from the provided data.
|
|
|
|
|
*
|
|
|
|
|
* Returns: TRUE on success, FALSE otherwise. Indexes should be skipped
|
|
|
|
|
* on error, but they are not fatal.
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
static gboolean
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_parse_superindex (GstElement * element,
|
|
|
|
|
GstBuffer * buf, guint64 ** _indexes)
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
{
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
guint8 *data = GST_BUFFER_DATA (buf);
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
gint bpe = 16, num, i;
|
|
|
|
|
guint64 *indexes;
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
*_indexes = NULL;
|
|
|
|
|
|
|
|
|
|
if (!buf || GST_BUFFER_SIZE (buf) < 24) {
|
|
|
|
|
GST_ERROR_OBJECT (element,
|
|
|
|
|
"Not enough data to parse superindex (%d available, %d needed)",
|
|
|
|
|
GST_BUFFER_SIZE (buf), 24);
|
|
|
|
|
if (buf)
|
|
|
|
|
gst_buffer_unref (buf);
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
return FALSE;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* check type of index. The opendml2 specs state that
|
|
|
|
|
* there should be 4 dwords per array entry. Type can be
|
|
|
|
|
* either frame or field (and we don't care). */
|
|
|
|
|
if (GST_READ_UINT16_LE (data) != 4 ||
|
|
|
|
|
(data[2] & 0xfe) != 0x0 || data[3] != 0x0) {
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_WARNING_OBJECT (element,
|
|
|
|
|
"Superindex for stream %d has unexpected "
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
"size_entry %d (bytes) or flags 0x%02x/0x%02x",
|
|
|
|
|
GST_READ_UINT16_LE (data), data[2], data[3]);
|
|
|
|
|
bpe = GST_READ_UINT16_LE (data) * 4;
|
|
|
|
|
}
|
|
|
|
|
num = GST_READ_UINT32_LE (&data[4]);
|
|
|
|
|
|
|
|
|
|
indexes = g_new (guint64, num + 1);
|
|
|
|
|
for (i = 0; i < num; i++) {
|
|
|
|
|
if (GST_BUFFER_SIZE (buf) < 24 + bpe * (i + 1))
|
|
|
|
|
break;
|
|
|
|
|
indexes[i] = GST_READ_UINT64_LE (&data[24 + bpe * i]);
|
|
|
|
|
}
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
indexes[i] = GST_BUFFER_OFFSET_NONE;
|
|
|
|
|
*_indexes = indexes;
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
|
|
|
|
|
gst_buffer_unref (buf);
|
|
|
|
|
|
|
|
|
|
return TRUE;
|
|
|
|
|
}
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
/**
|
|
|
|
|
* gst_avi_demux_parse_subindex:
|
|
|
|
|
* @element: caller element (used for errors/debug).
|
|
|
|
|
* @buf: input data to use for parsing.
|
|
|
|
|
* @stream: stream context.
|
|
|
|
|
* @entries_list: a list (returned by the function) containing all the
|
|
|
|
|
* indexes parsed in this specific subindex. The first
|
|
|
|
|
* entry is also a pointer to allocated memory that needs
|
|
|
|
|
* to be free´ed. May be NULL if no supported indexes were
|
|
|
|
|
* found.
|
|
|
|
|
*
|
|
|
|
|
* Reads superindex (openDML-2 spec stuff) from the provided data.
|
|
|
|
|
* The buffer will be discarded after use.
|
|
|
|
|
*
|
|
|
|
|
* Returns: TRUE on success, FALSE otherwise. Errors are fatal, we
|
|
|
|
|
* throw an error, caller should bail out asap.
|
|
|
|
|
*/
|
|
|
|
|
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
static gboolean
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_parse_subindex (GstElement * element,
|
|
|
|
|
GstBuffer * buf, avi_stream_context * stream, GList ** _entries_list)
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
{
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
guint8 *data = GST_BUFFER_DATA (buf);
|
|
|
|
|
gint bpe, num, x;
|
|
|
|
|
guint64 baseoff;
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
gst_avi_index_entry *entries, *entry;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GList *entries_list = NULL;
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
GstFormat format = GST_FORMAT_TIME;
|
2005-07-01 08:10:56 +00:00
|
|
|
|
gint64 tmp;
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
/* check size */
|
|
|
|
|
if (!buf || GST_BUFFER_SIZE (buf) < 24) {
|
|
|
|
|
GST_ERROR_OBJECT (element,
|
|
|
|
|
"Not enough data to parse subindex (%d available, %d needed)",
|
|
|
|
|
GST_BUFFER_SIZE (buf), 24);
|
|
|
|
|
if (buf)
|
|
|
|
|
gst_buffer_unref (buf);
|
|
|
|
|
*_entries_list = NULL;
|
|
|
|
|
return TRUE; /* continue */
|
|
|
|
|
}
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
/* We don't support index-data yet */
|
|
|
|
|
if (data[3] & 0x80) {
|
|
|
|
|
GST_ELEMENT_ERROR (element, STREAM, NOT_IMPLEMENTED, (NULL),
|
|
|
|
|
("Subindex-is-data is not implemented"));
|
|
|
|
|
gst_buffer_unref (buf);
|
|
|
|
|
return FALSE;
|
|
|
|
|
}
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
/* check type of index. The opendml2 specs state that
|
|
|
|
|
* there should be 4 dwords per array entry. Type can be
|
|
|
|
|
* either frame or field (and we don't care). */
|
|
|
|
|
bpe = (data[2] & 0x01) ? 12 : 8;
|
|
|
|
|
if (GST_READ_UINT16_LE (data) != bpe / 4 ||
|
|
|
|
|
(data[2] & 0xfe) != 0x0 || data[3] != 0x1) {
|
|
|
|
|
GST_WARNING_OBJECT (element,
|
|
|
|
|
"Superindex for stream %d has unexpected "
|
|
|
|
|
"size_entry %d (bytes) or flags 0x%02x/0x%02x",
|
|
|
|
|
GST_READ_UINT16_LE (data), data[2], data[3]);
|
|
|
|
|
bpe = GST_READ_UINT16_LE (data) * 4;
|
|
|
|
|
}
|
|
|
|
|
num = GST_READ_UINT32_LE (&data[4]);
|
|
|
|
|
baseoff = GST_READ_UINT64_LE (&data[12]);
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
entries = g_new (gst_avi_index_entry, num);
|
|
|
|
|
for (x = 0; x < num; x++) {
|
|
|
|
|
entry = &entries[x];
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (GST_BUFFER_SIZE (buf) < 24 + bpe * (x + 1))
|
|
|
|
|
break;
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
/* fill in */
|
|
|
|
|
entry->offset = baseoff + GST_READ_UINT32_LE (&data[24 + bpe * x]);
|
|
|
|
|
entry->size = GST_READ_UINT32_LE (&data[24 + bpe * x + 4]);
|
|
|
|
|
entry->flags = (entry->size & 0x80000000) ? 0 : GST_RIFF_IF_KEYFRAME;
|
|
|
|
|
entry->size &= ~0x80000000;
|
|
|
|
|
entry->index_nr = x;
|
|
|
|
|
entry->stream_nr = stream->num;
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
/* timestamps */
|
|
|
|
|
if (stream->strh->samplesize && stream->strh->type == GST_RIFF_FCC_auds) {
|
|
|
|
|
/* constant rate stream */
|
|
|
|
|
gst_avi_demux_src_convert (stream->pad, GST_FORMAT_BYTES,
|
2005-07-01 08:10:56 +00:00
|
|
|
|
stream->total_bytes, &format, &tmp);
|
|
|
|
|
entry->ts = tmp;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_src_convert (stream->pad, GST_FORMAT_BYTES,
|
2005-07-01 08:10:56 +00:00
|
|
|
|
stream->total_bytes + entry->size, &format, &tmp);
|
|
|
|
|
entry->dur = tmp;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
} else {
|
|
|
|
|
/* VBR stream */
|
|
|
|
|
gst_avi_demux_src_convert (stream->pad, GST_FORMAT_DEFAULT,
|
2005-07-01 08:10:56 +00:00
|
|
|
|
stream->total_frames, &format, &tmp);
|
|
|
|
|
entry->ts = tmp;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_src_convert (stream->pad, GST_FORMAT_DEFAULT,
|
2005-07-01 08:10:56 +00:00
|
|
|
|
stream->total_frames + 1, &format, &tmp);
|
|
|
|
|
entry->dur = tmp;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
}
|
|
|
|
|
entry->dur -= entry->ts;
|
|
|
|
|
|
|
|
|
|
/* stream position */
|
|
|
|
|
entry->bytes_before = stream->total_bytes;
|
|
|
|
|
stream->total_bytes += entry->size;
|
|
|
|
|
entry->frames_before = stream->total_frames;
|
|
|
|
|
stream->total_frames++;
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
entries_list = g_list_prepend (entries_list, entry);
|
|
|
|
|
}
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_LOG_OBJECT (element, "Read %d index entries", x);
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_buffer_unref (buf);
|
|
|
|
|
|
|
|
|
|
if (x > 0) {
|
|
|
|
|
*_entries_list = g_list_reverse (entries_list);
|
|
|
|
|
} else {
|
|
|
|
|
*_entries_list = NULL;
|
|
|
|
|
g_free (entries);
|
|
|
|
|
}
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
return TRUE;
|
|
|
|
|
}
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
static void
|
|
|
|
|
gst_avi_demux_read_subindexes (GstAviDemux * avi,
|
|
|
|
|
GList ** index, GList ** alloc_list)
|
|
|
|
|
{
|
|
|
|
|
GList *list;
|
|
|
|
|
guint32 tag;
|
|
|
|
|
GstBuffer *buf;
|
|
|
|
|
gint i, n;
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
for (n = 0; n < avi->num_streams; n++) {
|
|
|
|
|
avi_stream_context *stream = &avi->stream[n];
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
for (i = 0; stream->indexes[i] != GST_BUFFER_OFFSET_NONE; i++) {
|
|
|
|
|
if (gst_riff_read_chunk (GST_ELEMENT (avi), avi->sinkpad,
|
|
|
|
|
&stream->indexes[i], &tag, &buf) != GST_FLOW_OK)
|
|
|
|
|
continue;
|
|
|
|
|
else if (tag != GST_MAKE_FOURCC ('i', 'x', '0' + stream->num / 10,
|
|
|
|
|
'0' + stream->num % 10)) {
|
|
|
|
|
GST_ERROR_OBJECT (GST_ELEMENT (avi),
|
2005-11-21 13:38:24 +00:00
|
|
|
|
"Not an ix## chunk (%" GST_FOURCC_FORMAT ")",
|
|
|
|
|
GST_FOURCC_ARGS (tag));
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_buffer_unref (buf);
|
|
|
|
|
continue;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (!gst_avi_demux_parse_subindex (GST_ELEMENT (avi), buf, stream, &list))
|
|
|
|
|
continue;
|
|
|
|
|
if (list) {
|
|
|
|
|
*alloc_list = g_list_append (*alloc_list, list->data);
|
|
|
|
|
*index = g_list_concat (*index, list);
|
|
|
|
|
}
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
g_free (stream->indexes);
|
|
|
|
|
stream->indexes = NULL;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
/**
|
|
|
|
|
* gst_avi_demux_parse_stream:
|
|
|
|
|
* @element: calling element (used for debugging/errors).
|
|
|
|
|
* @buf: input buffer used to parse the stream.
|
|
|
|
|
*
|
|
|
|
|
* Parses all subchunks in a strl chunk (which defines a single
|
|
|
|
|
* stream). Discards the buffer after use. This function will
|
|
|
|
|
* increment the stream counter internally.
|
|
|
|
|
*
|
|
|
|
|
* Returns: whether the stream was identified successfully.
|
|
|
|
|
* Errors are not fatal. It does indicate the stream
|
|
|
|
|
* was skipped.
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
static gboolean
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_parse_stream (GstElement * element, GstBuffer * buf)
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
{
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GstAviDemux *avi = GST_AVI_DEMUX (element);
|
|
|
|
|
avi_stream_context *stream = &avi->stream[avi->num_streams];
|
|
|
|
|
GstElementClass *klass = GST_ELEMENT_GET_CLASS (element);
|
|
|
|
|
GstPadTemplate *templ;
|
|
|
|
|
GstBuffer *sub = NULL;
|
|
|
|
|
guint offset = 4;
|
|
|
|
|
guint32 tag = 0;
|
|
|
|
|
gchar *codec_name = NULL, *padname = NULL;
|
|
|
|
|
const gchar *tag_name;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
GstCaps *caps = NULL;
|
|
|
|
|
GstPad *pad;
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_DEBUG_OBJECT (element, "Parsing stream");
|
|
|
|
|
|
|
|
|
|
/* read strh */
|
|
|
|
|
if (!buf || !gst_riff_parse_chunk (element, buf, &offset, &tag, &sub) ||
|
|
|
|
|
tag != GST_RIFF_TAG_strh) {
|
|
|
|
|
GST_ERROR_OBJECT (element,
|
2005-11-21 13:38:24 +00:00
|
|
|
|
"Failed to find strh chunk (tag: %" GST_FOURCC_FORMAT ")",
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
buf ? GST_BUFFER_SIZE (buf) : 0, GST_FOURCC_ARGS (tag));
|
|
|
|
|
goto fail;
|
|
|
|
|
} else if (!gst_riff_parse_strh (element, sub, &stream->strh))
|
|
|
|
|
goto fail;
|
|
|
|
|
|
|
|
|
|
/* read strf */
|
|
|
|
|
if (!gst_riff_parse_chunk (element, buf, &offset, &tag, &sub) ||
|
|
|
|
|
tag != GST_RIFF_TAG_strf) {
|
|
|
|
|
GST_ERROR_OBJECT (element,
|
2005-11-21 13:38:24 +00:00
|
|
|
|
"Failed to find strh chunk (size: %d, tag: %"
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_FOURCC_FORMAT ")", buf ? GST_BUFFER_SIZE (buf) : 0,
|
|
|
|
|
GST_FOURCC_ARGS (tag));
|
|
|
|
|
goto fail;
|
|
|
|
|
} else {
|
|
|
|
|
gboolean res = FALSE;
|
2002-05-26 21:59:22 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
switch (stream->strh->type) {
|
|
|
|
|
case GST_RIFF_FCC_vids:
|
|
|
|
|
res = gst_riff_parse_strf_vids (element, sub,
|
|
|
|
|
&stream->strf.vids, &stream->extradata);
|
|
|
|
|
break;
|
|
|
|
|
case GST_RIFF_FCC_auds:
|
|
|
|
|
res = gst_riff_parse_strf_auds (element, sub,
|
|
|
|
|
&stream->strf.auds, &stream->extradata);
|
|
|
|
|
break;
|
|
|
|
|
case GST_RIFF_FCC_iavs:
|
|
|
|
|
res = gst_riff_parse_strf_iavs (element, sub,
|
|
|
|
|
&stream->strf.iavs, &stream->extradata);
|
|
|
|
|
break;
|
|
|
|
|
default:
|
|
|
|
|
GST_ERROR_OBJECT (element,
|
2005-11-21 13:38:24 +00:00
|
|
|
|
"Don´t know how to handle stream type %" GST_FOURCC_FORMAT,
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_FOURCC_ARGS (stream->strh->type));
|
|
|
|
|
break;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
2002-05-26 21:59:22 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (!res)
|
|
|
|
|
goto fail;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* read strd/strn */
|
|
|
|
|
while (gst_riff_parse_chunk (element, buf, &offset, &tag, &sub)) {
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
switch (tag) {
|
gst/: Add MS RLE support. I added some functions to read out strf chunks into strf chunks and the data behind it. Thi...
Original commit message from CVS:
reviewed by: <delete if not using a buddy>
* gst-libs/gst/riff/riff-media.c:
(gst_riff_create_video_caps_with_data),
(gst_riff_create_video_caps), (gst_riff_create_audio_caps),
(gst_riff_create_video_template_caps),
(gst_riff_create_audio_template_caps):
* gst-libs/gst/riff/riff-media.h:
* gst-libs/gst/riff/riff-read.c:
(gst_riff_read_strf_vids_with_data), (gst_riff_read_strf_vids):
* gst-libs/gst/riff/riff-read.h:
* gst/avi/gstavidemux.c: (gst_avi_demux_add_stream):
Add MS RLE support. I added some functions to read out strf chunks
into strf chunks and the data behind it. This is usually color
palettes (as in RLE, but also in 8-bit RGB). Also use those during
caps creation. Lastly, add ADPCM (similar to wavparse - which
should eventually be rifflib based).
* gst/matroska/matroska-demux.c: (gst_matroska_demux_class_init),
(gst_matroska_demux_init), (gst_matroska_demux_reset):
* gst/matroska/matroska-demux.h:
Remove placeholders for some prehistoric tagging system. Didn't add
support for any tag system really anyway.
* gst/qtdemux/qtdemux.c:
Add support for audio/x-m4a (MPEG-4) through spider.
* gst/wavparse/gstwavparse.c: (gst_wavparse_parse_fmt),
(gst_wavparse_loop):
ADPCM support (#135862). Increase max. buffer size because we
cannot split buffers for ADPCM (screws references) and I've seen
files with 2048 byte chunks. 4096 seems safe for now.
2004-04-16 01:20:44 +00:00
|
|
|
|
case GST_RIFF_TAG_strd:
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (stream->initdata)
|
|
|
|
|
gst_buffer_unref (stream->initdata);
|
|
|
|
|
stream->initdata = sub;
|
gst/: Add MS RLE support. I added some functions to read out strf chunks into strf chunks and the data behind it. Thi...
Original commit message from CVS:
reviewed by: <delete if not using a buddy>
* gst-libs/gst/riff/riff-media.c:
(gst_riff_create_video_caps_with_data),
(gst_riff_create_video_caps), (gst_riff_create_audio_caps),
(gst_riff_create_video_template_caps),
(gst_riff_create_audio_template_caps):
* gst-libs/gst/riff/riff-media.h:
* gst-libs/gst/riff/riff-read.c:
(gst_riff_read_strf_vids_with_data), (gst_riff_read_strf_vids):
* gst-libs/gst/riff/riff-read.h:
* gst/avi/gstavidemux.c: (gst_avi_demux_add_stream):
Add MS RLE support. I added some functions to read out strf chunks
into strf chunks and the data behind it. This is usually color
palettes (as in RLE, but also in 8-bit RGB). Also use those during
caps creation. Lastly, add ADPCM (similar to wavparse - which
should eventually be rifflib based).
* gst/matroska/matroska-demux.c: (gst_matroska_demux_class_init),
(gst_matroska_demux_init), (gst_matroska_demux_reset):
* gst/matroska/matroska-demux.h:
Remove placeholders for some prehistoric tagging system. Didn't add
support for any tag system really anyway.
* gst/qtdemux/qtdemux.c:
Add support for audio/x-m4a (MPEG-4) through spider.
* gst/wavparse/gstwavparse.c: (gst_wavparse_parse_fmt),
(gst_wavparse_loop):
ADPCM support (#135862). Increase max. buffer size because we
cannot split buffers for ADPCM (screws references) and I've seen
files with 2048 byte chunks. 4096 seems safe for now.
2004-04-16 01:20:44 +00:00
|
|
|
|
break;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
case GST_RIFF_TAG_strn:
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
g_free (stream->name);
|
2005-08-04 18:31:12 +00:00
|
|
|
|
stream->name = g_new (gchar, GST_BUFFER_SIZE (sub) + 1);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
memcpy (stream->name, GST_BUFFER_DATA (sub), GST_BUFFER_SIZE (sub));
|
|
|
|
|
stream->name[GST_BUFFER_SIZE (sub)] = '\0';
|
|
|
|
|
gst_buffer_unref (sub);
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
sub = NULL;
|
2004-03-15 19:32:27 +00:00
|
|
|
|
break;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
default:
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
if (tag == GST_MAKE_FOURCC ('i', 'n', 'd', 'x') ||
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
tag == GST_MAKE_FOURCC ('i', 'x', '0' + avi->num_streams / 10,
|
|
|
|
|
'0' + avi->num_streams % 10)) {
|
|
|
|
|
g_free (stream->indexes);
|
|
|
|
|
gst_avi_demux_parse_superindex (element, sub, &stream->indexes);
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
break;
|
|
|
|
|
}
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_WARNING_OBJECT (element,
|
2005-11-21 13:38:24 +00:00
|
|
|
|
"Unknown stream header tag %" GST_FOURCC_FORMAT ", ignoring",
|
2004-03-15 19:32:27 +00:00
|
|
|
|
GST_FOURCC_ARGS (tag));
|
|
|
|
|
/* fall-through */
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
case GST_RIFF_TAG_JUNK:
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_buffer_unref (sub);
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
sub = NULL;
|
2004-03-15 19:32:27 +00:00
|
|
|
|
break;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
/* we now have all info, let´s set up a pad and a caps and be done */
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
/* create stream name + pad */
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
switch (stream->strh->type) {
|
2005-09-26 13:06:27 +00:00
|
|
|
|
case GST_RIFF_FCC_vids:{
|
|
|
|
|
guint32 fourcc;
|
|
|
|
|
|
|
|
|
|
fourcc = (stream->strf.vids->compression) ?
|
|
|
|
|
stream->strf.vids->compression : stream->strh->fcc_handler;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
padname = g_strdup_printf ("video_%02d", avi->num_v_streams);
|
|
|
|
|
templ = gst_element_class_get_pad_template (klass, "video_%02d");
|
2005-09-26 13:06:27 +00:00
|
|
|
|
caps = gst_riff_create_video_caps (fourcc, stream->strh,
|
|
|
|
|
stream->strf.vids, stream->extradata, stream->initdata, &codec_name);
|
|
|
|
|
if (!caps) {
|
|
|
|
|
caps = gst_caps_new_simple ("video/x-avi-unknown", "fourcc",
|
|
|
|
|
GST_TYPE_FOURCC, fourcc, NULL);
|
|
|
|
|
}
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
tag_name = GST_TAG_VIDEO_CODEC;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
avi->num_v_streams++;
|
|
|
|
|
break;
|
2005-09-26 13:06:27 +00:00
|
|
|
|
}
|
|
|
|
|
case GST_RIFF_FCC_auds:{
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
padname = g_strdup_printf ("audio_%02d", avi->num_a_streams);
|
|
|
|
|
templ = gst_element_class_get_pad_template (klass, "audio_%02d");
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
caps = gst_riff_create_audio_caps (stream->strf.auds->format,
|
2005-09-26 13:06:27 +00:00
|
|
|
|
stream->strh, stream->strf.auds, stream->extradata,
|
|
|
|
|
stream->initdata, &codec_name);
|
|
|
|
|
if (!caps) {
|
|
|
|
|
caps = gst_caps_new_simple ("audio/x-avi-unknown", "codec_id",
|
|
|
|
|
G_TYPE_INT, stream->strf.auds->format, NULL);
|
|
|
|
|
}
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
tag_name = GST_TAG_AUDIO_CODEC;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
avi->num_a_streams++;
|
|
|
|
|
break;
|
2005-09-26 13:06:27 +00:00
|
|
|
|
}
|
|
|
|
|
case GST_RIFF_FCC_iavs:{
|
|
|
|
|
guint32 fourcc = stream->strh->fcc_handler;
|
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
padname = g_strdup_printf ("video_%02d", avi->num_v_streams);
|
|
|
|
|
templ = gst_element_class_get_pad_template (klass, "video_%02d");
|
2005-09-26 13:06:27 +00:00
|
|
|
|
caps = gst_riff_create_iavs_caps (fourcc, stream->strh,
|
|
|
|
|
stream->strf.iavs, stream->extradata, stream->initdata, &codec_name);
|
|
|
|
|
if (!caps) {
|
|
|
|
|
caps = gst_caps_new_simple ("video/x-avi-unknown", "fourcc",
|
|
|
|
|
GST_TYPE_FOURCC, fourcc, NULL);
|
|
|
|
|
}
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
tag_name = GST_TAG_VIDEO_CODEC;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
avi->num_v_streams++;
|
|
|
|
|
break;
|
2005-09-26 13:06:27 +00:00
|
|
|
|
}
|
2001-12-22 23:22:30 +00:00
|
|
|
|
default:
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
g_assert_not_reached ();
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* no caps means no stream */
|
|
|
|
|
if (!caps) {
|
|
|
|
|
GST_ERROR_OBJECT (element, "Did not find caps for stream %s", padname);
|
|
|
|
|
goto fail;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* set proper settings and add it */
|
2005-12-07 11:40:46 +00:00
|
|
|
|
if (stream->pad)
|
|
|
|
|
gst_object_unref (stream->pad);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
pad = stream->pad = gst_pad_new_from_template (templ, padname);
|
2006-02-04 15:41:43 +00:00
|
|
|
|
stream->last_flow = GST_FLOW_OK;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
g_free (padname);
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_pad_use_fixed_caps (pad);
|
|
|
|
|
#if 0
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
gst_pad_set_formats_function (pad, gst_avi_demux_get_src_formats);
|
|
|
|
|
gst_pad_set_event_mask_function (pad, gst_avi_demux_get_event_mask);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
#endif
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
gst_pad_set_event_function (pad, gst_avi_demux_handle_src_event);
|
|
|
|
|
gst_pad_set_query_type_function (pad, gst_avi_demux_get_src_query_types);
|
|
|
|
|
gst_pad_set_query_function (pad, gst_avi_demux_handle_src_query);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
#if 0
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
gst_pad_set_convert_function (pad, gst_avi_demux_src_convert);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
#endif
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
|
|
|
|
stream->num = avi->num_streams;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
stream->total_bytes = 0;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
stream->total_frames = 0;
|
|
|
|
|
stream->current_frame = 0;
|
|
|
|
|
stream->current_byte = 0;
|
|
|
|
|
stream->current_entry = -1;
|
|
|
|
|
gst_pad_set_element_private (pad, stream);
|
|
|
|
|
avi->num_streams++;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_pad_set_caps (pad, caps);
|
2005-11-07 10:31:32 +00:00
|
|
|
|
gst_pad_set_active (pad, TRUE);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_element_add_pad (GST_ELEMENT (avi), pad);
|
2005-11-07 10:31:32 +00:00
|
|
|
|
GST_LOG_OBJECT (element, "Added pad %s with caps %" GST_PTR_FORMAT,
|
|
|
|
|
GST_PAD_NAME (pad), caps);
|
2005-01-10 16:09:25 +00:00
|
|
|
|
|
2005-08-16 16:30:23 +00:00
|
|
|
|
if (codec_name) {
|
|
|
|
|
GstTagList *list = gst_tag_list_new ();
|
|
|
|
|
|
|
|
|
|
gst_tag_list_add (list, GST_TAG_MERGE_APPEND, tag_name, codec_name, NULL);
|
|
|
|
|
gst_element_found_tags_for_pad (GST_ELEMENT (avi), pad, list);
|
|
|
|
|
g_free (codec_name);
|
|
|
|
|
}
|
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
return TRUE;
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
fail:
|
|
|
|
|
/* unref any mem that may be in use */
|
|
|
|
|
if (buf)
|
|
|
|
|
gst_buffer_unref (buf);
|
|
|
|
|
if (sub)
|
|
|
|
|
gst_buffer_unref (sub);
|
|
|
|
|
g_free (stream->strh);
|
|
|
|
|
g_free (stream->strf.data);
|
|
|
|
|
g_free (stream->name);
|
|
|
|
|
g_free (stream->indexes);
|
|
|
|
|
if (stream->initdata)
|
|
|
|
|
gst_buffer_unref (stream->initdata);
|
|
|
|
|
if (stream->extradata)
|
|
|
|
|
gst_buffer_unref (stream->extradata);
|
|
|
|
|
memset (stream, 0, sizeof (avi_stream_context));
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
avi->num_streams++;
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
return FALSE;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
/**
|
|
|
|
|
* gst_avi_demux_parse_odml:
|
|
|
|
|
* @element: calling element (used for debug/error).
|
|
|
|
|
* @buf: input buffer to be used for parsing.
|
|
|
|
|
*
|
|
|
|
|
* Read an openDML-2.0 extension header. Fills in the frame number
|
|
|
|
|
* in the avi demuxer object when reading succeeds.
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
*/
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
static void
|
|
|
|
|
gst_avi_demux_parse_odml (GstElement * element, GstBuffer * buf)
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
{
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GstAviDemux *avi = GST_AVI_DEMUX (element);
|
|
|
|
|
guint32 tag = 0;
|
|
|
|
|
guint offset = 4;
|
|
|
|
|
GstBuffer *sub = NULL;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
while (gst_riff_parse_chunk (element, buf, &offset, &tag, &sub)) {
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
switch (tag) {
|
2004-03-14 22:34:33 +00:00
|
|
|
|
case GST_RIFF_TAG_dmlh:{
|
2004-03-15 19:32:27 +00:00
|
|
|
|
gst_riff_dmlh dmlh, *_dmlh;
|
|
|
|
|
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
if (sub && (GST_BUFFER_SIZE (sub) < sizeof (gst_riff_dmlh))) {
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_ERROR_OBJECT (element,
|
|
|
|
|
"DMLH entry is too small (%d bytes, %d needed)",
|
2004-03-15 19:32:27 +00:00
|
|
|
|
GST_BUFFER_SIZE (buf), (int) sizeof (gst_riff_dmlh));
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_buffer_unref (sub);
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
sub = NULL;
|
2004-03-15 19:32:27 +00:00
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
_dmlh = (gst_riff_dmlh *) GST_BUFFER_DATA (buf);
|
|
|
|
|
dmlh.totalframes = GUINT32_FROM_LE (_dmlh->totalframes);
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_INFO_OBJECT (element, "dmlh tag found:");
|
|
|
|
|
GST_INFO_OBJECT (element, " totalframes: %u", dmlh.totalframes);
|
2004-03-15 19:32:27 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
avi->avih->tot_frames = dmlh.totalframes;
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
if (sub) {
|
|
|
|
|
gst_buffer_unref (sub);
|
|
|
|
|
sub = NULL;
|
|
|
|
|
}
|
2004-03-15 19:32:27 +00:00
|
|
|
|
break;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
default:
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_WARNING_OBJECT (element,
|
2005-11-21 13:38:24 +00:00
|
|
|
|
"Unknown tag %" GST_FOURCC_FORMAT " in ODML header",
|
2004-03-15 19:32:27 +00:00
|
|
|
|
GST_FOURCC_ARGS (tag));
|
|
|
|
|
/* fall-through */
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
case GST_RIFF_TAG_JUNK:
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
if (sub) {
|
|
|
|
|
gst_buffer_unref (sub);
|
|
|
|
|
sub = NULL;
|
|
|
|
|
}
|
2004-03-15 19:32:27 +00:00
|
|
|
|
break;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
if (buf)
|
|
|
|
|
gst_buffer_unref (buf);
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
/**
|
|
|
|
|
* gst_avi_demux_parse_index:
|
|
|
|
|
* @element: calling element (used for debugging/errors).
|
|
|
|
|
* @buf: buffer containing the full index.
|
|
|
|
|
* @entries_list: list (returned by this function) containing the index
|
|
|
|
|
* entries parsed from the buffer. The first in the list
|
|
|
|
|
* is also a pointer to the allocated data and should be
|
|
|
|
|
* free'ed at some point.
|
|
|
|
|
*
|
|
|
|
|
* Read index entries from the provided buffer.
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
*/
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
static void
|
|
|
|
|
gst_avi_demux_parse_index (GstElement * element,
|
|
|
|
|
GstBuffer * buf, GList ** _entries_list)
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
{
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GstAviDemux *avi = GST_AVI_DEMUX (element);
|
|
|
|
|
guint64 pos_before = avi->offset;
|
|
|
|
|
gst_avi_index_entry *entries = NULL;
|
|
|
|
|
guint8 *data;
|
|
|
|
|
GList *entries_list = NULL;
|
|
|
|
|
guint i, num, n;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (!buf) {
|
|
|
|
|
*_entries_list = NULL;
|
|
|
|
|
return;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
data = GST_BUFFER_DATA (buf);
|
|
|
|
|
num = GST_BUFFER_SIZE (buf) / sizeof (gst_riff_index_entry);
|
|
|
|
|
entries = g_new (gst_avi_index_entry, num);
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
for (i = 0, n = 0; i < num; i++) {
|
2004-01-25 00:43:16 +00:00
|
|
|
|
gst_riff_index_entry entry, *_entry;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
avi_stream_context *stream;
|
|
|
|
|
gint stream_nr;
|
|
|
|
|
gst_avi_index_entry *target;
|
|
|
|
|
GstFormat format;
|
2005-07-01 08:10:56 +00:00
|
|
|
|
gint64 ts;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
_entry = &((gst_riff_index_entry *) data)[i];
|
2004-03-14 22:34:33 +00:00
|
|
|
|
entry.id = GUINT32_FROM_LE (_entry->id);
|
2004-01-25 00:43:16 +00:00
|
|
|
|
entry.offset = GUINT32_FROM_LE (_entry->offset);
|
2004-03-14 22:34:33 +00:00
|
|
|
|
entry.flags = GUINT32_FROM_LE (_entry->flags);
|
|
|
|
|
entry.size = GUINT32_FROM_LE (_entry->size);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
target = &entries[n];
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
2004-12-23 10:53:27 +00:00
|
|
|
|
if (entry.id == GST_RIFF_rec || entry.id == 0 ||
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
(entry.offset == 0 && n > 0))
|
2004-04-11 21:36:45 +00:00
|
|
|
|
continue;
|
|
|
|
|
|
2004-01-23 17:45:54 +00:00
|
|
|
|
stream_nr = CHUNKID_TO_STREAMNR (entry.id);
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
if (stream_nr >= avi->num_streams || stream_nr < 0) {
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_WARNING_OBJECT (element,
|
|
|
|
|
"Index entry %d has invalid stream nr %d", i, stream_nr);
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
continue;
|
|
|
|
|
}
|
|
|
|
|
target->stream_nr = stream_nr;
|
|
|
|
|
stream = &avi->stream[stream_nr];
|
|
|
|
|
|
|
|
|
|
target->index_nr = i;
|
2004-03-14 22:34:33 +00:00
|
|
|
|
target->flags = entry.flags;
|
|
|
|
|
target->size = entry.size;
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
target->offset = entry.offset + 8;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
|
|
|
|
/* figure out if the index is 0 based or relative to the MOVI start */
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (n == 0) {
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
if (target->offset < pos_before)
|
2004-03-15 19:32:27 +00:00
|
|
|
|
avi->index_offset = pos_before + 8;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
else
|
2004-03-15 19:32:27 +00:00
|
|
|
|
avi->index_offset = 0;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
target->bytes_before = stream->total_bytes;
|
|
|
|
|
target->frames_before = stream->total_frames;
|
|
|
|
|
|
|
|
|
|
format = GST_FORMAT_TIME;
|
|
|
|
|
if (stream->strh->type == GST_RIFF_FCC_auds) {
|
|
|
|
|
/* all audio frames are keyframes */
|
|
|
|
|
target->flags |= GST_RIFF_IF_KEYFRAME;
|
|
|
|
|
}
|
2004-03-14 22:34:33 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
if (stream->strh->samplesize && stream->strh->type == GST_RIFF_FCC_auds) {
|
|
|
|
|
/* constant rate stream */
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_src_convert (stream->pad, GST_FORMAT_BYTES,
|
2005-07-01 08:10:56 +00:00
|
|
|
|
stream->total_bytes, &format, &ts);
|
|
|
|
|
target->ts = ts;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_src_convert (stream->pad, GST_FORMAT_BYTES,
|
2005-07-01 08:10:56 +00:00
|
|
|
|
stream->total_bytes + target->size, &format, &ts);
|
|
|
|
|
target->dur = ts;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
} else {
|
|
|
|
|
/* VBR stream */
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_src_convert (stream->pad, GST_FORMAT_DEFAULT,
|
2005-07-01 08:10:56 +00:00
|
|
|
|
stream->total_frames, &format, &ts);
|
|
|
|
|
target->ts = ts;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_src_convert (stream->pad, GST_FORMAT_DEFAULT,
|
2005-07-01 08:10:56 +00:00
|
|
|
|
stream->total_frames + 1, &format, &ts);
|
|
|
|
|
target->dur = ts;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
target->dur -= target->ts;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
|
|
|
|
stream->total_bytes += target->size;
|
|
|
|
|
stream->total_frames++;
|
2004-10-04 13:32:20 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_DEBUG ("Adding index entry %d (%d) for stream %d of size %u "
|
2004-11-29 12:41:03 +00:00
|
|
|
|
"at offset %" G_GUINT64_FORMAT " and time %" GST_TIME_FORMAT,
|
|
|
|
|
target->index_nr, stream->total_frames - 1,
|
|
|
|
|
target->stream_nr, target->size, target->offset,
|
|
|
|
|
GST_TIME_ARGS (target->ts));
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
entries_list = g_list_prepend (entries_list, target);
|
|
|
|
|
|
|
|
|
|
n++;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
*_entries_list = g_list_reverse (entries_list);
|
|
|
|
|
gst_buffer_unref (buf);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
|
* gst_avi_demux_stream_index:
|
|
|
|
|
* @avi: avi demuxer object.
|
|
|
|
|
* @index: list of index entries, returned by this function.
|
|
|
|
|
* @alloc_list: list of allocated data, returned by this function.
|
|
|
|
|
*
|
|
|
|
|
* Seeks to index and reads it.
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
static void
|
|
|
|
|
gst_avi_demux_stream_index (GstAviDemux * avi,
|
|
|
|
|
GList ** index, GList ** alloc_list)
|
|
|
|
|
{
|
|
|
|
|
guint64 offset = avi->offset;
|
|
|
|
|
GstBuffer *buf;
|
|
|
|
|
guint32 tag;
|
|
|
|
|
gint i;
|
|
|
|
|
|
|
|
|
|
*alloc_list = NULL;
|
|
|
|
|
*index = NULL;
|
|
|
|
|
|
|
|
|
|
/* get position */
|
|
|
|
|
if (gst_pad_pull_range (avi->sinkpad, offset, 8, &buf) != GST_FLOW_OK)
|
|
|
|
|
return;
|
|
|
|
|
else if (!buf || GST_BUFFER_SIZE (buf) < 8) {
|
|
|
|
|
if (buf)
|
|
|
|
|
gst_buffer_unref (buf);
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
offset += 8 + GST_READ_UINT32_LE (GST_BUFFER_DATA (buf) + 4);
|
|
|
|
|
gst_buffer_unref (buf);
|
|
|
|
|
|
|
|
|
|
/* get size */
|
|
|
|
|
if (gst_riff_read_chunk (GST_ELEMENT (avi),
|
|
|
|
|
avi->sinkpad, &offset, &tag, &buf) != GST_FLOW_OK)
|
|
|
|
|
return;
|
|
|
|
|
else if (tag != GST_RIFF_TAG_idx1) {
|
|
|
|
|
GST_ERROR_OBJECT (avi,
|
2005-11-21 13:38:24 +00:00
|
|
|
|
"No index data after movi chunk, but %" GST_FOURCC_FORMAT,
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_FOURCC_ARGS (tag));
|
|
|
|
|
gst_buffer_unref (buf);
|
|
|
|
|
return;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_parse_index (GST_ELEMENT (avi), buf, index);
|
|
|
|
|
if (*index)
|
|
|
|
|
*alloc_list = g_list_append (*alloc_list, (*index)->data);
|
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
/* debug our indexes */
|
|
|
|
|
for (i = 0; i < avi->num_streams; i++) {
|
|
|
|
|
avi_stream_context *stream;
|
|
|
|
|
|
|
|
|
|
stream = &avi->stream[i];
|
2005-01-13 17:11:08 +00:00
|
|
|
|
GST_DEBUG_OBJECT (avi, "stream %u: %u frames, %" G_GINT64_FORMAT " bytes",
|
2004-03-15 19:32:27 +00:00
|
|
|
|
i, stream->total_frames, stream->total_bytes);
|
2001-12-22 23:22:30 +00:00
|
|
|
|
}
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
return;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
2002-02-03 20:10:04 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
#if 0
|
|
|
|
|
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
/*
|
|
|
|
|
* Sync to next data chunk.
|
|
|
|
|
*/
|
|
|
|
|
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
static gboolean
|
|
|
|
|
gst_avi_demux_skip (GstAviDemux * avi, gboolean prevent_eos)
|
|
|
|
|
{
|
|
|
|
|
GstRiffRead *riff = GST_RIFF_READ (avi);
|
|
|
|
|
|
|
|
|
|
if (prevent_eos) {
|
|
|
|
|
guint64 pos, length;
|
|
|
|
|
guint size;
|
|
|
|
|
guint8 *data;
|
|
|
|
|
|
|
|
|
|
pos = gst_bytestream_tell (riff->bs);
|
|
|
|
|
length = gst_bytestream_length (riff->bs);
|
|
|
|
|
|
|
|
|
|
if (pos + 8 > length)
|
|
|
|
|
return FALSE;
|
|
|
|
|
|
|
|
|
|
if (gst_bytestream_peek_bytes (riff->bs, &data, 8) != 8)
|
|
|
|
|
return FALSE;
|
|
|
|
|
|
|
|
|
|
size = GST_READ_UINT32_LE (&data[4]);
|
|
|
|
|
if (size & 1)
|
|
|
|
|
size++;
|
|
|
|
|
|
|
|
|
|
/* Note, we're going to skip which might involve seeks. Therefore,
|
|
|
|
|
* we need 1 byte more! */
|
|
|
|
|
if (pos + 8 + size >= length)
|
|
|
|
|
return FALSE;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return gst_riff_read_skip (riff);
|
|
|
|
|
}
|
|
|
|
|
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
static gboolean
|
|
|
|
|
gst_avi_demux_sync (GstAviDemux * avi, guint32 * ret_tag, gboolean prevent_eos)
|
|
|
|
|
{
|
|
|
|
|
GstRiffRead *riff = GST_RIFF_READ (avi);
|
|
|
|
|
guint32 tag;
|
|
|
|
|
guint64 length = gst_bytestream_length (riff->bs);
|
|
|
|
|
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
if (prevent_eos && gst_bytestream_tell (riff->bs) + 12 >= length)
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
return FALSE;
|
|
|
|
|
|
|
|
|
|
/* peek first (for the end of this 'list/movi' section) */
|
|
|
|
|
if (!(tag = gst_riff_peek_tag (riff, &avi->level_up)))
|
|
|
|
|
return FALSE;
|
|
|
|
|
|
|
|
|
|
/* if we're at top-level, we didn't read the 'movi'
|
|
|
|
|
* list tag yet. This can also be 'AVIX' in case of
|
|
|
|
|
* openDML-2.0 AVI files. Lastly, it might be idx1,
|
|
|
|
|
* in which case we skip it so we come at EOS. */
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
while (1) {
|
|
|
|
|
if (prevent_eos && gst_bytestream_tell (riff->bs) + 12 >= length)
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
return FALSE;
|
|
|
|
|
|
|
|
|
|
if (!(tag = gst_riff_peek_tag (riff, NULL)))
|
|
|
|
|
return FALSE;
|
|
|
|
|
|
|
|
|
|
switch (tag) {
|
|
|
|
|
case GST_RIFF_TAG_LIST:
|
|
|
|
|
if (!(tag = gst_riff_peek_list (riff)))
|
|
|
|
|
return FALSE;
|
|
|
|
|
|
|
|
|
|
switch (tag) {
|
|
|
|
|
case GST_RIFF_LIST_AVIX:
|
|
|
|
|
if (!gst_riff_read_list (riff, &tag))
|
|
|
|
|
return FALSE;
|
|
|
|
|
break;
|
|
|
|
|
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
case GST_RIFF_LIST_movi:
|
|
|
|
|
if (!gst_riff_read_list (riff, &tag))
|
|
|
|
|
return FALSE;
|
|
|
|
|
/* fall-through */
|
|
|
|
|
|
|
|
|
|
case GST_RIFF_rec:
|
|
|
|
|
goto done;
|
|
|
|
|
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
default:
|
2005-11-21 13:38:24 +00:00
|
|
|
|
GST_WARNING ("Unknown list %" GST_FOURCC_FORMAT " before AVI data",
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
GST_FOURCC_ARGS (tag));
|
|
|
|
|
/* fall-through */
|
|
|
|
|
|
|
|
|
|
case GST_RIFF_TAG_JUNK:
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
if (!gst_avi_demux_skip (avi, prevent_eos))
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
return FALSE;
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
default:
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
if ((tag & 0xff) >= '0' &&
|
|
|
|
|
(tag & 0xff) <= '9' &&
|
|
|
|
|
((tag >> 8) & 0xff) >= '0' && ((tag >> 8) & 0xff) <= '9') {
|
|
|
|
|
goto done;
|
|
|
|
|
}
|
|
|
|
|
/* pass-through */
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
|
|
|
|
|
case GST_RIFF_TAG_idx1:
|
|
|
|
|
case GST_RIFF_TAG_JUNK:
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
if (!gst_avi_demux_skip (avi, prevent_eos)) {
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
return FALSE;
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
}
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
}
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
done:
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
/* And then, we get the data */
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
if (prevent_eos && gst_bytestream_tell (riff->bs) + 12 >= length)
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
return FALSE;
|
|
|
|
|
|
|
|
|
|
if (!(tag = gst_riff_peek_tag (riff, NULL)))
|
|
|
|
|
return FALSE;
|
|
|
|
|
|
|
|
|
|
/* Support for rec-list files */
|
|
|
|
|
switch (tag) {
|
|
|
|
|
case GST_RIFF_TAG_LIST:
|
|
|
|
|
if (!(tag = gst_riff_peek_list (riff)))
|
|
|
|
|
return FALSE;
|
|
|
|
|
if (tag == GST_RIFF_rec) {
|
|
|
|
|
/* Simply skip the list */
|
|
|
|
|
if (!gst_riff_read_list (riff, &tag))
|
|
|
|
|
return FALSE;
|
|
|
|
|
if (!(tag = gst_riff_peek_tag (riff, NULL)))
|
|
|
|
|
return FALSE;
|
|
|
|
|
}
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case GST_RIFF_TAG_JUNK:
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
gst_avi_demux_skip (avi, prevent_eos);
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
return FALSE;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (ret_tag)
|
|
|
|
|
*ret_tag = tag;
|
|
|
|
|
|
|
|
|
|
return TRUE;
|
|
|
|
|
}
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
|
* gst_avi_demux_peek_tag:
|
|
|
|
|
*
|
|
|
|
|
* Returns the tag and size of the next chunk
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
static GstFlowReturn
|
|
|
|
|
gst_avi_demux_peek_tag (GstAviDemux * avi, guint64 offset, guint32 * tag,
|
|
|
|
|
guint * size)
|
|
|
|
|
{
|
|
|
|
|
GstFlowReturn res = GST_FLOW_OK;
|
|
|
|
|
GstBuffer *buf = NULL;
|
|
|
|
|
|
|
|
|
|
if ((res = gst_pad_pull_range (avi->sinkpad, offset, 8, &buf)) != GST_FLOW_OK)
|
|
|
|
|
goto beach;
|
|
|
|
|
if (GST_BUFFER_SIZE (buf) != 8)
|
|
|
|
|
return GST_FLOW_ERROR;
|
|
|
|
|
*tag = GST_READ_UINT32_LE (GST_BUFFER_DATA (buf));
|
|
|
|
|
*size = GST_READ_UINT32_LE (GST_BUFFER_DATA (buf) + 4);
|
|
|
|
|
gst_buffer_unref (buf);
|
|
|
|
|
|
|
|
|
|
GST_LOG_OBJECT (avi,
|
|
|
|
|
"Tag[%" GST_FOURCC_FORMAT "] (size:%d) %lld -- %lld",
|
|
|
|
|
GST_FOURCC_ARGS (*tag), *size, offset, offset + (guint64) * size);
|
|
|
|
|
beach:
|
|
|
|
|
return res;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
|
* gst_avi_demux_next_data_buffer:
|
|
|
|
|
*
|
|
|
|
|
* Returns the offset and size of the next buffer
|
|
|
|
|
* Position is the position of the buffer (after tag and size)
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
static GstFlowReturn
|
|
|
|
|
gst_avi_demux_next_data_buffer (GstAviDemux * avi, guint64 * offset,
|
|
|
|
|
guint32 * tag, guint * size)
|
|
|
|
|
{
|
|
|
|
|
guint64 off = *offset;
|
|
|
|
|
guint siz;
|
|
|
|
|
GstFlowReturn res = GST_FLOW_OK;
|
|
|
|
|
|
|
|
|
|
while (1) {
|
|
|
|
|
if ((res = gst_avi_demux_peek_tag (avi, off, tag, &siz)) != GST_FLOW_OK)
|
|
|
|
|
break;
|
|
|
|
|
if (*tag == GST_RIFF_TAG_LIST)
|
|
|
|
|
off += 8;
|
|
|
|
|
else if (*tag == GST_RIFF_LIST_movi)
|
|
|
|
|
off += 4;
|
|
|
|
|
else {
|
|
|
|
|
*offset = off + 8;
|
|
|
|
|
*size = siz;
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
return res;
|
|
|
|
|
}
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
/*
|
|
|
|
|
* Scan the file for all chunks to "create" a new index.
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
* Return value indicates if we can continue reading the stream. It
|
|
|
|
|
* does not say anything about whether we created an index.
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
*
|
|
|
|
|
* pull-range based
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
*/
|
|
|
|
|
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
static gboolean
|
2004-10-04 13:32:20 +00:00
|
|
|
|
gst_avi_demux_stream_scan (GstAviDemux * avi,
|
|
|
|
|
GList ** index, GList ** alloc_list)
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
{
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
GstFlowReturn res;
|
2004-10-04 13:32:20 +00:00
|
|
|
|
gst_avi_index_entry *entry, *entries = NULL;
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
avi_stream_context *stream;
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
GstFormat format = GST_FORMAT_BYTES;
|
|
|
|
|
guint64 pos = avi->offset;
|
|
|
|
|
guint64 length;
|
2005-12-19 15:43:30 +00:00
|
|
|
|
gint64 tmplength;
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
guint32 tag;
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
GstPad *peer;
|
2004-10-04 13:32:20 +00:00
|
|
|
|
GList *list = NULL;
|
|
|
|
|
guint index_size = 0;
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
|
|
|
|
|
/* FIXME:
|
|
|
|
|
* - implement non-seekable source support.
|
|
|
|
|
*/
|
|
|
|
|
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
GST_LOG_OBJECT (avi, "Creating index %s existing index",
|
|
|
|
|
(*index) ? "with" : "without");
|
|
|
|
|
|
|
|
|
|
if (!(peer = gst_pad_get_peer (avi->sinkpad)))
|
|
|
|
|
return FALSE;
|
2005-12-19 15:43:30 +00:00
|
|
|
|
if (!(gst_pad_query_duration (peer, &format, &tmplength)))
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
return FALSE;
|
2005-12-19 15:43:30 +00:00
|
|
|
|
length = tmplength;
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
gst_object_unref (peer);
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
if (*index) {
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
entry = g_list_last (*index)->data;
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
pos = entry->offset + avi->index_offset + entry->size;
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
if (entry->size & 1)
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
pos++;
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
if (pos < length) {
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
GST_LOG ("Incomplete index, seeking to last valid entry @ %"
|
|
|
|
|
G_GUINT64_FORMAT " of %" G_GUINT64_FORMAT " (%"
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
G_GUINT64_FORMAT "+%u)", pos, length, entry->offset, entry->size);
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
|
|
|
|
|
} else {
|
|
|
|
|
return TRUE;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
while (1) {
|
|
|
|
|
gint stream_nr;
|
|
|
|
|
guint size;
|
2005-12-19 15:43:30 +00:00
|
|
|
|
gint64 tmpts, tmpdur;
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
|
|
|
|
|
if ((res =
|
|
|
|
|
gst_avi_demux_next_data_buffer (avi, &pos, &tag,
|
|
|
|
|
&size)) != GST_FLOW_OK)
|
|
|
|
|
break;
|
|
|
|
|
stream_nr = CHUNKID_TO_STREAMNR (tag);
|
|
|
|
|
if (stream_nr < 0 || stream_nr >= avi->num_streams)
|
|
|
|
|
continue;
|
|
|
|
|
stream = &avi->stream[stream_nr];
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
/* pre-allocate */
|
|
|
|
|
if (index_size % 1024 == 0) {
|
|
|
|
|
entries = g_new (gst_avi_index_entry, 1024);
|
|
|
|
|
*alloc_list = g_list_prepend (*alloc_list, entries);
|
|
|
|
|
}
|
|
|
|
|
entry = &entries[index_size % 1024];
|
|
|
|
|
|
|
|
|
|
entry->index_nr = index_size++;
|
|
|
|
|
entry->stream_nr = stream_nr;
|
|
|
|
|
entry->flags = GST_RIFF_IF_KEYFRAME;
|
|
|
|
|
entry->offset = pos - avi->index_offset;
|
|
|
|
|
entry->size = size;
|
|
|
|
|
|
|
|
|
|
/* timestamps */
|
|
|
|
|
if (stream->strh->samplesize && stream->strh->type == GST_RIFF_FCC_auds) {
|
|
|
|
|
format = GST_FORMAT_TIME;
|
|
|
|
|
/* constant rate stream */
|
|
|
|
|
gst_avi_demux_src_convert (stream->pad, GST_FORMAT_BYTES,
|
2005-12-19 15:43:30 +00:00
|
|
|
|
stream->total_bytes, &format, &tmpts);
|
|
|
|
|
entry->ts = tmpts;
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
gst_avi_demux_src_convert (stream->pad, GST_FORMAT_BYTES,
|
2005-12-19 15:43:30 +00:00
|
|
|
|
stream->total_bytes + entry->size, &format, &tmpdur);
|
|
|
|
|
entry->dur = tmpdur;
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
} else {
|
|
|
|
|
format = GST_FORMAT_TIME;
|
|
|
|
|
/* VBR stream */
|
|
|
|
|
gst_avi_demux_src_convert (stream->pad, GST_FORMAT_DEFAULT,
|
2005-12-19 15:43:30 +00:00
|
|
|
|
stream->total_frames, &format, &tmpts);
|
|
|
|
|
entry->ts = tmpts;
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
gst_avi_demux_src_convert (stream->pad, GST_FORMAT_DEFAULT,
|
2005-12-19 15:43:30 +00:00
|
|
|
|
stream->total_frames + 1, &format, &tmpdur);
|
|
|
|
|
entry->dur = tmpdur;
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
}
|
|
|
|
|
entry->dur -= entry->ts;
|
|
|
|
|
|
|
|
|
|
/* stream position */
|
|
|
|
|
entry->bytes_before = stream->total_bytes;
|
|
|
|
|
stream->total_bytes += entry->size;
|
|
|
|
|
entry->frames_before = stream->total_frames;
|
|
|
|
|
stream->total_frames++;
|
|
|
|
|
|
|
|
|
|
list = g_list_prepend (list, entry);
|
|
|
|
|
GST_DEBUG_OBJECT (avi, "Added index entry %d (in stream: %d), offset %"
|
|
|
|
|
G_GUINT64_FORMAT ", time %" GST_TIME_FORMAT " for stream %d",
|
|
|
|
|
index_size - 1, entry->frames_before, entry->offset,
|
|
|
|
|
GST_TIME_ARGS (entry->ts), entry->stream_nr);
|
|
|
|
|
|
|
|
|
|
/* update position */
|
|
|
|
|
pos += ((size + 1) & ~1);
|
|
|
|
|
if (pos > length) {
|
|
|
|
|
GST_WARNING_OBJECT (avi,
|
|
|
|
|
"Stopping index lookup since we are further than EOF");
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#if 0
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
while (gst_avi_demux_sync (avi, &tag, TRUE)) {
|
|
|
|
|
gint stream_nr = CHUNKID_TO_STREAMNR (tag);
|
|
|
|
|
guint8 *data;
|
|
|
|
|
GstFormat format = GST_FORMAT_TIME;
|
|
|
|
|
|
|
|
|
|
if (stream_nr < 0 || stream_nr >= avi->num_streams)
|
|
|
|
|
goto next;
|
|
|
|
|
stream = &avi->stream[stream_nr];
|
|
|
|
|
|
|
|
|
|
/* get chunk size */
|
|
|
|
|
if (gst_bytestream_peek_bytes (riff->bs, &data, 8) != 8)
|
|
|
|
|
goto next;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
|
|
|
|
|
/* fill in */
|
2004-10-04 13:32:20 +00:00
|
|
|
|
entry->index_nr = index_size++;
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
entry->stream_nr = stream_nr;
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
entry->flags = GST_RIFF_IF_KEYFRAME;
|
|
|
|
|
entry->offset = gst_bytestream_tell (riff->bs) + 8 - avi->index_offset;
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
entry->size = GST_READ_UINT32_LE (&data[4]);
|
|
|
|
|
|
|
|
|
|
/* timestamps */
|
|
|
|
|
if (stream->strh->samplesize && stream->strh->type == GST_RIFF_FCC_auds) {
|
|
|
|
|
/* constant rate stream */
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_src_convert (stream->pad, GST_FORMAT_BYTES,
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
stream->total_bytes, &format, &entry->ts);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_src_convert (stream->pad, GST_FORMAT_BYTES,
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
stream->total_bytes + entry->size, &format, &entry->dur);
|
|
|
|
|
} else {
|
|
|
|
|
/* VBR stream */
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_src_convert (stream->pad, GST_FORMAT_DEFAULT,
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
stream->total_frames, &format, &entry->ts);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_src_convert (stream->pad, GST_FORMAT_DEFAULT,
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
stream->total_frames + 1, &format, &entry->dur);
|
|
|
|
|
}
|
|
|
|
|
entry->dur -= entry->ts;
|
|
|
|
|
|
|
|
|
|
/* stream position */
|
|
|
|
|
entry->bytes_before = stream->total_bytes;
|
|
|
|
|
stream->total_bytes += entry->size;
|
|
|
|
|
entry->frames_before = stream->total_frames;
|
|
|
|
|
stream->total_frames++;
|
|
|
|
|
|
2004-10-04 13:32:20 +00:00
|
|
|
|
list = g_list_prepend (list, entry);
|
2005-01-13 17:11:08 +00:00
|
|
|
|
GST_DEBUG_OBJECT (avi, "Added index entry %d (in stream: %d), offset %"
|
2004-12-23 11:44:03 +00:00
|
|
|
|
G_GUINT64_FORMAT ", time %" GST_TIME_FORMAT " for stream %d",
|
|
|
|
|
index_size - 1, entry->frames_before, entry->offset,
|
|
|
|
|
GST_TIME_ARGS (entry->ts), entry->stream_nr);
|
2004-10-04 13:32:20 +00:00
|
|
|
|
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
next:
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
if (!gst_avi_demux_skip (avi, TRUE))
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
/* seek back */
|
2004-10-04 13:32:20 +00:00
|
|
|
|
if (!(event = gst_riff_read_seek (riff, pos))) {
|
|
|
|
|
g_list_free (list);
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
return FALSE;
|
2004-10-04 13:32:20 +00:00
|
|
|
|
}
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
gst_event_unref (event);
|
|
|
|
|
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
#endif
|
2004-10-04 13:32:20 +00:00
|
|
|
|
GST_LOG_OBJECT (avi, "index created, %d items", index_size);
|
|
|
|
|
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
*index = g_list_concat (*index, g_list_reverse (list));
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
|
|
|
|
return TRUE;
|
2002-01-15 22:49:11 +00:00
|
|
|
|
}
|
|
|
|
|
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
/*
|
|
|
|
|
* Massage index.
|
|
|
|
|
* We're going to go over each entry in the index and finetune
|
|
|
|
|
* some things we don't like about AVI. For example, a single
|
|
|
|
|
* chunk might be too long. Also, individual streams might be
|
|
|
|
|
* out-of-sync. In the first case, we cut the chunk in several
|
|
|
|
|
* smaller pieces. In the second case, we re-order chunk reading
|
|
|
|
|
* order. The end result should be a smoother playing AVI.
|
|
|
|
|
*/
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
static gint
|
2004-10-04 13:32:20 +00:00
|
|
|
|
sort (gst_avi_index_entry * a, gst_avi_index_entry * b)
|
|
|
|
|
{
|
|
|
|
|
if (a->ts > b->ts)
|
|
|
|
|
return 1;
|
|
|
|
|
else if (a->ts < b->ts)
|
|
|
|
|
return -1;
|
|
|
|
|
else
|
|
|
|
|
return a->stream_nr - b->stream_nr;
|
|
|
|
|
}
|
|
|
|
|
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
static void
|
2004-10-04 13:32:20 +00:00
|
|
|
|
gst_avi_demux_massage_index (GstAviDemux * avi,
|
|
|
|
|
GList * list, GList * alloc_list)
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
{
|
|
|
|
|
gst_avi_index_entry *entry;
|
|
|
|
|
avi_stream_context *stream;
|
|
|
|
|
gint i;
|
2004-10-04 13:32:20 +00:00
|
|
|
|
GList *one;
|
|
|
|
|
|
|
|
|
|
GST_LOG ("Starting index massage");
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
|
|
|
|
|
/* init frames */
|
|
|
|
|
for (i = 0; i < avi->num_streams; i++) {
|
2004-11-29 12:41:03 +00:00
|
|
|
|
GstFormat fmt = GST_FORMAT_TIME;
|
2005-07-01 08:10:56 +00:00
|
|
|
|
gint64 delay = 0;
|
2004-11-29 12:41:03 +00:00
|
|
|
|
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
stream = &avi->stream[i];
|
2004-10-02 20:27:04 +00:00
|
|
|
|
if (stream->strh->type == GST_RIFF_FCC_vids) {
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (!gst_avi_demux_src_convert (stream->pad,
|
|
|
|
|
GST_FORMAT_DEFAULT, stream->strh->init_frames, &fmt, &delay)) {
|
|
|
|
|
delay = 0;
|
2004-11-29 12:41:03 +00:00
|
|
|
|
}
|
2004-10-02 20:27:04 +00:00
|
|
|
|
} else {
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (!gst_avi_demux_src_convert (stream->pad,
|
|
|
|
|
GST_FORMAT_DEFAULT, stream->strh->init_frames, &fmt, &delay)) {
|
|
|
|
|
delay = 0;
|
2004-11-29 12:41:03 +00:00
|
|
|
|
}
|
2004-10-02 20:27:04 +00:00
|
|
|
|
}
|
2004-11-29 12:41:03 +00:00
|
|
|
|
GST_LOG ("Adding init_time=%" GST_TIME_FORMAT " to stream %d",
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_TIME_ARGS (delay), i);
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
for (one = list; one != NULL; one = one->next) {
|
|
|
|
|
entry = one->data;
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (entry->stream_nr == i)
|
|
|
|
|
entry->ts += delay;
|
|
|
|
|
}
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
}
|
|
|
|
|
|
2004-10-04 13:32:20 +00:00
|
|
|
|
GST_LOG ("I'm now going to cut large chunks into smaller pieces");
|
|
|
|
|
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
/* cut chunks in small (seekable) pieces */
|
2004-10-09 19:33:21 +00:00
|
|
|
|
for (i = 0; i < avi->num_streams; i++) {
|
|
|
|
|
if (avi->stream[i].total_frames != 1)
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
continue;
|
|
|
|
|
|
2004-10-09 19:33:21 +00:00
|
|
|
|
for (one = list; one != NULL; one = one->next) {
|
|
|
|
|
entry = one->data;
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
|
2004-10-09 19:33:21 +00:00
|
|
|
|
if (entry->stream_nr != i)
|
|
|
|
|
continue;
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
|
2004-10-09 19:33:21 +00:00
|
|
|
|
#define MAX_DURATION (GST_SECOND / 2)
|
2004-10-04 13:32:20 +00:00
|
|
|
|
|
2004-10-09 19:33:21 +00:00
|
|
|
|
/* check for max duration of a single buffer. I suppose that
|
|
|
|
|
* the allocation of index entries could be improved. */
|
|
|
|
|
stream = &avi->stream[entry->stream_nr];
|
|
|
|
|
if (entry->dur > MAX_DURATION && stream->strh->type == GST_RIFF_FCC_auds) {
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
guint32 ideal_size = stream->strf.auds->av_bps / 10;
|
2004-10-09 19:33:21 +00:00
|
|
|
|
gst_avi_index_entry *entries;
|
|
|
|
|
gint old_size, num_added;
|
|
|
|
|
GList *one2;
|
|
|
|
|
|
|
|
|
|
/* copy index */
|
|
|
|
|
old_size = entry->size;
|
|
|
|
|
num_added = (entry->size - 1) / ideal_size;
|
|
|
|
|
avi->index_size += num_added;
|
|
|
|
|
entries = g_malloc (sizeof (gst_avi_index_entry) * num_added);
|
|
|
|
|
alloc_list = g_list_prepend (alloc_list, entries);
|
|
|
|
|
for (one2 = one->next; one2 != NULL; one2 = one2->next) {
|
|
|
|
|
gst_avi_index_entry *entry2 = one2->data;
|
|
|
|
|
|
|
|
|
|
entry2->index_nr += num_added;
|
|
|
|
|
if (entry2->stream_nr == entry->stream_nr)
|
|
|
|
|
entry2->frames_before += num_added;
|
2004-10-05 15:28:08 +00:00
|
|
|
|
}
|
2004-10-04 13:32:20 +00:00
|
|
|
|
|
2004-10-09 19:33:21 +00:00
|
|
|
|
/* new sized index chunks */
|
|
|
|
|
for (i = 0; i < num_added + 1; i++) {
|
|
|
|
|
gst_avi_index_entry *entry2;
|
|
|
|
|
|
|
|
|
|
if (i == 0) {
|
|
|
|
|
entry2 = entry;
|
|
|
|
|
} else {
|
|
|
|
|
entry2 = &entries[i - 1];
|
|
|
|
|
list = g_list_insert_before (list, one->next, entry2);
|
|
|
|
|
entry = one->data;
|
|
|
|
|
one = one->next;
|
|
|
|
|
memcpy (entry2, entry, sizeof (gst_avi_index_entry));
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (old_size >= ideal_size) {
|
|
|
|
|
entry2->size = ideal_size;
|
|
|
|
|
old_size -= ideal_size;
|
|
|
|
|
} else {
|
|
|
|
|
entry2->size = old_size;
|
|
|
|
|
}
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
entry2->dur = GST_SECOND * entry2->size / stream->strf.auds->av_bps;
|
2004-10-09 19:33:21 +00:00
|
|
|
|
if (i != 0) {
|
|
|
|
|
entry2->index_nr++;
|
|
|
|
|
entry2->ts += entry->dur;
|
|
|
|
|
entry2->offset += entry->size;
|
|
|
|
|
entry2->bytes_before += entry->size;
|
|
|
|
|
entry2->frames_before++;
|
|
|
|
|
}
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2004-10-04 13:32:20 +00:00
|
|
|
|
GST_LOG ("I'm now going to reorder the index entries for time");
|
|
|
|
|
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
/* re-order for time */
|
2004-10-04 13:32:20 +00:00
|
|
|
|
list = g_list_sort (list, (GCompareFunc) sort);
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
|
2004-10-04 13:32:20 +00:00
|
|
|
|
GST_LOG ("Filling in index array");
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
|
2004-10-04 13:32:20 +00:00
|
|
|
|
avi->index_size = g_list_length (list);
|
|
|
|
|
avi->index_entries = g_new (gst_avi_index_entry, avi->index_size);
|
|
|
|
|
for (i = 0, one = list; one != NULL; one = one->next, i++) {
|
|
|
|
|
entry = one->data;
|
|
|
|
|
memcpy (&avi->index_entries[i], entry, sizeof (gst_avi_index_entry));
|
2004-10-04 17:29:27 +00:00
|
|
|
|
avi->index_entries[i].index_nr = i;
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
}
|
2004-10-04 13:32:20 +00:00
|
|
|
|
|
|
|
|
|
GST_LOG ("Freeing original index list");
|
|
|
|
|
|
|
|
|
|
g_list_foreach (alloc_list, (GFunc) g_free, NULL);
|
|
|
|
|
g_list_free (alloc_list);
|
|
|
|
|
g_list_free (list);
|
|
|
|
|
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
for (i = 0; i < avi->num_streams; i++) {
|
|
|
|
|
GST_LOG ("Stream %d, %d frames, %" G_GUINT64_FORMAT " bytes", i,
|
|
|
|
|
avi->stream[i].total_frames, avi->stream[i].total_bytes);
|
|
|
|
|
}
|
|
|
|
|
|
2004-10-04 13:32:20 +00:00
|
|
|
|
GST_LOG ("Index massaging done");
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
}
|
|
|
|
|
|
2005-07-16 16:43:53 +00:00
|
|
|
|
static gboolean
|
|
|
|
|
gst_avi_demux_send_event (GstAviDemux * avi, GstEvent * event)
|
|
|
|
|
{
|
|
|
|
|
gint i;
|
|
|
|
|
|
|
|
|
|
for (i = 0; i < avi->num_streams; i++) {
|
|
|
|
|
avi_stream_context *stream = &avi->stream[i];
|
|
|
|
|
|
|
|
|
|
gst_event_ref (event);
|
|
|
|
|
gst_pad_push_event (stream->pad, event);
|
|
|
|
|
}
|
|
|
|
|
gst_event_unref (event);
|
|
|
|
|
return TRUE;
|
|
|
|
|
}
|
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
/*
|
|
|
|
|
* Read full AVI headers.
|
|
|
|
|
*/
|
2002-06-18 19:17:04 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
static GstFlowReturn
|
2004-03-14 22:34:33 +00:00
|
|
|
|
gst_avi_demux_stream_header (GstAviDemux * avi)
|
2001-12-22 23:22:30 +00:00
|
|
|
|
{
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GstFlowReturn res;
|
|
|
|
|
GstBuffer *buf, *sub = NULL;
|
|
|
|
|
guint32 tag;
|
2004-10-04 13:32:20 +00:00
|
|
|
|
GList *index = NULL, *alloc = NULL;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
guint offset = 4;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
|
|
|
|
/* the header consists of a 'hdrl' LIST tag */
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if ((res = gst_riff_read_chunk (GST_ELEMENT (avi), avi->sinkpad,
|
|
|
|
|
&avi->offset, &tag, &buf)) != GST_FLOW_OK)
|
|
|
|
|
return res;
|
|
|
|
|
else if (tag != GST_RIFF_TAG_LIST) {
|
2004-02-02 17:23:33 +00:00
|
|
|
|
GST_ELEMENT_ERROR (avi, STREAM, DEMUX, (NULL),
|
2005-11-21 13:38:24 +00:00
|
|
|
|
("Invalid AVI header (no LIST at start): %"
|
2004-03-15 19:32:27 +00:00
|
|
|
|
GST_FOURCC_FORMAT, GST_FOURCC_ARGS (tag)));
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
return GST_FLOW_ERROR;
|
2006-02-14 16:58:30 +00:00
|
|
|
|
} else if (GST_BUFFER_SIZE (buf) < 4) {
|
|
|
|
|
GST_ELEMENT_ERROR (avi, STREAM, DEMUX, (NULL), ("No header found"));
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
return GST_FLOW_ERROR;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
|
|
|
|
|
2006-02-14 16:58:30 +00:00
|
|
|
|
/* Find the 'hdrl' LIST tag */
|
|
|
|
|
while (GST_READ_UINT32_LE (GST_BUFFER_DATA (buf)) != GST_RIFF_LIST_hdrl) {
|
|
|
|
|
|
|
|
|
|
GST_LOG_OBJECT (avi, "buffer contains %" GST_FOURCC_FORMAT,
|
|
|
|
|
GST_FOURCC_ARGS (GST_READ_UINT32_LE (GST_BUFFER_DATA (buf))));
|
|
|
|
|
|
|
|
|
|
/* Eat up */
|
|
|
|
|
gst_buffer_unref (buf);
|
|
|
|
|
if ((res = gst_riff_read_chunk (GST_ELEMENT (avi), avi->sinkpad,
|
|
|
|
|
&avi->offset, &tag, &buf)) != GST_FLOW_OK)
|
|
|
|
|
return res;
|
|
|
|
|
else if (tag != GST_RIFF_TAG_LIST) {
|
|
|
|
|
GST_ELEMENT_ERROR (avi, STREAM, DEMUX, (NULL),
|
|
|
|
|
("Invalid AVI header (no LIST at start): %"
|
|
|
|
|
GST_FOURCC_FORMAT, GST_FOURCC_ARGS (tag)));
|
|
|
|
|
return GST_FLOW_ERROR;
|
|
|
|
|
} else if (GST_BUFFER_SIZE (buf) < 4) {
|
|
|
|
|
|
|
|
|
|
GST_ELEMENT_ERROR (avi, STREAM, DEMUX, (NULL),
|
|
|
|
|
("Invalid AVI header (no hdrl at start): %"
|
|
|
|
|
GST_FOURCC_FORMAT, GST_FOURCC_ARGS (tag)));
|
|
|
|
|
gst_buffer_unref (buf);
|
|
|
|
|
return GST_FLOW_ERROR;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
/* the hdrl starts with a 'avih' header */
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (!gst_riff_parse_chunk (GST_ELEMENT (avi), buf, &offset, &tag, &sub) ||
|
|
|
|
|
tag != GST_RIFF_TAG_avih) {
|
2004-02-02 17:23:33 +00:00
|
|
|
|
GST_ELEMENT_ERROR (avi, STREAM, DEMUX, (NULL),
|
2005-11-21 13:38:24 +00:00
|
|
|
|
("Invalid AVI header (no avih at start): %"
|
2004-03-15 19:32:27 +00:00
|
|
|
|
GST_FOURCC_FORMAT, GST_FOURCC_ARGS (tag)));
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
if (sub) {
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_buffer_unref (sub);
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
sub = NULL;
|
|
|
|
|
}
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_buffer_unref (buf);
|
|
|
|
|
return GST_FLOW_ERROR;
|
|
|
|
|
} else if (!gst_avi_demux_parse_avih (GST_ELEMENT (avi), sub, &avi->avih)) {
|
|
|
|
|
gst_buffer_unref (buf);
|
|
|
|
|
return GST_FLOW_ERROR;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* now, read the elements from the header until the end */
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
while (gst_riff_parse_chunk (GST_ELEMENT (avi), buf, &offset, &tag, &sub)) {
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
switch (tag) {
|
|
|
|
|
case GST_RIFF_TAG_LIST:
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (!sub || GST_BUFFER_SIZE (sub) < 4) {
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
if (sub) {
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_buffer_unref (sub);
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
sub = NULL;
|
|
|
|
|
}
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
break;
|
|
|
|
|
}
|
2004-03-15 19:32:27 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
switch (GST_READ_UINT32_LE (GST_BUFFER_DATA (sub))) {
|
2004-03-15 19:32:27 +00:00
|
|
|
|
case GST_RIFF_LIST_strl:
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
if (!(gst_avi_demux_parse_stream (GST_ELEMENT (avi), sub)))
|
|
|
|
|
return GST_FLOW_ERROR;
|
2005-11-11 16:48:58 +00:00
|
|
|
|
gst_buffer_unref (sub);
|
2004-03-15 19:32:27 +00:00
|
|
|
|
break;
|
|
|
|
|
case GST_RIFF_LIST_odml:
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_parse_odml (GST_ELEMENT (avi), sub);
|
2004-03-15 19:32:27 +00:00
|
|
|
|
break;
|
|
|
|
|
default:
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_WARNING_OBJECT (avi,
|
2005-11-21 13:38:24 +00:00
|
|
|
|
"Unknown list %" GST_FOURCC_FORMAT " in AVI header",
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_FOURCC_ARGS (GST_READ_UINT32_LE (GST_BUFFER_DATA (sub))));
|
2004-03-15 19:32:27 +00:00
|
|
|
|
/* fall-through */
|
|
|
|
|
case GST_RIFF_TAG_JUNK:
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
if (sub) {
|
2005-12-07 11:40:46 +00:00
|
|
|
|
gst_buffer_unref (sub);
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
sub = NULL;
|
|
|
|
|
}
|
2004-03-15 19:32:27 +00:00
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
break;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
default:
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_WARNING_OBJECT (avi,
|
2005-11-21 13:38:24 +00:00
|
|
|
|
"Unknown off %d tag %" GST_FOURCC_FORMAT " in AVI header",
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
offset, GST_FOURCC_ARGS (tag));
|
2004-03-15 19:32:27 +00:00
|
|
|
|
/* fall-through */
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
case GST_RIFF_TAG_JUNK:
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
if (sub) {
|
2005-12-07 11:40:46 +00:00
|
|
|
|
gst_buffer_unref (sub);
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
sub = NULL;
|
|
|
|
|
}
|
2004-03-15 19:32:27 +00:00
|
|
|
|
break;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
2003-01-20 20:49:34 +00:00
|
|
|
|
}
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_buffer_unref (buf);
|
2002-06-18 19:17:04 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
/* check parsed streams */
|
|
|
|
|
if (avi->num_streams == 0) {
|
|
|
|
|
GST_ELEMENT_ERROR (avi, STREAM, DEMUX, (NULL), ("No streams found"));
|
|
|
|
|
return GST_FLOW_ERROR;
|
|
|
|
|
} else if (avi->num_streams != avi->avih->streams) {
|
|
|
|
|
GST_WARNING_OBJECT (avi,
|
|
|
|
|
"Stream header mentioned %d streams, but %d available",
|
|
|
|
|
avi->avih->streams, avi->num_streams);
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
|
|
|
|
|
2005-08-09 14:40:02 +00:00
|
|
|
|
GST_DEBUG_OBJECT (avi, "skipping junk between header and data ...");
|
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
/* Now, find the data (i.e. skip all junk between header and data) */
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
do {
|
|
|
|
|
guint size;
|
|
|
|
|
guint32 tag, ltag;
|
|
|
|
|
|
|
|
|
|
if ((res = gst_pad_pull_range (avi->sinkpad, avi->offset,
|
|
|
|
|
12, &buf)) != GST_FLOW_OK)
|
|
|
|
|
return res;
|
|
|
|
|
else if (!buf || GST_BUFFER_SIZE (buf) < 12)
|
|
|
|
|
return GST_FLOW_ERROR;
|
|
|
|
|
|
|
|
|
|
tag = GST_READ_UINT32_LE (GST_BUFFER_DATA (buf));
|
|
|
|
|
size = GST_READ_UINT32_LE (GST_BUFFER_DATA (buf) + 4);
|
|
|
|
|
ltag = GST_READ_UINT32_LE (GST_BUFFER_DATA (buf) + 8);
|
|
|
|
|
gst_buffer_unref (buf);
|
|
|
|
|
|
|
|
|
|
if (tag == GST_RIFF_TAG_LIST) {
|
|
|
|
|
switch (ltag) {
|
|
|
|
|
case GST_RIFF_LIST_movi:
|
|
|
|
|
goto done;
|
|
|
|
|
case GST_RIFF_LIST_INFO:
|
|
|
|
|
if ((res = gst_riff_read_chunk (GST_ELEMENT (avi), avi->sinkpad,
|
|
|
|
|
&avi->offset, &tag, &buf)) != GST_FLOW_OK)
|
|
|
|
|
return res;
|
|
|
|
|
else {
|
|
|
|
|
GstTagList *t;
|
|
|
|
|
|
|
|
|
|
sub = gst_buffer_create_sub (buf, 4, GST_BUFFER_SIZE (buf) - 4);
|
|
|
|
|
gst_riff_parse_info (GST_ELEMENT (avi), sub, &t);
|
2005-08-16 16:30:23 +00:00
|
|
|
|
if (t) {
|
|
|
|
|
gst_element_found_tags (GST_ELEMENT (avi), t);
|
|
|
|
|
}
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
if (sub) {
|
|
|
|
|
gst_buffer_unref (sub);
|
|
|
|
|
sub = NULL;
|
|
|
|
|
}
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_buffer_unref (buf);
|
|
|
|
|
}
|
2005-08-09 14:40:02 +00:00
|
|
|
|
/* gst_riff_read_chunk() has already advanced avi->offset */
|
|
|
|
|
break;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
default:
|
|
|
|
|
avi->offset += 8 + ((size + 1) & ~1);
|
|
|
|
|
break;
|
2004-01-23 13:22:16 +00:00
|
|
|
|
}
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
} else {
|
|
|
|
|
avi->offset += 8 + ((size + 1) & ~1);
|
2003-01-20 20:49:34 +00:00
|
|
|
|
}
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
} while (1);
|
|
|
|
|
done:
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
|
|
|
|
/* create or read stream index (for seeking) */
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
if (avi->stream[0].indexes != NULL) {
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_read_subindexes (avi, &index, &alloc);
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
}
|
2004-10-04 13:32:20 +00:00
|
|
|
|
if (!index) {
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (avi->avih->flags & GST_RIFF_AVIH_HASINDEX) {
|
|
|
|
|
gst_avi_demux_stream_index (avi, &index, &alloc);
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
}
|
|
|
|
|
/* some indexes are incomplete, continue streaming from there */
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
if (!index)
|
|
|
|
|
gst_avi_demux_stream_scan (avi, &index, &alloc);
|
2004-10-04 13:32:20 +00:00
|
|
|
|
}
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
if (index) {
|
2004-10-04 13:32:20 +00:00
|
|
|
|
gst_avi_demux_massage_index (avi, index, alloc);
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
} else {
|
2004-10-04 13:32:20 +00:00
|
|
|
|
g_list_free (index);
|
|
|
|
|
g_list_foreach (alloc, (GFunc) g_free, NULL);
|
|
|
|
|
g_list_free (alloc);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
|
|
|
|
|
GST_ELEMENT_ERROR (avi, STREAM, NOT_IMPLEMENTED, (NULL),
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
("Could not get/create index"));
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
return GST_FLOW_ERROR;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
|
|
|
|
|
2005-07-16 16:43:53 +00:00
|
|
|
|
/* send initial discont */
|
2005-11-11 16:48:58 +00:00
|
|
|
|
avi->segment_start = 0;
|
|
|
|
|
avi->segment_stop = (gint64) (((gfloat) avi->stream[0].strh->scale) *
|
|
|
|
|
avi->stream[0].strh->length / avi->stream[0].strh->rate) * GST_SECOND;
|
2005-11-22 11:53:34 +00:00
|
|
|
|
avi->seek_event = gst_event_new_new_segment
|
2005-11-11 16:48:58 +00:00
|
|
|
|
(FALSE, avi->segment_rate, GST_FORMAT_TIME,
|
|
|
|
|
avi->segment_start, avi->segment_stop, avi->segment_start);
|
2005-07-16 16:43:53 +00:00
|
|
|
|
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
/* at this point we know all the streams and we can signal the no more
|
|
|
|
|
* pads signal */
|
2005-01-13 17:11:08 +00:00
|
|
|
|
GST_DEBUG_OBJECT (avi, "signaling no more pads");
|
ext/gnomevfs/gstgnomevfssrc.c: Some debug.
Original commit message from CVS:
* ext/gnomevfs/gstgnomevfssrc.c: (gst_gnomevfssrc_get),
(gst_gnomevfssrc_srcpad_query), (gst_gnomevfssrc_srcpad_event):
Some debug.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_read_superindex),
(gst_avi_demux_read_subindexes), (gst_avi_demux_add_stream),
(gst_avi_demux_stream_index), (gst_avi_demux_skip),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header):
* gst/avi/gstavidemux.h:
Support for openDML-2.0 indx/ix## chunks. Support for broken index
recovery (where, if part of the index is broken, we will still read
the rest of the index and recover the broken part by stream
scanning). More broken media support. EOS workarounds. General AVI
braindamage headache recovery. Aspirin included.
2004-10-12 12:29:45 +00:00
|
|
|
|
gst_element_no_more_pads (GST_ELEMENT (avi));
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
return GST_FLOW_OK;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* Handle seek.
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
static gboolean
|
2005-11-11 16:48:58 +00:00
|
|
|
|
gst_avi_demux_handle_seek (GstAviDemux * avi, gboolean update)
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
{
|
2005-11-11 16:48:58 +00:00
|
|
|
|
GstClockTime start_time;
|
|
|
|
|
gboolean flush, keyframe;
|
|
|
|
|
guint stream;
|
|
|
|
|
gst_avi_index_entry *entry;
|
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
/* FIXME: if we seek in an openDML file, we will have multiple
|
|
|
|
|
* primary levels. Seeking in between those will cause havoc. */
|
2002-02-03 20:10:04 +00:00
|
|
|
|
|
2005-11-11 16:48:58 +00:00
|
|
|
|
flush = avi->segment_flags & GST_SEEK_FLAG_FLUSH;
|
|
|
|
|
keyframe = avi->segment_flags & GST_SEEK_FLAG_KEY_UNIT;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
2005-11-11 16:48:58 +00:00
|
|
|
|
if (flush) {
|
|
|
|
|
for (stream = avi->num_streams; stream--;)
|
|
|
|
|
gst_pad_push_event (avi->stream[stream].pad,
|
|
|
|
|
gst_event_new_flush_start ());
|
|
|
|
|
gst_pad_push_event (avi->sinkpad, gst_event_new_flush_start ());
|
|
|
|
|
} else
|
|
|
|
|
gst_pad_pause_task (avi->sinkpad);
|
2003-01-20 20:49:34 +00:00
|
|
|
|
|
2005-11-21 18:09:02 +00:00
|
|
|
|
GST_PAD_STREAM_LOCK (avi->sinkpad);
|
2005-01-29 15:47:18 +00:00
|
|
|
|
|
2005-11-11 16:48:58 +00:00
|
|
|
|
/* fill current_entry according to flags and update */
|
|
|
|
|
if (update) {
|
|
|
|
|
entry = gst_avi_demux_index_entry_for_time (avi, 0, avi->segment_start,
|
|
|
|
|
(guint32) GST_RIFF_IF_KEYFRAME);
|
|
|
|
|
if (entry) {
|
|
|
|
|
GST_DEBUG_OBJECT (avi,
|
|
|
|
|
"Got keyframe entry %d [stream:%d / ts:%" GST_TIME_FORMAT
|
|
|
|
|
" / duration:%" GST_TIME_FORMAT "]", entry->index_nr,
|
|
|
|
|
entry->stream_nr, GST_TIME_ARGS (entry->ts),
|
|
|
|
|
GST_TIME_ARGS (entry->dur));
|
|
|
|
|
avi->current_entry = entry->index_nr;
|
|
|
|
|
} else {
|
|
|
|
|
GST_WARNING_OBJECT (avi,
|
|
|
|
|
"Couldn't find AviIndexEntry for time:%" GST_TIME_FORMAT,
|
|
|
|
|
GST_TIME_ARGS (avi->segment_start));
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
GST_DEBUG ("seek: %" GST_TIME_FORMAT " -- %" GST_TIME_FORMAT
|
|
|
|
|
" keyframe seeking:%d update:%d", GST_TIME_ARGS (avi->segment_start),
|
|
|
|
|
GST_TIME_ARGS (avi->segment_stop), keyframe, update);
|
|
|
|
|
|
|
|
|
|
if (keyframe)
|
|
|
|
|
start_time = avi->index_entries[avi->current_entry].ts;
|
|
|
|
|
else
|
|
|
|
|
start_time = avi->segment_start;
|
|
|
|
|
|
2005-11-22 11:53:34 +00:00
|
|
|
|
avi->seek_event = gst_event_new_new_segment
|
2005-11-11 16:48:58 +00:00
|
|
|
|
(!update, avi->segment_rate, GST_FORMAT_TIME,
|
|
|
|
|
start_time, avi->segment_stop, start_time);
|
|
|
|
|
|
|
|
|
|
if (flush) {
|
|
|
|
|
for (stream = avi->num_streams; stream--;)
|
|
|
|
|
gst_pad_push_event (avi->stream[stream].pad, gst_event_new_flush_stop ());
|
|
|
|
|
gst_pad_push_event (avi->sinkpad, gst_event_new_flush_stop ());
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (avi->segment_flags & GST_SEEK_FLAG_SEGMENT) {
|
|
|
|
|
gst_element_post_message (GST_ELEMENT (avi),
|
|
|
|
|
gst_message_new_segment_start (GST_OBJECT (avi), GST_FORMAT_TIME,
|
|
|
|
|
start_time));
|
|
|
|
|
}
|
2005-02-17 19:21:42 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_pad_start_task (avi->sinkpad, (GstTaskFunction) gst_avi_demux_loop,
|
|
|
|
|
avi->sinkpad);
|
2005-01-29 15:47:18 +00:00
|
|
|
|
|
2005-11-21 18:09:02 +00:00
|
|
|
|
GST_PAD_STREAM_UNLOCK (avi->sinkpad);
|
2005-01-29 15:47:18 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
return TRUE;
|
2005-11-11 16:48:58 +00:00
|
|
|
|
|
2005-01-29 15:47:18 +00:00
|
|
|
|
}
|
|
|
|
|
|
2005-11-15 12:16:00 +00:00
|
|
|
|
/*
|
|
|
|
|
* Invert DIB buffers... Takes existing buffer and
|
|
|
|
|
* returns either the buffer or a new one (with old
|
|
|
|
|
* one dereferenced).
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
static inline void
|
|
|
|
|
swap_line (guint8 * d1, guint8 * d2, guint8 * tmp, gint bytes)
|
|
|
|
|
{
|
|
|
|
|
memcpy (tmp, d1, bytes);
|
|
|
|
|
memcpy (d1, d2, bytes);
|
|
|
|
|
memcpy (d2, tmp, bytes);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static GstBuffer *
|
|
|
|
|
gst_avi_demux_invert (avi_stream_context * stream, GstBuffer * buf)
|
|
|
|
|
{
|
|
|
|
|
gint y, h = stream->strf.vids->height, w = stream->strf.vids->width;
|
2005-12-07 11:40:46 +00:00
|
|
|
|
guint8 *tmp = NULL;
|
|
|
|
|
|
2006-01-23 09:59:03 +00:00
|
|
|
|
buf = gst_buffer_make_writable (buf);
|
2005-12-07 11:40:46 +00:00
|
|
|
|
if (GST_BUFFER_SIZE (buf) < (w * h)) {
|
|
|
|
|
GST_WARNING ("Buffer is smaller than reported Width x Height");
|
|
|
|
|
return buf;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
tmp = g_malloc (w);
|
2005-11-15 12:16:00 +00:00
|
|
|
|
|
|
|
|
|
for (y = 0; y < h / 2; y++) {
|
|
|
|
|
swap_line (GST_BUFFER_DATA (buf) + w * y,
|
|
|
|
|
GST_BUFFER_DATA (buf) + w * (h - 1 - y), tmp, w);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
g_free (tmp);
|
|
|
|
|
|
|
|
|
|
return buf;
|
|
|
|
|
}
|
|
|
|
|
|
2006-02-02 12:29:24 +00:00
|
|
|
|
static gboolean
|
|
|
|
|
gst_avi_demux_all_source_pads_unlinked (GstAviDemux * avi)
|
|
|
|
|
{
|
2006-02-04 15:41:43 +00:00
|
|
|
|
gint i;
|
2006-02-02 12:29:24 +00:00
|
|
|
|
|
|
|
|
|
for (i = 0; i < avi->num_streams; ++i) {
|
2006-02-04 15:41:43 +00:00
|
|
|
|
if (gst_pad_is_linked (avi->stream[i].pad))
|
|
|
|
|
return FALSE;
|
|
|
|
|
/* ignore unlinked state if we haven't tried to push on this pad yet */
|
|
|
|
|
if (avi->stream[i].last_flow == GST_FLOW_OK)
|
|
|
|
|
return FALSE;
|
2006-02-02 12:29:24 +00:00
|
|
|
|
}
|
|
|
|
|
|
2006-02-04 15:41:43 +00:00
|
|
|
|
return TRUE;
|
2006-02-02 12:29:24 +00:00
|
|
|
|
}
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
static GstFlowReturn
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
gst_avi_demux_process_next_entry (GstAviDemux * avi)
|
|
|
|
|
{
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GstFlowReturn res;
|
2004-12-23 10:53:27 +00:00
|
|
|
|
gboolean processed = FALSE;
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
|
2004-12-23 10:53:27 +00:00
|
|
|
|
do {
|
2005-11-11 16:48:58 +00:00
|
|
|
|
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
if (avi->current_entry >= avi->index_size) {
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GST_LOG_OBJECT (avi, "Handled last index entry, setting EOS (%d > %d)",
|
|
|
|
|
avi->current_entry, avi->index_size);
|
2006-02-13 10:43:15 +00:00
|
|
|
|
goto eos;
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
} else {
|
|
|
|
|
GstBuffer *buf;
|
|
|
|
|
gst_avi_index_entry *entry = &avi->index_entries[avi->current_entry++];
|
|
|
|
|
avi_stream_context *stream;
|
|
|
|
|
|
|
|
|
|
if (entry->stream_nr >= avi->num_streams) {
|
2005-01-13 17:11:08 +00:00
|
|
|
|
GST_DEBUG_OBJECT (avi,
|
|
|
|
|
"Entry has non-existing stream nr %d", entry->stream_nr);
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
continue;
|
|
|
|
|
}
|
2005-11-11 16:48:58 +00:00
|
|
|
|
|
2006-02-13 10:43:15 +00:00
|
|
|
|
if ((entry->flags & GST_RIFF_IF_KEYFRAME)
|
|
|
|
|
&& GST_CLOCK_TIME_IS_VALID (entry->ts)
|
|
|
|
|
&& GST_CLOCK_TIME_IS_VALID (avi->segment_stop)
|
|
|
|
|
&& (entry->ts > avi->segment_stop)) {
|
|
|
|
|
GST_LOG_OBJECT (avi, "Found keyframe after segment,"
|
|
|
|
|
" setting EOS (%" GST_TIME_FORMAT " > %" GST_TIME_FORMAT ")",
|
|
|
|
|
GST_TIME_ARGS (entry->ts), GST_TIME_ARGS (avi->segment_stop));
|
|
|
|
|
goto eos;
|
|
|
|
|
}
|
|
|
|
|
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
stream = &avi->stream[entry->stream_nr];
|
2005-11-11 16:48:58 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (entry->size == 0 || !stream->pad) {
|
|
|
|
|
GST_DEBUG_OBJECT (avi, "Skipping entry %d (%d, %p)",
|
|
|
|
|
avi->current_entry - 1, entry->size, stream->pad);
|
|
|
|
|
goto next;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if ((res = gst_pad_pull_range (avi->sinkpad, entry->offset +
|
|
|
|
|
avi->index_offset, entry->size, &buf)) != GST_FLOW_OK)
|
|
|
|
|
return res;
|
|
|
|
|
else {
|
2005-11-15 12:16:00 +00:00
|
|
|
|
if (stream->strh->fcc_handler == GST_MAKE_FOURCC ('D', 'I', 'B', ' ')) {
|
|
|
|
|
buf = gst_avi_demux_invert (stream, buf);
|
|
|
|
|
}
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (!(entry->flags & GST_RIFF_IF_KEYFRAME))
|
|
|
|
|
GST_BUFFER_FLAG_SET (buf, GST_BUFFER_FLAG_DELTA_UNIT);
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
GST_BUFFER_TIMESTAMP (buf) = entry->ts;
|
|
|
|
|
GST_BUFFER_DURATION (buf) = entry->dur;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_buffer_set_caps (buf, GST_PAD_CAPS (stream->pad));
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
GST_DEBUG_OBJECT (avi, "Processing buffer of size %d and time %"
|
|
|
|
|
GST_TIME_FORMAT " on pad %s",
|
|
|
|
|
GST_BUFFER_SIZE (buf), GST_TIME_ARGS (GST_BUFFER_TIMESTAMP (buf)),
|
2005-11-07 10:31:32 +00:00
|
|
|
|
GST_PAD_NAME (stream->pad));
|
2006-02-02 12:29:24 +00:00
|
|
|
|
res = gst_pad_push (stream->pad, buf);
|
2006-02-04 15:41:43 +00:00
|
|
|
|
stream->last_flow = res;
|
2006-02-02 12:29:24 +00:00
|
|
|
|
if (res != GST_FLOW_OK && res != GST_FLOW_NOT_LINKED) {
|
|
|
|
|
GST_DEBUG_OBJECT (avi, "Flow on pad %s: %s",
|
|
|
|
|
GST_PAD_NAME (stream->pad), gst_flow_get_name (res));
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
return res;
|
2006-02-02 12:29:24 +00:00
|
|
|
|
}
|
2004-12-23 10:53:27 +00:00
|
|
|
|
processed = TRUE;
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
}
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
next:
|
2004-12-06 10:28:13 +00:00
|
|
|
|
stream->current_frame = entry->frames_before + 1;
|
|
|
|
|
stream->current_byte = entry->bytes_before + entry->size;
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
}
|
2004-12-23 10:53:27 +00:00
|
|
|
|
} while (!processed);
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
return GST_FLOW_OK;
|
2006-02-13 10:43:15 +00:00
|
|
|
|
|
|
|
|
|
eos:
|
|
|
|
|
/* handle end-of-stream/segment */
|
|
|
|
|
if (avi->segment_flags & GST_SEEK_FLAG_SEGMENT)
|
|
|
|
|
gst_element_post_message
|
|
|
|
|
(GST_ELEMENT (avi),
|
|
|
|
|
gst_message_new_segment_done (GST_OBJECT (avi), GST_FORMAT_TIME,
|
|
|
|
|
avi->segment_stop));
|
|
|
|
|
else
|
|
|
|
|
gst_avi_demux_send_event (avi, gst_event_new_eos ());
|
|
|
|
|
return GST_FLOW_WRONG_STATE;
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
}
|
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
/*
|
|
|
|
|
* Read data.
|
|
|
|
|
*/
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
static GstFlowReturn
|
2004-03-14 22:34:33 +00:00
|
|
|
|
gst_avi_demux_stream_data (GstAviDemux * avi)
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
{
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
/* if we have a avi->index_entries[], we don't want to read
|
|
|
|
|
* the stream linearly, but seek to the next ts/index_entry. */
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
//if (avi->index_entries != NULL) {
|
|
|
|
|
return gst_avi_demux_process_next_entry (avi);
|
|
|
|
|
//}
|
|
|
|
|
#if 0
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
if (!gst_avi_demux_sync (avi, &tag, FALSE))
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
return FALSE;
|
|
|
|
|
stream_nr = CHUNKID_TO_STREAMNR (tag);
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
if (stream_nr < 0 || stream_nr >= avi->num_streams) {
|
|
|
|
|
/* recoverable */
|
2005-11-21 13:38:24 +00:00
|
|
|
|
GST_WARNING ("Invalid stream ID %d (%" GST_FOURCC_FORMAT ")",
|
2004-03-15 19:32:27 +00:00
|
|
|
|
stream_nr, GST_FOURCC_ARGS (tag));
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
if (!gst_riff_read_skip (riff))
|
|
|
|
|
return FALSE;
|
|
|
|
|
} else {
|
|
|
|
|
avi_stream_context *stream;
|
|
|
|
|
GstClockTime next_ts;
|
|
|
|
|
GstFormat format;
|
|
|
|
|
GstBuffer *buf;
|
2003-01-20 20:49:34 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
/* get buffer */
|
|
|
|
|
if (!gst_riff_read_data (riff, &tag, &buf))
|
|
|
|
|
return FALSE;
|
|
|
|
|
|
|
|
|
|
/* get time of this buffer */
|
|
|
|
|
stream = &avi->stream[stream_nr];
|
|
|
|
|
format = GST_FORMAT_TIME;
|
2004-03-14 22:34:33 +00:00
|
|
|
|
gst_pad_query (stream->pad, GST_QUERY_POSITION, &format, &next_ts);
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
|
|
|
|
/* set delay (if any) */
|
|
|
|
|
if (stream->strh->init_frames == stream->current_frame &&
|
2004-03-15 19:32:27 +00:00
|
|
|
|
stream->delay == 0)
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
stream->delay = next_ts;
|
|
|
|
|
|
|
|
|
|
stream->current_frame++;
|
|
|
|
|
stream->current_byte += GST_BUFFER_SIZE (buf);
|
|
|
|
|
|
|
|
|
|
/* should we skip this data? */
|
|
|
|
|
if (stream->skip) {
|
|
|
|
|
stream->skip--;
|
|
|
|
|
gst_buffer_unref (buf);
|
|
|
|
|
} else {
|
|
|
|
|
if (!stream->pad || !GST_PAD_IS_USABLE (stream->pad)) {
|
2004-03-15 19:32:27 +00:00
|
|
|
|
gst_buffer_unref (buf);
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
} else {
|
2004-03-15 19:32:27 +00:00
|
|
|
|
GstClockTime dur_ts;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
|
2005-11-15 12:16:00 +00:00
|
|
|
|
if (stream->strh->fcc_handler == GST_MAKE_FOURCC ('D', 'I', 'B', ' ')) {
|
|
|
|
|
buf = gst_avi_demux_invert (stream, buf);
|
|
|
|
|
}
|
|
|
|
|
|
2004-03-15 19:32:27 +00:00
|
|
|
|
GST_BUFFER_TIMESTAMP (buf) = next_ts;
|
|
|
|
|
gst_pad_query (stream->pad, GST_QUERY_POSITION, &format, &dur_ts);
|
|
|
|
|
GST_BUFFER_DURATION (buf) = dur_ts - next_ts;
|
2005-01-13 17:11:08 +00:00
|
|
|
|
GST_DEBUG_OBJECT (avi,
|
|
|
|
|
"Pushing buffer with time=%" GST_TIME_FORMAT " over pad %s",
|
2005-11-07 10:31:32 +00:00
|
|
|
|
GST_TIME_ARGS (next_ts), GST_PAD_NAME (stream->pad));
|
2004-03-15 19:32:27 +00:00
|
|
|
|
gst_pad_push (stream->pad, GST_DATA (buf));
|
2001-12-22 23:22:30 +00:00
|
|
|
|
}
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
2001-12-22 23:22:30 +00:00
|
|
|
|
}
|
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
return TRUE;
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
#endif
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
}
|
2003-01-21 22:23:32 +00:00
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
static void
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
gst_avi_demux_loop (GstPad * pad)
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
{
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
GstFlowReturn res;
|
|
|
|
|
GstAviDemux *avi = GST_AVI_DEMUX (GST_PAD_PARENT (pad));
|
|
|
|
|
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
switch (avi->state) {
|
|
|
|
|
case GST_AVI_DEMUX_START:
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if ((res = gst_avi_demux_stream_init (avi)) != GST_FLOW_OK)
|
|
|
|
|
goto pause;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
avi->state = GST_AVI_DEMUX_HEADER;
|
|
|
|
|
/* fall-through */
|
|
|
|
|
case GST_AVI_DEMUX_HEADER:
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if ((res = gst_avi_demux_stream_header (avi)) != GST_FLOW_OK)
|
|
|
|
|
goto pause;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
avi->state = GST_AVI_DEMUX_MOVI;
|
ext/flac/gstflacdec.c: Only return true if we actually filled something in. Prevents player applications from showing...
Original commit message from CVS:
* ext/flac/gstflacdec.c: (gst_flacdec_src_query):
Only return true if we actually filled something in. Prevents
player applications from showing a random length for flac files.
* gst-libs/gst/riff/riff-read.c: (gst_riff_read_class_init),
(gst_riff_read_use_event), (gst_riff_read_handle_event),
(gst_riff_read_seek), (gst_riff_read_skip), (gst_riff_read_strh),
(gst_riff_read_strf_vids_with_data),
(gst_riff_read_strf_auds_with_data), (gst_riff_read_strf_iavs):
OK, ok, so I implemented event handling. Apparently it's normal
that we receive random events at random points without asking
for it.
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_src_convert), (gst_avi_demux_handle_src_query),
(gst_avi_demux_handle_src_event), (gst_avi_demux_stream_index),
(gst_avi_demux_sync), (gst_avi_demux_stream_scan),
(gst_avi_demux_massage_index), (gst_avi_demux_stream_header),
(gst_avi_demux_handle_seek), (gst_avi_demux_process_next_entry),
(gst_avi_demux_stream_data), (gst_avi_demux_loop):
* gst/avi/gstavidemux.h:
Implement non-lineair chunk handling and subchunk processing.
The first solves playback of AVI files where the audio and video
data of individual buffers that we read are not synchronized.
This should not happen according to the wonderful AVI specs, but
of course it does happen in reality. It is also a prerequisite for
the second. Subchunk processing allows us to cut chunks in small
pieces and process each of these pieces separately. This is
required because I've seen several AVI files with incredibly large
audio chunks, even some files with only one audio chunk for the
whole file. This allows for proper playback including seeking.
This patch is supposed to fix all AVI A/V sync issues.
* gst/flx/gstflxdec.c: (gst_flxdec_class_init),
(flx_decode_chunks), (flx_decode_color), (gst_flxdec_loop):
Work.
* gst/modplug/gstmodplug.cc:
Proper return value setting for the query() function.
* gst/playback/gstplaybasebin.c: (setup_source):
Being in non-playing state (after, e.g., EOS) is not necessarily
a bad thing. Allow for that. This fixes playback of short files.
They don't actually playback fully now, because the clock already
runs. This means that small files (<500kB) with a small length
(<2sec) will still not or barely play. Other files, such as mod
or flx, will work correctly, however.
2004-09-29 09:45:40 +00:00
|
|
|
|
break;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
case GST_AVI_DEMUX_MOVI:
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
if (avi->seek_event) {
|
2005-07-16 16:43:53 +00:00
|
|
|
|
gst_avi_demux_send_event (avi, avi->seek_event);
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
avi->seek_event = NULL;
|
|
|
|
|
}
|
2006-02-02 12:29:24 +00:00
|
|
|
|
if ((res = gst_avi_demux_stream_data (avi)) != GST_FLOW_OK) {
|
|
|
|
|
GST_DEBUG_OBJECT (avi, "stream_data flow: %s", gst_flow_get_name (res));
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
goto pause;
|
2006-02-02 12:29:24 +00:00
|
|
|
|
}
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
break;
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
default:
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
g_assert_not_reached ();
|
|
|
|
|
}
|
|
|
|
|
|
2006-02-02 12:29:24 +00:00
|
|
|
|
if (gst_avi_demux_all_source_pads_unlinked (avi)) {
|
|
|
|
|
GST_DEBUG_OBJECT (avi, "all source pads unlinked, pausing");
|
|
|
|
|
goto pause;
|
|
|
|
|
}
|
|
|
|
|
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
return;
|
|
|
|
|
|
|
|
|
|
pause:
|
|
|
|
|
GST_LOG_OBJECT (avi, "pausing task");
|
|
|
|
|
gst_pad_pause_task (avi->sinkpad);
|
2005-11-11 16:48:58 +00:00
|
|
|
|
if (GST_FLOW_IS_FATAL (res)) {
|
|
|
|
|
guint stream = avi->num_streams;
|
|
|
|
|
|
|
|
|
|
/* for fatal errors we post an error message */
|
|
|
|
|
GST_ELEMENT_ERROR (avi, STREAM, FAILED,
|
|
|
|
|
(_("Internal data stream error.")),
|
|
|
|
|
("streaming stopped, reason %s", gst_flow_get_name (res)));
|
gst/avi/gstavidemux.c: Construct index for indexless files.
Original commit message from CVS:
* gst/avi/gstavidemux.c: (gst_avi_demux_reset),
(gst_avi_demux_handle_src_event), (gst_avi_demux_parse_stream),
(gst_avi_demux_parse_odml), (gst_avi_demux_peek_tag),
(gst_avi_demux_next_data_buffer), (gst_avi_demux_stream_scan),
(gst_avi_demux_stream_header), (gst_avi_demux_loop):
Construct index for indexless files.
Make sure pad/buffers are correctly reset to NULL once we don't need
them anymore, else we get lovely segfaults/assertions.
* gst/wavparse/gstwavparse.c:
Yes, you can have 96KHz audio and wma in wav :(
2005-12-19 15:06:27 +00:00
|
|
|
|
while (stream--) {
|
|
|
|
|
if (avi->stream[stream].pad)
|
|
|
|
|
gst_pad_push_event (avi->stream[stream].pad, gst_event_new_eos ());
|
|
|
|
|
}
|
2005-11-11 16:48:58 +00:00
|
|
|
|
}
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static gboolean
|
|
|
|
|
gst_avi_demux_sink_activate (GstPad * sinkpad)
|
|
|
|
|
{
|
|
|
|
|
if (gst_pad_check_pull_range (sinkpad))
|
|
|
|
|
return gst_pad_activate_pull (sinkpad, TRUE);
|
|
|
|
|
|
|
|
|
|
return FALSE;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static gboolean
|
|
|
|
|
gst_avi_demux_sink_activate_pull (GstPad * sinkpad, gboolean active)
|
|
|
|
|
{
|
|
|
|
|
if (active) {
|
|
|
|
|
/* if we have a scheduler we can start the task */
|
|
|
|
|
gst_pad_start_task (sinkpad, (GstTaskFunction) gst_avi_demux_loop, sinkpad);
|
|
|
|
|
} else {
|
|
|
|
|
gst_pad_stop_task (sinkpad);
|
2003-01-21 22:23:32 +00:00
|
|
|
|
}
|
gst/avi/: Port from -THREADED to HEAD, and fix for latest API changes of the day. Keep avimux dead for now.
Original commit message from CVS:
* gst/avi/Makefile.am:
* gst/avi/gstavi.c: (plugin_init):
* gst/avi/gstavidemux.c: (gst_avi_demux_get_type),
(gst_avi_demux_class_init), (gst_avi_demux_init),
(gst_avi_demux_reset), (gst_avi_demux_index_next),
(gst_avi_demux_index_entry_for_time),
(gst_avi_demux_index_entry_for_byte),
(gst_avi_demux_index_entry_for_frame), (gst_avi_demux_src_convert),
(gst_avi_demux_handle_src_query), (gst_avi_demux_handle_src_event),
(gst_avi_demux_parse_file_header), (gst_avi_demux_stream_init),
(gst_avi_demux_parse_avih), (gst_avi_demux_parse_superindex),
(gst_avi_demux_parse_subindex), (gst_avi_demux_read_subindexes),
(gst_avi_demux_parse_stream), (gst_avi_demux_parse_odml),
(gst_avi_demux_parse_index), (gst_avi_demux_stream_index),
(gst_avi_demux_stream_scan), (gst_avi_demux_massage_index),
(gst_avi_demux_stream_header), (gst_avi_demux_handle_seek),
(gst_avi_demux_process_next_entry), (gst_avi_demux_stream_data),
(gst_avi_demux_loop), (gst_avi_demux_sink_activate),
(gst_avi_demux_sink_activate_pull):
* gst/avi/gstavidemux.h:
Port from -THREADED to HEAD, and fix for latest API changes of
the day. Keep avimux dead for now.
2005-06-30 09:00:36 +00:00
|
|
|
|
|
|
|
|
|
return TRUE;
|
2001-12-23 22:14:25 +00:00
|
|
|
|
}
|
|
|
|
|
|
2005-09-02 15:44:50 +00:00
|
|
|
|
static GstStateChangeReturn
|
|
|
|
|
gst_avi_demux_change_state (GstElement * element, GstStateChange transition)
|
2001-12-23 22:14:25 +00:00
|
|
|
|
{
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
GstAviDemux *avi = GST_AVI_DEMUX (element);
|
2001-12-22 23:22:30 +00:00
|
|
|
|
|
2005-09-26 13:06:27 +00:00
|
|
|
|
switch (transition) {
|
|
|
|
|
case GST_STATE_CHANGE_READY_TO_PAUSED:
|
|
|
|
|
break;
|
|
|
|
|
default:
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (GST_ELEMENT_CLASS (parent_class)->change_state) {
|
|
|
|
|
GstStateChangeReturn ret;
|
|
|
|
|
|
|
|
|
|
ret = GST_ELEMENT_CLASS (parent_class)->change_state (element, transition);
|
|
|
|
|
if (ret != GST_STATE_CHANGE_SUCCESS)
|
|
|
|
|
return ret;
|
|
|
|
|
}
|
|
|
|
|
|
2005-09-02 15:44:50 +00:00
|
|
|
|
switch (transition) {
|
|
|
|
|
case GST_STATE_CHANGE_PAUSED_TO_READY:
|
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b...
Original commit message from CVS:
Riff, EBML, fourcc etc. work. Not fully finished, but better than
what we used to have and definately worth a first broad testing.
I've revived rifflib. Rifflib used to be a bytestream-for-riff, which
just dup'ed bytestream. I've rewritten rifflib to be a modern riff-
chunk parser that uses bytestream fully, plus adds some extra functions
so that riff file parsing becomes extremely easy. It also contains some
small usability functions for strh/strf and metadata parsing. Note that
it doesn't use the new tagging yet, that's a TODO.
Avidemux has been rewritten to use this. I think we all agreed that
avidemux was pretty much a big mess, which is because it used all
sort of bytestream magic all around the place. It was just ugly.
This is a lot nicer, very complete and safe. I think this is far more
robust than what the old avidemux could ever have been. Of course, it
might contain bugs, please let me know.
EBML writing has also been implemented. This is useful for matroska.
I'm intending to modify avidemux (with a riffwriter) similarly. Maybe
I'll change wavparse/-enc too to use rifflib.
Lastly, several plugins have been modified to use rifflib's fourcc
parsing instead of their own. this puts fourcc parsing in one central
place, which should make it a lot simpler to add new fourccs. We might
want to move this to its own lib instead of rifflib.
Enjoy!
2003-12-07 20:00:41 +00:00
|
|
|
|
gst_avi_demux_reset (avi);
|
2001-12-23 22:14:25 +00:00
|
|
|
|
break;
|
|
|
|
|
default:
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
|
2005-09-02 15:44:50 +00:00
|
|
|
|
return GST_STATE_CHANGE_SUCCESS;
|
2001-12-22 23:22:30 +00:00
|
|
|
|
}
|