gstreamer/gst/matroska/matroska-mux.c

1607 lines
48 KiB
C
Raw Normal View History

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
/* GStreamer Matroska muxer/demuxer
* (c) 2003 Ronald Bultje <rbultje@ronald.bitfreak.net>
* (c) 2005 Michal Benes <michal.benes@xeris.cz>
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
*
* matroska-mux.c: matroska file/stream muxer
*
* 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.
*/
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include <math.h>
#include <string.h>
#include "matroska-mux.h"
#include "matroska-ids.h"
GST_DEBUG_CATEGORY_STATIC (matroskamux_debug);
#define GST_CAT_DEFAULT matroskamux_debug
enum
{
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
/* FILL ME */
LAST_SIGNAL
};
enum
{
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
ARG_0,
ARG_WRITING_APP,
ARG_MATROSKA_VERSION
/* FILL ME */
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 GstStaticPadTemplate src_templ = GST_STATIC_PAD_TEMPLATE ("src",
GST_PAD_SRC,
GST_PAD_ALWAYS,
GST_STATIC_CAPS ("video/x-matroska")
);
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
#define COMMON_VIDEO_CAPS \
"width = (int) [ 16, 4096 ], " \
"height = (int) [ 16, 4096 ], " \
"framerate = (fraction) [ 0, MAX ]"
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 GstStaticPadTemplate videosink_templ =
GST_STATIC_PAD_TEMPLATE ("video_%d",
GST_PAD_SINK,
GST_PAD_REQUEST,
GST_STATIC_CAPS ("video/mpeg, "
"mpegversion = (int) { 1, 2, 4 }, "
"systemstream = (boolean) false, "
COMMON_VIDEO_CAPS "; "
"video/x-h264, "
COMMON_VIDEO_CAPS "; "
"video/x-divx, "
COMMON_VIDEO_CAPS "; "
"video/x-xvid, "
COMMON_VIDEO_CAPS "; "
"video/x-msmpeg, "
COMMON_VIDEO_CAPS "; "
"image/jpeg, "
COMMON_VIDEO_CAPS "; "
"video/x-raw-yuv, "
"format = (fourcc) { YUY2, I420 }, " COMMON_VIDEO_CAPS)
);
#define COMMON_AUDIO_CAPS \
"channels = (int) [ 1, 8 ], " \
"rate = (int) [ 8000, 96000 ]"
configure.ac: Improve mpeg2enc detection. This is for distributions that do ship mjpegtools, but without mpeg2enc. Al... Original commit message from CVS: 2003-12-21 Ronald Bultje <rbultje@ronald.bitfreak.net> * configure.ac: Improve mpeg2enc detection. This is for distributions that do ship mjpegtools, but without mpeg2enc. Also does object check for might there ever be ABI incompatibility. * ext/mpeg2enc/gstmpeg2enc.cc: Add Andrew as second maintainer (he's helping me), and also add an error if no caps was set. This happens if I pull before capsnego and that's something I should solve sometime else. * gst/matroska/matroska-demux.c: (gst_matroska_demux_parse_blockgroup): Fix time parsing. * gst/matroska/matroska-mux.c: (gst_matroska_mux_audio_pad_link), (gst_matroska_mux_track_header): Add caps to templates. * gst/mpegaudioparse/gstmpegaudioparse.c: (mp3_sink_factory): Add mpegversion=1 to prevent confusion with MPEG/AAC. * gst/mpegstream/gstmpegdemux.c: Remove layer since it causes warnings about unfixed caps. * gst/videotestsrc/gstvideotestsrc.c: (gst_videotestsrc_get): Fix obvious typo (we error out if caps were set, we should of course error out if *no* caps were set). * sys/oss/gstosselement.c: (gst_osselement_convert): Fix format conversion, we confused bits/bytes. * sys/oss/gstosselement.h: Improve documentation for 'bps'. * sys/v4l/TODO: Remove stuff about plugins that need removing - this was done ages ago. * sys/v4l/gstv4lmjpegsrc.c: (gst_v4lmjpegsrc_init), (gst_v4lmjpegsrc_src_convert), (gst_v4lmjpegsrc_src_query): * sys/v4l/gstv4lsrc.c: (gst_v4lsrc_init), (gst_v4lsrc_src_convert), (gst_v4lsrc_src_query): * sys/v4l2/gstv4l2src.c: (gst_v4l2src_init), (gst_v4l2src_src_convert), (gst_v4l2src_src_query): Add get_query_types(), get_formats() and query() functions.
2003-12-21 18:44:34 +00:00
/* FIXME:
* * audio/x-raw-float: endianness needs defining.
*/
static GstStaticPadTemplate audiosink_templ =
GST_STATIC_PAD_TEMPLATE ("audio_%d",
GST_PAD_SINK,
GST_PAD_REQUEST,
GST_STATIC_CAPS ("audio/mpeg, "
"mpegversion = (int) 1, "
"layer = (int) [ 1, 3 ], "
COMMON_AUDIO_CAPS "; "
"audio/mpeg, "
"mpegversion = (int) { 2, 4 }, "
COMMON_AUDIO_CAPS "; "
"audio/x-ac3, "
COMMON_AUDIO_CAPS "; "
"audio/x-vorbis, "
COMMON_AUDIO_CAPS "; "
"audio/x-raw-int, "
"width = (int) { 8, 16, 24 }, "
"depth = (int) { 8, 16, 24 }, "
"endianness = (int) { BIG_ENDIAN, LITTLE_ENDIAN }, "
"signed = (boolean) { true, false }, "
COMMON_AUDIO_CAPS ";"
"audio/x-tta, "
"width = (int) { 8, 16, 24 }, "
"channels = (int) { 1, 2 }, " "rate = (int) [ 8000, 96000 ]")
);
static GstStaticPadTemplate subtitlesink_templ =
GST_STATIC_PAD_TEMPLATE ("subtitle_%d",
GST_PAD_SINK,
GST_PAD_REQUEST,
GST_STATIC_CAPS_ANY);
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 GArray *used_uids;
static void
_do_init (GType matroskamux_type)
{
GST_DEBUG_CATEGORY_INIT (matroskamux_debug, "matroskamux", 0,
"Matroska muxer");
}
GST_BOILERPLATE_FULL (GstMatroskaMux, gst_matroska_mux, GstElement,
GST_TYPE_ELEMENT, _do_init);
/* Matroska muxer destructor */
static void gst_matroska_mux_finalize (GObject * object);
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
/* Pads collected callback */
static GstFlowReturn
gst_matroska_mux_collected (GstCollectPads * pads, gpointer user_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
/* pad functions */
static gboolean gst_matroska_mux_handle_src_event (GstPad * pad,
GstEvent * event);
static GstPad *gst_matroska_mux_request_new_pad (GstElement * element,
GstPadTemplate * templ, const gchar * 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
/* gst internal change state handler */
static GstStateChangeReturn
gst_matroska_mux_change_state (GstElement * element, GstStateChange transition);
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
/* gobject bla bla */
static void gst_matroska_mux_set_property (GObject * object,
guint prop_id, const GValue * value, GParamSpec * pspec);
static void gst_matroska_mux_get_property (GObject * object,
guint prop_id, GValue * value, GParamSpec * pspec);
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
/* reset muxer */
static void gst_matroska_mux_reset (GstElement * element);
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
/* uid generation */
static guint32 gst_matroska_mux_create_uid ();
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 guint gst_matroska_mux_signals[LAST_SIGNAL] = { 0 };*/
static void
gst_matroska_mux_base_init (gpointer g_class)
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 (g_class);
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 GstElementDetails gst_matroska_mux_details = {
"Matroska muxer",
"Codec/Muxer",
"Muxes video/audio/subtitle streams into a matroska stream",
"Ronald Bultje <rbultje@ronald.bitfreak.net>"
};
gst_element_class_add_pad_template (element_class,
gst_static_pad_template_get (&videosink_templ));
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_class_add_pad_template (element_class,
gst_static_pad_template_get (&audiosink_templ));
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_class_add_pad_template (element_class,
gst_static_pad_template_get (&subtitlesink_templ));
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_class_add_pad_template (element_class,
gst_static_pad_template_get (&src_templ));
gst_element_class_set_details (element_class, &gst_matroska_mux_details);
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_matroska_mux_class_init (GstMatroskaMuxClass * klass)
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
{
GObjectClass *gobject_class;
GstElementClass *gstelement_class;
gobject_class = (GObjectClass *) klass;
gstelement_class = (GstElementClass *) klass;
gobject_class->finalize = GST_DEBUG_FUNCPTR (gst_matroska_mux_finalize);
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
gobject_class->get_property = gst_matroska_mux_get_property;
gobject_class->set_property = gst_matroska_mux_set_property;
g_object_class_install_property (gobject_class, ARG_WRITING_APP,
g_param_spec_string ("writing-app", "Writing application.",
"The name the application that creates the matroska file.",
NULL, G_PARAM_READWRITE));
g_object_class_install_property (gobject_class, ARG_MATROSKA_VERSION,
g_param_spec_int ("version", "Matroska version",
"This parameter determines what matroska features can be used.",
1, 2, 1, G_PARAM_READWRITE));
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
gstelement_class->change_state = gst_matroska_mux_change_state;
gstelement_class->request_new_pad = gst_matroska_mux_request_new_pad;
}
/**
* gst_matroska_mux_init:
* @mux: #GstMatroskaMux that should be initialized.
* @g_class: Class of the muxer.
*
* Matroska muxer constructor.
*/
static void
gst_matroska_mux_init (GstMatroskaMux * mux, GstMatroskaMuxClass * g_class)
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 *gstelement_class = GST_ELEMENT_CLASS (g_class);
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
mux->srcpad =
gst_pad_new_from_template (gst_element_class_get_pad_template
(gstelement_class, "src"), "src");
gst_pad_set_event_function (mux->srcpad, gst_matroska_mux_handle_src_event);
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 (mux), mux->srcpad);
mux->collect = gst_collect_pads_new ();
gst_collect_pads_set_function (mux->collect,
(GstCollectPadsFunction) gst_matroska_mux_collected, mux);
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
mux->ebml_write = gst_ebml_write_new (mux->srcpad);
/* initialize internal variables */
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
mux->index = NULL;
mux->matroska_version = 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
/* Initialize all variables */
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_matroska_mux_reset (GST_ELEMENT (mux));
}
/**
* gst_matroska_mux_finalize:
* @object: #GstMatroskaMux that should be finalized.
*
* Finalize matroska muxer.
*/
static void
gst_matroska_mux_finalize (GObject * object)
{
GstMatroskaMux *mux = GST_MATROSKA_MUX (object);
gst_object_unref (mux->collect);
gst_object_unref (mux->ebml_write);
if (mux->writing_app)
g_free (mux->writing_app);
G_OBJECT_CLASS (parent_class)->finalize (object);
}
/**
* gst_matroska_mux_create_uid:
*
* Generate new unused track UID.
*
* Returns: New track UID.
*/
static guint32
gst_matroska_mux_create_uid (void)
{
guint32 uid = 0;
GRand *rand = g_rand_new ();
while (!uid) {
guint i;
uid = g_rand_int (rand);
for (i = 0; i < used_uids->len; i++) {
if (g_array_index (used_uids, guint32, i) == uid) {
uid = 0;
break;
}
}
g_array_append_val (used_uids, uid);
}
g_free (rand);
return uid;
}
/**
* gst_matroska_mux_reset:
* @element: #GstMatroskaMux that should be reseted.
*
* Reset matroska muxer back to initial state.
*/
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_matroska_mux_reset (GstElement * element)
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
{
GstMatroskaMux *mux = GST_MATROSKA_MUX (element);
GSList *walk;
/* reset EBML write */
gst_ebml_write_reset (mux->ebml_write);
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
/* reset input */
mux->state = GST_MATROSKA_MUX_STATE_START;
/* clean up existing streams */
while ((walk = mux->collect->data) != NULL) {
GstMatroskaPad *collect_pad;
GstPad *thepad;
collect_pad = (GstMatroskaPad *) walk->data;
thepad = collect_pad->collect.pad;
/* free track information */
if (collect_pad->track != NULL) {
g_free (collect_pad->track->codec_id);
g_free (collect_pad->track->codec_name);
g_free (collect_pad->track->name);
g_free (collect_pad->track->language);
g_free (collect_pad->track->codec_priv);
g_free (collect_pad->track);
collect_pad->track = 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
}
/* free cached buffer */
if (collect_pad->buffer != NULL) {
gst_buffer_unref (collect_pad->buffer);
collect_pad->buffer = 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
}
/* remove from collectpads */
gst_collect_pads_remove_pad (mux->collect, thepad);
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
}
mux->num_streams = 0;
mux->num_a_streams = 0;
mux->num_t_streams = 0;
mux->num_v_streams = 0;
/* reset writing_app */
if (mux->writing_app) {
g_free (mux->writing_app);
}
mux->writing_app = g_strdup ("GStreamer Matroska muxer");
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
/* reset indexes */
mux->num_indexes = 0;
g_free (mux->index);
mux->index = NULL;
/* reset timers */
mux->time_scale = 1000000;
mux->duration = 0;
/* reset uid array */
if (used_uids) {
g_array_free (used_uids, TRUE);
}
/* arbitrary size, 10 should be enough in most cases */
used_uids = g_array_sized_new (FALSE, FALSE, sizeof (guint32), 10);
/* reset cluster */
mux->cluster = 0;
mux->cluster_time = 0;
mux->cluster_pos = 0;
/* reset meta-seek index */
mux->num_meta_indexes = 0;
g_free (mux->meta_index);
mux->meta_index = 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_matroska_mux_handle_src_event:
* @pad: Pad which received the event.
* @event: Received event.
*
* handle events - copied from oggmux without understanding
*
* Returns: #TRUE on success.
*/
static gboolean
gst_matroska_mux_handle_src_event (GstPad * pad, GstEvent * event)
{
GstEventType type;
type = event ? GST_EVENT_TYPE (event) : GST_EVENT_UNKNOWN;
switch (type) {
case GST_EVENT_SEEK:
/* disable seeking for now */
return FALSE;
default:
break;
}
return gst_pad_event_default (pad, event);
}
/**
* gst_matroska_mux_video_pad_setcaps:
* @pad: Pad which got the caps.
* @caps: New caps.
*
* Setcaps function for video sink pad.
*
* Returns: #TRUE on success.
*/
static gboolean
gst_matroska_mux_video_pad_setcaps (GstPad * pad, GstCaps * caps)
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
{
GstMatroskaTrackContext *context = NULL;
GstMatroskaTrackVideoContext *videocontext;
GstMatroskaPad *collect_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
const gchar *mimetype;
gint width, height, pixel_width, pixel_height;
const GValue *framerate;
GstStructure *structure;
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
/* find context */
collect_pad = (GstMatroskaPad *) gst_pad_get_element_private (pad);
g_assert (collect_pad);
context = collect_pad->track;
g_assert (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
g_assert (context->type == GST_MATROSKA_TRACK_TYPE_VIDEO);
videocontext = (GstMatroskaTrackVideoContext *) context;
/* gst -> matroska ID'ing */
structure = gst_caps_get_structure (caps, 0);
mimetype = gst_structure_get_name (structure);
/* get general properties */
gst_structure_get_int (structure, "width", &width);
gst_structure_get_int (structure, "height", &height);
framerate = gst_structure_get_value (structure, "framerate");
if (framerate == NULL || !GST_VALUE_HOLDS_FRACTION (framerate))
return FALSE;
videocontext->pixel_width = width;
videocontext->pixel_height = height;
context->default_duration = gst_util_uint64_scale_int (GST_SECOND,
gst_value_get_fraction_denominator (framerate),
gst_value_get_fraction_numerator (framerate));
if (gst_structure_get_fraction (structure, "pixel-aspect-ratio",
&pixel_width, &pixel_height)) {
if (pixel_width > pixel_height) {
videocontext->display_width = width * pixel_width / pixel_height;
videocontext->display_height = height;
} else if (pixel_width < pixel_height) {
videocontext->display_width = width;
videocontext->display_height = height * pixel_height / pixel_width;
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 {
videocontext->display_width = 0;
videocontext->display_height = 0;
}
} else {
videocontext->display_width = 0;
videocontext->display_height = 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
videocontext->asr_mode = GST_MATROSKA_ASPECT_RATIO_MODE_FREE;
videocontext->eye_mode = GST_MATROSKA_EYE_MODE_MONO;
videocontext->fourcc = 0;
/* find type */
if (!strcmp (mimetype, "video/x-raw-yuv")) {
context->codec_id = g_strdup (GST_MATROSKA_CODEC_ID_VIDEO_UNCOMPRESSED);
gst_structure_get_fourcc (structure, "format", &videocontext->fourcc);
return TRUE;
} else if (!strcmp (mimetype, "image/jpeg")) {
context->codec_id = g_strdup (GST_MATROSKA_CODEC_ID_VIDEO_MJPEG);
return TRUE;
} else if (!strcmp (mimetype, "video/x-divx")) {
gint divxversion;
BITMAPINFOHEADER *bih;
bih = (BITMAPINFOHEADER *) g_malloc0 (sizeof (BITMAPINFOHEADER));
GST_WRITE_UINT32_LE (&bih->bi_size, sizeof (BITMAPINFOHEADER));
GST_WRITE_UINT32_LE (&bih->bi_width, videocontext->pixel_width);
GST_WRITE_UINT32_LE (&bih->bi_height, videocontext->pixel_height);
GST_WRITE_UINT16_LE (&bih->bi_planes, (guint16) 1);
GST_WRITE_UINT16_LE (&bih->bi_bit_count, (guint16) 24);
GST_WRITE_UINT32_LE (&bih->bi_size_image, videocontext->pixel_width *
videocontext->pixel_height * 3);
gst_structure_get_int (structure, "divxversion", &divxversion);
switch (divxversion) {
case 3:
GST_WRITE_UINT32_LE (&bih->bi_compression, GST_STR_FOURCC ("DIV3"));
break;
case 4:
GST_WRITE_UINT32_LE (&bih->bi_compression, GST_STR_FOURCC ("DIVX"));
break;
case 5:
GST_WRITE_UINT32_LE (&bih->bi_compression, GST_STR_FOURCC ("DX50"));
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
context->codec_id = g_strdup (GST_MATROSKA_CODEC_ID_VIDEO_VFW_FOURCC);
context->codec_priv = (gpointer) bih;
context->codec_priv_size = sizeof (BITMAPINFOHEADER);
return TRUE;
} else if (!strcmp (mimetype, "video/x-xvid")) {
BITMAPINFOHEADER *bih;
bih = (BITMAPINFOHEADER *) g_malloc0 (sizeof (BITMAPINFOHEADER));
GST_WRITE_UINT32_LE (&bih->bi_size, sizeof (BITMAPINFOHEADER));
GST_WRITE_UINT32_LE (&bih->bi_width, videocontext->pixel_width);
GST_WRITE_UINT32_LE (&bih->bi_height, videocontext->pixel_height);
GST_WRITE_UINT16_LE (&bih->bi_planes, (guint16) 1);
GST_WRITE_UINT16_LE (&bih->bi_bit_count, (guint16) 24);
GST_WRITE_UINT32_LE (&bih->bi_compression, GST_STR_FOURCC ("XVID"));
GST_WRITE_UINT32_LE (&bih->bi_size_image, videocontext->pixel_width *
videocontext->pixel_height * 3);
context->codec_id = g_strdup (GST_MATROSKA_CODEC_ID_VIDEO_VFW_FOURCC);
context->codec_priv = (gpointer) bih;
context->codec_priv_size = sizeof (BITMAPINFOHEADER);
return TRUE;
} else if (!strcmp (mimetype, "video/x-h264")) {
const GValue *codec_data;
context->codec_id = g_strdup (GST_MATROSKA_CODEC_ID_VIDEO_MPEG4_AVC);
if (context->codec_priv != NULL) {
g_free (context->codec_priv);
context->codec_priv = NULL;
context->codec_priv_size = 0;
}
/* Create avcC header */
codec_data = gst_structure_get_value (structure, "codec_data");
if (codec_data != NULL) {
guint8 *priv_data = NULL;
guint priv_data_size = 0;
GstBuffer *codec_data_buf = g_value_peek_pointer (codec_data);
priv_data_size = GST_BUFFER_SIZE (codec_data_buf);
priv_data = g_malloc0 (priv_data_size);
memcpy (priv_data, GST_BUFFER_DATA (codec_data_buf), priv_data_size);
context->codec_priv = priv_data;
context->codec_priv_size = priv_data_size;
}
return TRUE;
} else if (!strcmp (mimetype, "video/mpeg")) {
gint mpegversion;
gst_structure_get_int (structure, "mpegversion", &mpegversion);
switch (mpegversion) {
case 1:
context->codec_id = g_strdup (GST_MATROSKA_CODEC_ID_VIDEO_MPEG1);
break;
case 2:
context->codec_id = g_strdup (GST_MATROSKA_CODEC_ID_VIDEO_MPEG2);
break;
case 4:
context->codec_id = g_strdup (GST_MATROSKA_CODEC_ID_VIDEO_MPEG4_ASP);
break;
default:
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
return TRUE;
} else if (!strcmp (mimetype, "video/x-msmpeg")) {
context->codec_id = g_strdup (GST_MATROSKA_CODEC_ID_VIDEO_MSMPEG4V3);
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;
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;
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_matroska_mux_audio_pad_setcaps:
* @pad: Pad which got the caps.
* @caps: New caps.
*
* Setcaps function for audio sink pad.
*
* Returns: #TRUE on success.
*/
static gboolean
gst_matroska_mux_audio_pad_setcaps (GstPad * pad, GstCaps * caps)
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
{
GstMatroskaTrackContext *context = NULL;
GstMatroskaTrackAudioContext *audiocontext;
GstMatroskaPad *collect_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
const gchar *mimetype;
gint samplerate = 0, channels = 0;
GstStructure *structure;
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
/* find context */
collect_pad = (GstMatroskaPad *) gst_pad_get_element_private (pad);
g_assert (collect_pad);
context = collect_pad->track;
g_assert (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
g_assert (context->type == GST_MATROSKA_TRACK_TYPE_AUDIO);
audiocontext = (GstMatroskaTrackAudioContext *) context;
structure = gst_caps_get_structure (caps, 0);
mimetype = gst_structure_get_name (structure);
/* general setup */
gst_structure_get_int (structure, "rate", &samplerate);
gst_structure_get_int (structure, "channels", &channels);
audiocontext->samplerate = samplerate;
audiocontext->channels = channels;
audiocontext->bitdepth = 0;
context->default_duration = 0;
if (!strcmp (mimetype, "audio/mpeg")) {
gint mpegversion = 0;
gst_structure_get_int (structure, "mpegversion", &mpegversion);
switch (mpegversion) {
case 1:{
gint layer;
gst_structure_get_int (structure, "layer", &layer);
switch (layer) {
case 1:
context->codec_id = g_strdup (GST_MATROSKA_CODEC_ID_AUDIO_MPEG1_L1);
context->default_duration =
384 * GST_SECOND / audiocontext->samplerate;
break;
case 2:
context->codec_id = g_strdup (GST_MATROSKA_CODEC_ID_AUDIO_MPEG1_L2);
context->default_duration =
1152 * GST_SECOND / audiocontext->samplerate;
break;
case 3:
context->codec_id = g_strdup (GST_MATROSKA_CODEC_ID_AUDIO_MPEG1_L3);
context->default_duration =
1152 * GST_SECOND / audiocontext->samplerate;
break;
default:
return FALSE;
}
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 2:
context->codec_id = g_strdup (GST_MATROSKA_CODEC_ID_AUDIO_MPEG2 "MAIN");
break;
case 4:
context->codec_id = g_strdup (GST_MATROSKA_CODEC_ID_AUDIO_MPEG4 "MAIN");
break;
default:
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
}
return TRUE;
} else if (!strcmp (mimetype, "audio/x-raw-int")) {
gint endianness, width, depth;
gboolean signedness;
if (!gst_structure_get_int (structure, "endianness", &endianness) ||
!gst_structure_get_int (structure, "width", &width) ||
!gst_structure_get_int (structure, "depth", &depth) ||
!gst_structure_get_int (structure, "signed", &signedness))
return FALSE;
if (width != depth ||
(width == 8 && signedness) || (width == 16 && !signedness))
return FALSE;
audiocontext->bitdepth = depth;
if (endianness == G_BIG_ENDIAN)
context->codec_id = g_strdup (GST_MATROSKA_CODEC_ID_AUDIO_PCM_INT_BE);
else
context->codec_id = g_strdup (GST_MATROSKA_CODEC_ID_AUDIO_PCM_INT_LE);
return TRUE;
} else if (!strcmp (mimetype, "audio/x-raw-float")) {
/* FIXME: endianness is undefined */
} else if (!strcmp (mimetype, "audio/x-vorbis")) {
const GValue *streamheader;
guint8 *priv_data = NULL;
guint priv_data_size = 0;
context->codec_id = g_strdup (GST_MATROSKA_CODEC_ID_AUDIO_VORBIS);
if (context->codec_priv != NULL) {
g_free (context->codec_priv);
context->codec_priv = NULL;
context->codec_priv_size = 0;
}
streamheader = gst_structure_get_value (structure, "streamheader");
if (streamheader != NULL) {
if (G_VALUE_TYPE (streamheader) == GST_TYPE_ARRAY) {
GArray *bufarr = g_value_peek_pointer (streamheader);
gint i;
gint offset;
if (bufarr->len == 3) {
GstBuffer *buf[3];
for (i = 0; i < bufarr->len; i++) {
GValue *bufval = &g_array_index (bufarr, GValue, i);
if (G_VALUE_TYPE (bufval) == GST_TYPE_BUFFER) {
buf[i] = g_value_peek_pointer (bufval);
}
}
priv_data_size = 1;
priv_data_size += GST_BUFFER_SIZE (buf[0]) / 0xff + 1;
priv_data_size += GST_BUFFER_SIZE (buf[1]) / 0xff + 1;
for (i = 0; i < 3; ++i) {
priv_data_size += GST_BUFFER_SIZE (buf[i]);
}
priv_data = g_malloc0 (priv_data_size);
priv_data[0] = 2;
offset = 1;
for (i = 0; i < GST_BUFFER_SIZE (buf[0]) / 0xff; ++i) {
priv_data[offset++] = 0xff;
}
priv_data[offset++] = GST_BUFFER_SIZE (buf[0]) % 0xff;
for (i = 0; i < GST_BUFFER_SIZE (buf[1]) / 0xff; ++i) {
priv_data[offset++] = 0xff;
}
priv_data[offset++] = GST_BUFFER_SIZE (buf[1]) % 0xff;
for (i = 0; i < 3; ++i) {
memcpy (priv_data + offset, GST_BUFFER_DATA (buf[i]),
GST_BUFFER_SIZE (buf[i]));
offset += GST_BUFFER_SIZE (buf[i]);
}
if (GST_BUFFER_SIZE (buf[0]) < 1 + 6 + 4) {
GST_WARNING ("First vorbis header too small, ignoring");
} else {
if (memcmp (GST_BUFFER_DATA (buf[0]) + 1, "vorbis", 6) == 0) {
guint8 *hdr = GST_BUFFER_DATA (buf[0]) + 1 + 6 + 4;
audiocontext->channels = GST_READ_UINT8 (hdr);
audiocontext->samplerate = GST_READ_UINT32_LE (hdr + 1);
}
}
} else {
GST_WARNING ("Vorbis header does not contain "
"three buffers (found %d buffers), Ignoring.", bufarr->len);
}
}
}
if (priv_data == NULL) {
GST_WARNING ("Could not write Vorbis header into codec private data. "
"You will probably not be able to play the stream.");
}
context->codec_priv_size = priv_data_size;
context->codec_priv = (gpointer) priv_data;
return TRUE;
} else if (!strcmp (mimetype, "audio/x-ac3")) {
context->codec_id = g_strdup (GST_MATROSKA_CODEC_ID_AUDIO_AC3);
return TRUE;
} else if (!strcmp (mimetype, "audio/x-tta")) {
gint width;
/* TTA frame duration */
context->default_duration = 1.04489795918367346939 * GST_SECOND;
gst_structure_get_int (structure, "width", &width);
audiocontext->bitdepth = width;
context->codec_id = g_strdup (GST_MATROSKA_CODEC_ID_AUDIO_TTA);
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
}
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_matroska_mux_subtitle_pad_setcaps:
* @pad: Pad which got the caps.
* @caps: New caps.
*
* Setcaps function for subtitle sink pad.
*
* Returns: #TRUE on success.
*/
static gboolean
gst_matroska_mux_subtitle_pad_setcaps (GstPad * pad, GstCaps * caps)
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
{
/* Consider this as boilerplate code for now. There is
* no single subtitle creation element in GStreamer,
* neither do I know how subtitling works at all. */
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_matroska_mux_request_new_pad:
* @element: #GstMatroskaMux.
* @templ: #GstPadTemplate.
* @pad_name: New pad name.
*
* Request pad function for sink templates.
*
* Returns: New #GstPad.
*/
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 GstPad *
gst_matroska_mux_request_new_pad (GstElement * element,
GstPadTemplate * templ, const gchar * pad_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
{
GstElementClass *klass = GST_ELEMENT_GET_CLASS (element);
GstMatroskaMux *mux = GST_MATROSKA_MUX (element);
GstMatroskaPad *collect_pad;
GstPad *newpad = 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
gchar *name = NULL;
GstPadSetCapsFunction setcapsfunc = 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
GstMatroskaTrackContext *context = NULL;
if (templ == gst_element_class_get_pad_template (klass, "audio_%d")) {
name = g_strdup_printf ("audio_%d", mux->num_a_streams++);
setcapsfunc = gst_matroska_mux_audio_pad_setcaps;
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
context = (GstMatroskaTrackContext *)
g_new0 (GstMatroskaTrackAudioContext, 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
context->type = GST_MATROSKA_TRACK_TYPE_AUDIO;
context->name = g_strdup ("Audio");
} else if (templ == gst_element_class_get_pad_template (klass, "video_%d")) {
name = g_strdup_printf ("video_%d", mux->num_v_streams++);
setcapsfunc = gst_matroska_mux_video_pad_setcaps;
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
context = (GstMatroskaTrackContext *)
g_new0 (GstMatroskaTrackVideoContext, 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
context->type = GST_MATROSKA_TRACK_TYPE_VIDEO;
context->name = g_strdup ("Video");
} else if (templ == gst_element_class_get_pad_template (klass, "subtitle_%d")) {
name = g_strdup_printf ("subtitle_%d", mux->num_t_streams++);
setcapsfunc = gst_matroska_mux_subtitle_pad_setcaps;
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
context = (GstMatroskaTrackContext *)
g_new0 (GstMatroskaTrackSubtitleContext, 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
context->type = GST_MATROSKA_TRACK_TYPE_SUBTITLE;
context->name = g_strdup ("Subtitle");
} else {
GST_WARNING_OBJECT (mux, "This is not our template!");
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 NULL;
}
newpad = gst_pad_new_from_template (templ, 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
g_free (name);
collect_pad = (GstMatroskaPad *)
gst_collect_pads_add_pad (mux->collect, newpad, sizeof (GstMatroskaPad));
context->flags = GST_MATROSKA_TRACK_ENABLED | GST_MATROSKA_TRACK_DEFAULT;
collect_pad->track = context;
collect_pad->buffer = NULL;
gst_pad_set_setcaps_function (newpad, setcapsfunc);
gst_element_add_pad (element, newpad);
return newpad;
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_matroska_mux_track_header:
* @mux: #GstMatroskaMux
* @context: Tack context.
*
* Write a track header.
*/
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_matroska_mux_track_header (GstMatroskaMux * mux,
GstMatroskaTrackContext * 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
{
GstEbmlWrite *ebml = mux->ebml_write;
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
guint64 master;
/* track type goes before the type-specific stuff */
gst_ebml_write_uint (ebml, GST_MATROSKA_ID_TRACKNUMBER, context->num);
gst_ebml_write_uint (ebml, GST_MATROSKA_ID_TRACKTYPE, context->type);
gst_ebml_write_uint (ebml, GST_MATROSKA_ID_TRACKUID,
gst_matroska_mux_create_uid ());
if (context->default_duration) {
gst_ebml_write_uint (ebml, GST_MATROSKA_ID_TRACKDEFAULTDURATION,
context->default_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
/* type-specific stuff */
switch (context->type) {
case GST_MATROSKA_TRACK_TYPE_VIDEO:{
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
GstMatroskaTrackVideoContext *videocontext =
(GstMatroskaTrackVideoContext *) 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
master = gst_ebml_write_master_start (ebml, GST_MATROSKA_ID_TRACKVIDEO);
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_ebml_write_uint (ebml, GST_MATROSKA_ID_VIDEOPIXELWIDTH,
videocontext->pixel_width);
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_ebml_write_uint (ebml, GST_MATROSKA_ID_VIDEOPIXELHEIGHT,
videocontext->pixel_height);
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 (videocontext->display_width && videocontext->display_height) {
gst_ebml_write_uint (ebml, GST_MATROSKA_ID_VIDEODISPLAYWIDTH,
videocontext->display_width);
gst_ebml_write_uint (ebml, GST_MATROSKA_ID_VIDEODISPLAYHEIGHT,
videocontext->display_height);
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 (context->flags & GST_MATROSKA_VIDEOTRACK_INTERLACED)
gst_ebml_write_uint (ebml, GST_MATROSKA_ID_VIDEOFLAGINTERLACED, 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 (videocontext->fourcc) {
guint32 fcc_le = GUINT32_TO_LE (videocontext->fourcc);
gst_ebml_write_binary (ebml, GST_MATROSKA_ID_VIDEOCOLOURSPACE,
(gpointer) & fcc_le, 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
}
gst_ebml_write_master_finish (ebml, master);
break;
}
case GST_MATROSKA_TRACK_TYPE_AUDIO:{
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
GstMatroskaTrackAudioContext *audiocontext =
(GstMatroskaTrackAudioContext *) 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
master = gst_ebml_write_master_start (ebml, GST_MATROSKA_ID_TRACKAUDIO);
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 (audiocontext->samplerate != 8000)
gst_ebml_write_float (ebml, GST_MATROSKA_ID_AUDIOSAMPLINGFREQ,
audiocontext->samplerate);
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 (audiocontext->channels != 1)
gst_ebml_write_uint (ebml, GST_MATROSKA_ID_AUDIOCHANNELS,
audiocontext->channels);
configure.ac: Improve mpeg2enc detection. This is for distributions that do ship mjpegtools, but without mpeg2enc. Al... Original commit message from CVS: 2003-12-21 Ronald Bultje <rbultje@ronald.bitfreak.net> * configure.ac: Improve mpeg2enc detection. This is for distributions that do ship mjpegtools, but without mpeg2enc. Also does object check for might there ever be ABI incompatibility. * ext/mpeg2enc/gstmpeg2enc.cc: Add Andrew as second maintainer (he's helping me), and also add an error if no caps was set. This happens if I pull before capsnego and that's something I should solve sometime else. * gst/matroska/matroska-demux.c: (gst_matroska_demux_parse_blockgroup): Fix time parsing. * gst/matroska/matroska-mux.c: (gst_matroska_mux_audio_pad_link), (gst_matroska_mux_track_header): Add caps to templates. * gst/mpegaudioparse/gstmpegaudioparse.c: (mp3_sink_factory): Add mpegversion=1 to prevent confusion with MPEG/AAC. * gst/mpegstream/gstmpegdemux.c: Remove layer since it causes warnings about unfixed caps. * gst/videotestsrc/gstvideotestsrc.c: (gst_videotestsrc_get): Fix obvious typo (we error out if caps were set, we should of course error out if *no* caps were set). * sys/oss/gstosselement.c: (gst_osselement_convert): Fix format conversion, we confused bits/bytes. * sys/oss/gstosselement.h: Improve documentation for 'bps'. * sys/v4l/TODO: Remove stuff about plugins that need removing - this was done ages ago. * sys/v4l/gstv4lmjpegsrc.c: (gst_v4lmjpegsrc_init), (gst_v4lmjpegsrc_src_convert), (gst_v4lmjpegsrc_src_query): * sys/v4l/gstv4lsrc.c: (gst_v4lsrc_init), (gst_v4lsrc_src_convert), (gst_v4lsrc_src_query): * sys/v4l2/gstv4l2src.c: (gst_v4l2src_init), (gst_v4l2src_src_convert), (gst_v4l2src_src_query): Add get_query_types(), get_formats() and query() functions.
2003-12-21 18:44:34 +00:00
if (audiocontext->bitdepth) {
gst_ebml_write_uint (ebml, GST_MATROSKA_ID_AUDIOBITDEPTH,
audiocontext->bitdepth);
configure.ac: Improve mpeg2enc detection. This is for distributions that do ship mjpegtools, but without mpeg2enc. Al... Original commit message from CVS: 2003-12-21 Ronald Bultje <rbultje@ronald.bitfreak.net> * configure.ac: Improve mpeg2enc detection. This is for distributions that do ship mjpegtools, but without mpeg2enc. Also does object check for might there ever be ABI incompatibility. * ext/mpeg2enc/gstmpeg2enc.cc: Add Andrew as second maintainer (he's helping me), and also add an error if no caps was set. This happens if I pull before capsnego and that's something I should solve sometime else. * gst/matroska/matroska-demux.c: (gst_matroska_demux_parse_blockgroup): Fix time parsing. * gst/matroska/matroska-mux.c: (gst_matroska_mux_audio_pad_link), (gst_matroska_mux_track_header): Add caps to templates. * gst/mpegaudioparse/gstmpegaudioparse.c: (mp3_sink_factory): Add mpegversion=1 to prevent confusion with MPEG/AAC. * gst/mpegstream/gstmpegdemux.c: Remove layer since it causes warnings about unfixed caps. * gst/videotestsrc/gstvideotestsrc.c: (gst_videotestsrc_get): Fix obvious typo (we error out if caps were set, we should of course error out if *no* caps were set). * sys/oss/gstosselement.c: (gst_osselement_convert): Fix format conversion, we confused bits/bytes. * sys/oss/gstosselement.h: Improve documentation for 'bps'. * sys/v4l/TODO: Remove stuff about plugins that need removing - this was done ages ago. * sys/v4l/gstv4lmjpegsrc.c: (gst_v4lmjpegsrc_init), (gst_v4lmjpegsrc_src_convert), (gst_v4lmjpegsrc_src_query): * sys/v4l/gstv4lsrc.c: (gst_v4lsrc_init), (gst_v4lsrc_src_convert), (gst_v4lsrc_src_query): * sys/v4l2/gstv4l2src.c: (gst_v4l2src_init), (gst_v4l2src_src_convert), (gst_v4l2src_src_query): Add get_query_types(), get_formats() and query() functions.
2003-12-21 18:44:34 +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
gst_ebml_write_master_finish (ebml, master);
break;
}
default:
/* doesn't need type-specific data */
break;
}
gst_ebml_write_ascii (ebml, GST_MATROSKA_ID_CODECID, context->codec_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 (context->codec_priv)
gst_ebml_write_binary (ebml, GST_MATROSKA_ID_CODECPRIVATE,
context->codec_priv, context->codec_priv_size);
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: until we have a nice way of getting the codecname
* out of the caps, I'm not going to enable this. Too much
* (useless, double, boring) work... */
/*gst_ebml_write_utf8 (ebml, GST_MATROSKA_ID_CODECNAME,
context->codec_name); */
gst_ebml_write_utf8 (ebml, GST_MATROSKA_ID_TRACKNAME, context->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
}
/**
* gst_matroska_mux_start:
* @mux: #GstMatroskaMux
*
* Start a new matroska file (write headers etc...)
*/
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_matroska_mux_start (GstMatroskaMux * mux)
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
{
GstEbmlWrite *ebml = mux->ebml_write;
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 seekhead_id[] = { GST_MATROSKA_ID_INFO,
GST_MATROSKA_ID_TRACKS,
GST_MATROSKA_ID_CUES,
GST_MATROSKA_ID_SEEKHEAD,
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 0
GST_MATROSKA_ID_TAGS,
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
#endif
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
guint64 master, child;
GSList *collected;
int i;
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
guint tracknum = 1;
gdouble duration = 0;
guint32 *segment_uid = (guint32 *) g_malloc (16);
GRand *rand = g_rand_new ();
GTimeVal time = { 0, 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
/* we start with a EBML header */
gst_ebml_write_header (ebml, "matroska", mux->matroska_version);
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
/* start a segment */
mux->segment_pos =
gst_ebml_write_master_start (ebml, GST_MATROSKA_ID_SEGMENT);
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
mux->segment_master = ebml->pos;
/* the rest of the header is cached */
gst_ebml_write_set_cache (ebml, 0x1000);
/* seekhead (table of contents) - we set the positions later */
mux->seekhead_pos = ebml->pos;
master = gst_ebml_write_master_start (ebml, GST_MATROSKA_ID_SEEKHEAD);
for (i = 0; seekhead_id[i] != 0; i++) {
child = gst_ebml_write_master_start (ebml, GST_MATROSKA_ID_SEEKENTRY);
gst_ebml_write_uint (ebml, GST_MATROSKA_ID_SEEKID, seekhead_id[i]);
gst_ebml_write_uint (ebml, GST_MATROSKA_ID_SEEKPOSITION, -1);
gst_ebml_write_master_finish (ebml, child);
}
gst_ebml_write_master_finish (ebml, master);
/* segment info */
mux->info_pos = ebml->pos;
master = gst_ebml_write_master_start (ebml, GST_MATROSKA_ID_INFO);
for (i = 0; i < 4; i++) {
segment_uid[i] = g_rand_int (rand);
}
g_free (rand);
gst_ebml_write_binary (ebml, GST_MATROSKA_ID_SEGMENTUID,
(guint8 *) segment_uid, 16);
g_free (segment_uid);
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_ebml_write_uint (ebml, GST_MATROSKA_ID_TIMECODESCALE, mux->time_scale);
mux->duration_pos = ebml->pos;
/* get duration */
for (collected = mux->collect->data; collected;
collected = g_slist_next (collected)) {
GstMatroskaPad *collect_pad;
GstFormat format = GST_FORMAT_TIME;
GstPad *thepad, *peerpad;
gint64 trackduration;
collect_pad = (GstMatroskaPad *) collected->data;
thepad = collect_pad->collect.pad;
/* Query the total length of the track. */
peerpad = gst_pad_get_peer (thepad);
GST_DEBUG_OBJECT (thepad, "querying duration");
if (gst_pad_query_duration (peerpad, &format, &trackduration)) {
GST_DEBUG_OBJECT (thepad, "%duration: %" GST_TIME_FORMAT,
GST_TIME_ARGS (trackduration));
if ((gdouble) trackduration > duration) {
duration = (gdouble) trackduration;
}
}
gst_object_unref (peerpad);
}
gst_ebml_write_float (ebml, GST_MATROSKA_ID_DURATION,
duration / mux->time_scale);
gst_ebml_write_utf8 (ebml, GST_MATROSKA_ID_MUXINGAPP,
"GStreamer plugin version " GST_PLUGINS_GOOD_VERSION);
if (mux->writing_app && mux->writing_app[0]) {
gst_ebml_write_utf8 (ebml, GST_MATROSKA_ID_WRITINGAPP, mux->writing_app);
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_get_current_time (&time);
gst_ebml_write_date (ebml, GST_MATROSKA_ID_DATEUTC, time.tv_sec);
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_ebml_write_master_finish (ebml, master);
/* tracks */
mux->tracks_pos = ebml->pos;
master = gst_ebml_write_master_start (ebml, GST_MATROSKA_ID_TRACKS);
for (collected = mux->collect->data; collected;
collected = g_slist_next (collected)) {
GstMatroskaPad *collect_pad;
GstPad *thepad;
collect_pad = (GstMatroskaPad *) collected->data;
thepad = collect_pad->collect.pad;
if (gst_pad_is_linked (thepad) && gst_pad_is_active (thepad) &&
collect_pad->track->codec_id != 0) {
collect_pad->track->num = tracknum++;
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
child = gst_ebml_write_master_start (ebml, GST_MATROSKA_ID_TRACKENTRY);
gst_matroska_mux_track_header (mux, collect_pad->track);
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_ebml_write_master_finish (ebml, child);
}
}
gst_ebml_write_master_finish (ebml, master);
/* lastly, flush the cache */
gst_ebml_write_flush_cache (ebml);
}
/**
* gst_matroska_mux_finish:
* @mux: #GstMatroskaMux
*
* Finish a new matroska file (write index etc...)
*/
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_matroska_mux_finish (GstMatroskaMux * mux)
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
{
GstEbmlWrite *ebml = mux->ebml_write;
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
guint64 pos;
guint64 duration = 0;
GSList *collected;
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
/* finish last cluster */
if (mux->cluster) {
gst_ebml_write_master_finish (ebml, mux->cluster);
}
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
/* cues */
if (mux->index != NULL) {
guint n;
guint64 master, pointentry_master, trackpos_master;
mux->cues_pos = ebml->pos;
gst_ebml_write_set_cache (ebml, 12 + 41 * mux->num_indexes);
master = gst_ebml_write_master_start (ebml, GST_MATROSKA_ID_CUES);
for (n = 0; n < mux->num_indexes; n++) {
GstMatroskaIndex *idx = &mux->index[n];
pointentry_master = gst_ebml_write_master_start (ebml,
GST_MATROSKA_ID_POINTENTRY);
gst_ebml_write_uint (ebml, GST_MATROSKA_ID_CUETIME,
idx->time / mux->time_scale);
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
trackpos_master = gst_ebml_write_master_start (ebml,
GST_MATROSKA_ID_CUETRACKPOSITION);
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_ebml_write_uint (ebml, GST_MATROSKA_ID_CUETRACK, idx->track);
gst_ebml_write_uint (ebml, GST_MATROSKA_ID_CUECLUSTERPOSITION,
idx->pos - mux->segment_master);
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_ebml_write_master_finish (ebml, trackpos_master);
gst_ebml_write_master_finish (ebml, pointentry_master);
}
gst_ebml_write_master_finish (ebml, master);
gst_ebml_write_flush_cache (ebml);
}
if (mux->meta_index != NULL) {
guint n;
guint64 master, seekentry_master;
mux->meta_pos = ebml->pos;
gst_ebml_write_set_cache (ebml, 12 + 28 * mux->num_meta_indexes);
master = gst_ebml_write_master_start (ebml, GST_MATROSKA_ID_SEEKHEAD);
for (n = 0; n < mux->num_meta_indexes; n++) {
GstMatroskaMetaSeekIndex *idx = &mux->meta_index[n];
seekentry_master = gst_ebml_write_master_start (ebml,
GST_MATROSKA_ID_SEEKENTRY);
gst_ebml_write_uint (ebml, GST_MATROSKA_ID_SEEKID, idx->id);
gst_ebml_write_uint (ebml, GST_MATROSKA_ID_SEEKPOSITION,
idx->pos - mux->segment_master);
gst_ebml_write_master_finish (ebml, seekentry_master);
}
gst_ebml_write_master_finish (ebml, master);
}
gst_ebml_write_flush_cache (ebml);
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: tags */
/* update seekhead. We know that:
* - a seekhead contains 4 entries.
* - order of entries is as above.
* - a seekhead has a 4-byte header + 8-byte length
* - each entry is 2-byte master, 2-byte ID pointer,
* 2-byte length pointer, all 8/1-byte length, 4-
* byte ID and 8-byte length pointer, where the
* length pointer starts at 20.
* - all entries are local to the segment (so pos - segment_master).
* - so each entry is at 12 + 20 + num * 28. */
gst_ebml_replace_uint (ebml, mux->seekhead_pos + 32,
mux->info_pos - mux->segment_master);
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_ebml_replace_uint (ebml, mux->seekhead_pos + 60,
mux->tracks_pos - mux->segment_master);
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 (mux->index != NULL) {
gst_ebml_replace_uint (ebml, mux->seekhead_pos + 88,
mux->cues_pos - mux->segment_master);
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 {
/* void'ify */
guint64 my_pos = ebml->pos;
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_ebml_write_seek (ebml, mux->seekhead_pos + 68);
gst_ebml_write_buffer_header (ebml, GST_EBML_ID_VOID, 26);
gst_ebml_write_seek (ebml, my_pos);
}
if (mux->meta_index != NULL) {
gst_ebml_replace_uint (ebml, mux->seekhead_pos + 116,
mux->meta_pos - mux->segment_master);
} else {
/* void'ify */
guint64 my_pos = ebml->pos;
gst_ebml_write_seek (ebml, mux->seekhead_pos + 96);
gst_ebml_write_buffer_header (ebml, GST_EBML_ID_VOID, 26);
gst_ebml_write_seek (ebml, my_pos);
}
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 0
gst_ebml_replace_uint (ebml, mux->seekhead_pos + 116,
mux->tags_pos - mux->segment_master);
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
#endif
/* update duration */
/* first get the overall duration */
for (collected = mux->collect->data; collected;
collected = g_slist_next (collected)) {
GstMatroskaPad *collect_pad;
collect_pad = (GstMatroskaPad *) collected->data;
if (collect_pad->duration > duration)
duration = collect_pad->duration;
}
if (duration != 0) {
pos = mux->ebml_write->pos;
gst_ebml_write_seek (ebml, mux->duration_pos);
gst_ebml_write_float (ebml, GST_MATROSKA_ID_DURATION,
(gdouble) duration / mux->time_scale);
gst_ebml_write_seek (ebml, pos);
}
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
/* finish segment - this also writes element length */
gst_ebml_write_master_finish (ebml, mux->segment_pos);
}
/**
* gst_matroska_mux_best_pad:
* @mux: #GstMatroskaMux
*
* Find a pad with the oldest data
* (data from this pad should be written first).
*
* Returns: Selected pad.
*/
static GstMatroskaPad *
gst_matroska_mux_best_pad (GstMatroskaMux * mux)
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
{
GSList *collected;
GstMatroskaPad *best = NULL;
for (collected = mux->collect->data; collected;
collected = g_slist_next (collected)) {
GstMatroskaPad *collect_pad;
collect_pad = (GstMatroskaPad *) collected->data;
/* fetch a new buffer if needed */
if (collect_pad->buffer == NULL) {
collect_pad->buffer = gst_collect_pads_pop (mux->collect,
(GstCollectData *) collect_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
}
/* if we have a buffer check if it is better then the current best one */
if (collect_pad->buffer != NULL) {
if (best == NULL
|| GST_BUFFER_TIMESTAMP (collect_pad->buffer) <
GST_BUFFER_TIMESTAMP (best->buffer)) {
best = collect_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
}
}
return best;
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_matroska_mux_buffer_header:
* @track: Track context.
* @relative_timestamp: relative timestamp of the buffer
* @flags: Buffer flags.
*
* Create a buffer containing buffer header.
*
* Returns: New buffer.
*/
GstBuffer *
gst_matroska_mux_create_buffer_header (GstMatroskaTrackContext * track,
guint16 relative_timestamp, int flags)
{
GstBuffer *hdr;
hdr = gst_buffer_new_and_alloc (4);
/* track num - FIXME: what if num >= 0x80 (unlikely)? */
GST_BUFFER_DATA (hdr)[0] = track->num | 0x80;
/* time relative to clustertime */
GST_WRITE_UINT16_BE (GST_BUFFER_DATA (hdr) + 1, relative_timestamp);
/* flags */
GST_BUFFER_DATA (hdr)[3] = flags;
return hdr;
}
/**
* gst_matroska_mux_write_data:
* @mux: #GstMatroskaMux
*
* Write collected data (called from gst_matroska_mux_collected).
*
* Returns: Result of the gst_pad_push issued to write the data.
*/
static GstFlowReturn
gst_matroska_mux_write_data (GstMatroskaMux * mux)
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
{
GstMatroskaPad *best;
GstEbmlWrite *ebml = mux->ebml_write;
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
GstBuffer *buf, *hdr;
guint64 cluster, blockgroup;
gboolean write_duration;
guint16 relative_timestamp;
guint64 block_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
/* which stream to write from? */
best = gst_matroska_mux_best_pad (mux);
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 there is no best pad, we have reached EOS */
if (best == NULL) {
GST_DEBUG_OBJECT (mux, "No best pad finishing...");
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_matroska_mux_finish (mux);
gst_pad_push_event (mux->srcpad, gst_event_new_eos ());
return GST_FLOW_WRONG_STATE;
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_OBJECT (best->collect.pad, "best 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
/* write data */
buf = best->buffer;
best->buffer = 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
/* set the timestamp for outgoing buffers */
ebml->timestamp = GST_BUFFER_TIMESTAMP (buf);
if (mux->cluster) {
/* start a new cluster every two seconds */
if (mux->cluster_time + GST_SECOND * 2 < GST_BUFFER_TIMESTAMP (buf)) {
GstMatroskaMetaSeekIndex *idx;
gst_ebml_write_master_finish (ebml, mux->cluster);
mux->cluster_pos = ebml->pos;
mux->cluster =
gst_ebml_write_master_start (ebml, GST_MATROSKA_ID_CLUSTER);
gst_ebml_write_uint (ebml, GST_MATROSKA_ID_CLUSTERTIMECODE,
GST_BUFFER_TIMESTAMP (buf) / mux->time_scale);
mux->cluster_time = GST_BUFFER_TIMESTAMP (buf);
if (mux->num_meta_indexes % 32 == 0) {
mux->meta_index = g_renew (GstMatroskaMetaSeekIndex, mux->meta_index,
mux->num_meta_indexes + 32);
}
idx = &mux->meta_index[mux->num_meta_indexes++];
idx->id = GST_MATROSKA_ID_CLUSTER;
idx->pos = mux->cluster_pos;
}
} else {
/* first cluster */
GstMatroskaMetaSeekIndex *idx;
mux->cluster_pos = ebml->pos;
mux->cluster = gst_ebml_write_master_start (ebml, GST_MATROSKA_ID_CLUSTER);
gst_ebml_write_uint (ebml, GST_MATROSKA_ID_CLUSTERTIMECODE,
GST_BUFFER_TIMESTAMP (buf) / mux->time_scale);
mux->cluster_time = GST_BUFFER_TIMESTAMP (buf);
if (mux->num_meta_indexes % 32 == 0) {
mux->meta_index = g_renew (GstMatroskaMetaSeekIndex, mux->meta_index,
mux->num_meta_indexes + 32);
}
idx = &mux->meta_index[mux->num_meta_indexes++];
idx->id = GST_MATROSKA_ID_CLUSTER;
idx->pos = mux->cluster_pos;
}
cluster = mux->cluster;
/* update duration of this track */
if (GST_BUFFER_DURATION_IS_VALID (buf))
best->duration += GST_BUFFER_DURATION (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
/* We currently write an index entry for each keyframe in a
* video track or one entry for each cluster in an audio track
* for audio only files. This can be largely improved, such as doing
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
* one for each keyframe or each second (for all-keyframe
* streams), only the *first* video track. But that'll come later... */
if (best->track->type == GST_MATROSKA_TRACK_TYPE_VIDEO &&
!GST_BUFFER_FLAG_IS_SET (buf, GST_BUFFER_FLAG_DELTA_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
GstMatroskaIndex *idx;
if (mux->num_indexes % 32 == 0) {
mux->index = g_renew (GstMatroskaIndex, mux->index,
mux->num_indexes + 32);
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
}
idx = &mux->index[mux->num_indexes++];
idx->pos = mux->cluster_pos;
idx->time = GST_BUFFER_TIMESTAMP (buf);
idx->track = best->track->num;
} else if ((best->track->type == GST_MATROSKA_TRACK_TYPE_AUDIO) &&
(mux->num_streams == 1)) {
GstMatroskaIndex *idx;
if (mux->num_indexes % 32 == 0) {
mux->index = g_renew (GstMatroskaIndex, mux->index,
mux->num_indexes + 32);
}
idx = &mux->index[mux->num_indexes++];
idx->pos = mux->cluster_pos;
idx->time = GST_BUFFER_TIMESTAMP (buf);
idx->track = best->track->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
}
/* Check if the duration differs from the default duration. */
write_duration = FALSE;
block_duration = GST_BUFFER_DURATION (buf);
if (GST_BUFFER_DURATION_IS_VALID (buf)) {
if (block_duration != best->track->default_duration) {
write_duration = TRUE;
}
}
/* write the block, for matroska v2 use SimpleBlock if possible
* one slice (*breath*).
* FIXME: lacing, etc. */
relative_timestamp =
(GST_BUFFER_TIMESTAMP (buf) - mux->cluster_time) / mux->time_scale;
if (mux->matroska_version > 1 && !write_duration) {
int flags =
GST_BUFFER_FLAG_IS_SET (buf, GST_BUFFER_FLAG_DELTA_UNIT) ? 0 : 0x80;
hdr =
gst_matroska_mux_create_buffer_header (best->track, relative_timestamp,
flags);
gst_ebml_write_buffer_header (ebml, GST_MATROSKA_ID_SIMPLEBLOCK,
GST_BUFFER_SIZE (buf) + GST_BUFFER_SIZE (hdr));
gst_ebml_write_buffer (ebml, hdr);
gst_ebml_write_buffer (ebml, buf);
return gst_ebml_last_write_result (ebml);
} else {
blockgroup = gst_ebml_write_master_start (ebml, GST_MATROSKA_ID_BLOCKGROUP);
hdr =
gst_matroska_mux_create_buffer_header (best->track, relative_timestamp,
0);
gst_ebml_write_buffer_header (ebml, GST_MATROSKA_ID_BLOCK,
GST_BUFFER_SIZE (buf) + GST_BUFFER_SIZE (hdr));
gst_ebml_write_buffer (ebml, hdr);
gst_ebml_write_buffer (ebml, buf);
if (write_duration) {
gst_ebml_write_uint (ebml, GST_MATROSKA_ID_BLOCKDURATION,
block_duration / mux->time_scale);
}
gst_ebml_write_master_finish (ebml, blockgroup);
return gst_ebml_last_write_result (ebml);
}
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_matroska_mux_collected:
* @pads: #GstCollectPads
* @uuser_data: #GstMatroskaMux
*
* Collectpads callback.
*
* Returns: #GstFlowReturn
*/
static GstFlowReturn
gst_matroska_mux_collected (GstCollectPads * pads, gpointer user_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
{
GstMatroskaMux *mux = GST_MATROSKA_MUX (user_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
GST_DEBUG_OBJECT (mux, "Collected pads");
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
/* start with a header */
if (mux->state == GST_MATROSKA_MUX_STATE_START) {
if (mux->collect->data == NULL) {
GST_ELEMENT_ERROR (mux, STREAM, MUX, (NULL),
("No input streams configured"));
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
mux->state = GST_MATROSKA_MUX_STATE_HEADER;
gst_matroska_mux_start (mux);
mux->state = GST_MATROSKA_MUX_STATE_DATA;
}
/* do one single buffer */
return gst_matroska_mux_write_data (mux);
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_matroska_mux_change_state:
* @element: #GstMatroskaMux
* @transition: State change transition.
*
* Change the muxer state.
*
* Returns: #GstStateChangeReturn
*/
static GstStateChangeReturn
gst_matroska_mux_change_state (GstElement * element, GstStateChange transition)
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
{
GstStateChangeReturn ret;
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
GstMatroskaMux *mux = GST_MATROSKA_MUX (element);
switch (transition) {
case GST_STATE_CHANGE_NULL_TO_READY:
break;
case GST_STATE_CHANGE_READY_TO_PAUSED:
gst_collect_pads_start (mux->collect);
break;
case GST_STATE_CHANGE_PAUSED_TO_PLAYING:
break;
default:
break;
}
ret = GST_ELEMENT_CLASS (parent_class)->change_state (element, transition);
switch (transition) {
case GST_STATE_CHANGE_PLAYING_TO_PAUSED:
break;
case GST_STATE_CHANGE_PAUSED_TO_READY:
gst_collect_pads_stop (mux->collect);
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_matroska_mux_reset (GST_ELEMENT (mux));
break;
case GST_STATE_CHANGE_READY_TO_NULL:
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:
break;
}
return ret;
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_matroska_mux_set_property (GObject * object,
guint prop_id, const GValue * value, GParamSpec * pspec)
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
{
GstMatroskaMux *mux;
g_return_if_fail (GST_IS_MATROSKA_MUX (object));
mux = GST_MATROSKA_MUX (object);
switch (prop_id) {
case ARG_WRITING_APP:
if (!g_value_get_string (value)) {
GST_WARNING_OBJECT (mux, "writing-app property can not be NULL");
break;
}
g_free (mux->writing_app);
mux->writing_app = g_strdup (g_value_get_string (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
break;
case ARG_MATROSKA_VERSION:
mux->matroska_version = g_value_get_int (value);
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:
G_OBJECT_WARN_INVALID_PROPERTY_ID (object, prop_id, pspec);
break;
}
}
static void
gst_matroska_mux_get_property (GObject * object,
guint prop_id, GValue * value, GParamSpec * pspec)
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
{
GstMatroskaMux *mux;
g_return_if_fail (GST_IS_MATROSKA_MUX (object));
mux = GST_MATROSKA_MUX (object);
switch (prop_id) {
case ARG_WRITING_APP:
g_value_set_string (value, mux->writing_app);
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
break;
case ARG_MATROSKA_VERSION:
g_value_set_int (value, mux->matroska_version);
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:
G_OBJECT_WARN_INVALID_PROPERTY_ID (object, prop_id, pspec);
break;
}
}
gboolean
gst_matroska_mux_plugin_init (GstPlugin * plugin)
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 gst_element_register (plugin, "matroskamux",
GST_RANK_NONE, GST_TYPE_MATROSKA_MUX);
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
}