gstreamer/gst/matroska/ebml-write.c

587 lines
14 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 EBML I/O
* (c) 2003 Ronald Bultje <rbultje@ronald.bitfreak.net>
*
* ebml-write.c: write EBML data to file/stream
*
* 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 <string.h>
#include "ebml-write.h"
#include "ebml-ids.h"
enum {
/* FILL ME */
LAST_SIGNAL
};
static void gst_ebml_write_class_init (GstEbmlWriteClass *klass);
static void gst_ebml_write_init (GstEbmlWrite *ebml);
static GstElementStateReturn
gst_ebml_write_change_state (GstElement *element);
static GstElementClass *parent_class = NULL;
GType
gst_ebml_write_get_type (void)
{
static GType gst_ebml_write_type = 0;
if (!gst_ebml_write_type) {
static const GTypeInfo gst_ebml_write_info = {
sizeof (GstEbmlWriteClass),
NULL,
NULL,
(GClassInitFunc) gst_ebml_write_class_init,
NULL,
NULL,
sizeof (GstEbmlWrite),
0,
(GInstanceInitFunc) gst_ebml_write_init,
};
gst_ebml_write_type =
g_type_register_static (GST_TYPE_ELEMENT, "GstEbmlWrite",
&gst_ebml_write_info, 0);
}
return gst_ebml_write_type;
}
static void
gst_ebml_write_class_init (GstEbmlWriteClass *klass)
{
GstElementClass *gstelement_class = (GstElementClass *) klass;
parent_class = g_type_class_ref (GST_TYPE_ELEMENT);
gstelement_class->change_state = gst_ebml_write_change_state;
}
static void
gst_ebml_write_init (GstEbmlWrite *ebml)
{
ebml->srcpad = NULL;
ebml->pos = 0;
ebml->cache = NULL;
}
static GstElementStateReturn
gst_ebml_write_change_state (GstElement *element)
{
GstEbmlWrite *ebml = GST_EBML_WRITE (element);
switch (GST_STATE_TRANSITION (element)) {
case GST_STATE_PAUSED_TO_READY:
ebml->pos = 0;
break;
default:
break;
}
if (GST_ELEMENT_CLASS (parent_class)->change_state)
return GST_ELEMENT_CLASS (parent_class)->change_state (element);
return GST_STATE_SUCCESS;
}
/*
* Caching.
*
* The idea is that you use this for writing a lot
* of small elements. This will just "queue" all of
* them and they'll be pushed to the next element all
* at once. This saves memory and time for buffer
* allocation and init, and it looks better.
*/
void
gst_ebml_write_set_cache (GstEbmlWrite *ebml,
guint size)
{
/* This is currently broken. I don't know why yet. */
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b... Original commit message from CVS: Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first broad testing. I've revived rifflib. Rifflib used to be a bytestream-for-riff, which just dup'ed bytestream. I've rewritten rifflib to be a modern riff- chunk parser that uses bytestream fully, plus adds some extra functions so that riff file parsing becomes extremely easy. It also contains some small usability functions for strh/strf and metadata parsing. Note that it doesn't use the new tagging yet, that's a TODO. Avidemux has been rewritten to use this. I think we all agreed that avidemux was pretty much a big mess, which is because it used all sort of bytestream magic all around the place. It was just ugly. This is a lot nicer, very complete and safe. I think this is far more robust than what the old avidemux could ever have been. Of course, it might contain bugs, please let me know. EBML writing has also been implemented. This is useful for matroska. I'm intending to modify avidemux (with a riffwriter) similarly. Maybe I'll change wavparse/-enc too to use rifflib. Lastly, several plugins have been modified to use rifflib's fourcc parsing instead of their own. this puts fourcc parsing in one central place, which should make it a lot simpler to add 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;
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b... Original commit message from CVS: Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first broad testing. I've revived rifflib. Rifflib used to be a bytestream-for-riff, which just dup'ed bytestream. I've rewritten rifflib to be a modern riff- chunk parser that uses bytestream fully, plus adds some extra functions so that riff file parsing becomes extremely easy. It also contains some small usability functions for strh/strf and metadata parsing. Note that it doesn't use the new tagging yet, that's a TODO. Avidemux has been rewritten to use this. I think we all agreed that avidemux was pretty much a big mess, which is because it used all sort of bytestream magic all around the place. It was just ugly. This is a lot nicer, very complete and safe. I think this is far more robust than what the old avidemux could ever have been. Of course, it might contain bugs, please let me know. EBML writing has also been implemented. This is useful for matroska. I'm intending to modify avidemux (with a riffwriter) similarly. Maybe I'll change wavparse/-enc too to use rifflib. Lastly, several plugins have been modified to use rifflib's fourcc parsing instead of their own. this puts fourcc parsing in one central place, which should make it a lot simpler to add 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_return_if_fail (ebml->cache == NULL);
ebml->cache = gst_buffer_new_and_alloc (size);
GST_BUFFER_SIZE (ebml->cache) = 0;
GST_BUFFER_OFFSET (ebml->cache) = ebml->pos;
ebml->handled = 0;
}
void
gst_ebml_write_flush_cache (GstEbmlWrite *ebml)
{
if (!ebml->cache)
return;
/* this is very important. It may fail, in which case the client
* programmer didn't use the cache somewhere. That's fatal. */
g_assert (ebml->handled == GST_BUFFER_SIZE (ebml->cache));
g_assert (GST_BUFFER_SIZE (ebml->cache) +
GST_BUFFER_OFFSET (ebml->cache) == ebml->pos);
gst_pad_push (ebml->srcpad, GST_DATA (ebml->cache));
ebml->cache = NULL;
ebml->handled = 0;
}
/*
* One-element buffer, in case of no cache. If there is
* a cache, use that instead.
*/
static GstBuffer *
gst_ebml_write_element_new (GstEbmlWrite *ebml,
guint size)
{
/* Create new buffer of size + ID + length */
GstBuffer *buf;
/* length, ID */
size += 12;
/* prefer cache */
if (ebml->cache) {
if (GST_BUFFER_MAXSIZE (ebml->cache) -
GST_BUFFER_SIZE (ebml->cache) < size) {
GST_LOG ("Cache available, but too small. Clearing...");
gst_ebml_write_flush_cache (ebml);
} else {
return ebml->cache;
}
}
/* else, use a one-element buffer. This is slower */
buf = gst_buffer_new_and_alloc (size);
GST_BUFFER_SIZE (buf) = 0;
return buf;
}
/*
* Write element ID into a buffer.
*/
static void
gst_ebml_write_element_id (GstBuffer *buf,
guint32 id)
{
guint8 *data = GST_BUFFER_DATA (buf) + GST_BUFFER_SIZE (buf);
guint bytes = 4, mask = 0x10;
/* get ID length */
while (!(id & (mask << ((bytes - 1) * 8))) && bytes > 0) {
mask <<= 1;
bytes--;
}
/* if invalid ID, use dummy */
if (bytes == 0) {
GST_WARNING ("Invalid ID, voiding");
bytes = 1;
id = GST_EBML_ID_VOID;
}
/* write out, BE */
GST_BUFFER_SIZE (buf) += bytes;
while (bytes--) {
data[bytes] = id & 0xff;
id >>= 8;
}
}
/*
* Write element length into a buffer.
*/
static void
gst_ebml_write_element_size (GstBuffer *buf,
guint64 size)
{
guint8 *data = GST_BUFFER_DATA (buf) + GST_BUFFER_SIZE (buf);
guint bytes = 1, mask = 0x80;
/* how many bytes? */
while ((size >> ((bytes - 1) * 8)) >= mask && bytes <= 8) {
mask >>= 1;
bytes++;
}
/* if invalid size, use max. */
if (bytes > 8) {
GST_WARNING ("Invalid size, maximizing");
mask = 0x01;
bytes = 8;
/* Now here's a real FIXME: we cannot read those yet! */
size = 0x00ffffffffffffffLLU;
}
/* write out, BE, with length size marker */
GST_BUFFER_SIZE (buf) += bytes;
while (bytes-- > 0) {
data[bytes] = size & 0xff;
size >>= 8;
if (!bytes)
*data |= mask;
}
}
/*
* Write element data into a buffer.
*/
static void
gst_ebml_write_element_data (GstBuffer *buf,
guint8 *write,
guint64 length)
{
guint8 *data = GST_BUFFER_DATA (buf) + GST_BUFFER_SIZE (buf);
memcpy (data, write, length);
GST_BUFFER_SIZE (buf) += length;
}
/*
* Write out buffer by moving it to the next element.
*/
static void
gst_ebml_write_element_push (GstEbmlWrite *ebml,
GstBuffer *buf)
{
guint data_size = GST_BUFFER_SIZE (buf) - ebml->handled;
ebml->pos += data_size;
if (buf == ebml->cache) {
ebml->handled += data_size;
}
/* if there's no cache, then don't push it! */
if (ebml->cache) {
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b... Original commit message from CVS: Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first broad testing. I've revived rifflib. Rifflib used to be a bytestream-for-riff, which just dup'ed bytestream. I've rewritten rifflib to be a modern riff- chunk parser that uses bytestream fully, plus adds some extra functions so that riff file parsing becomes extremely easy. It also contains some small usability functions for strh/strf and metadata parsing. Note that it doesn't use the new tagging yet, that's a TODO. Avidemux has been rewritten to use this. I think we all agreed that avidemux was pretty much a big mess, which is because it used all sort of bytestream magic all around the place. It was just ugly. This is a lot nicer, very complete and safe. I think this is far more robust than what the old avidemux could ever have been. Of course, it might contain bugs, please let me know. EBML writing has also been implemented. This is useful for matroska. I'm intending to modify avidemux (with a riffwriter) similarly. Maybe I'll change wavparse/-enc too to use rifflib. Lastly, several plugins have been modified to use rifflib's fourcc parsing instead of their own. this puts fourcc parsing in one central place, which should make it a lot simpler to add 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 (buf == ebml->cache);
return;
}
gst_pad_push (ebml->srcpad, GST_DATA (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
}
/*
* Seek.
*/
void
gst_ebml_write_seek (GstEbmlWrite *ebml,
guint64 pos)
{
GstEvent *seek;
/* Cache seeking. A bit dangerous, we assume the client writer
* knows what he's doing... */
if (ebml->cache) {
/* within bounds? */
if (pos >= GST_BUFFER_OFFSET (ebml->cache) &&
pos < GST_BUFFER_OFFSET (ebml->cache) + GST_BUFFER_MAXSIZE (ebml->cache)) {
GST_BUFFER_SIZE (ebml->cache) = pos - GST_BUFFER_OFFSET (ebml->cache);
if (ebml->pos > pos)
ebml->handled -= ebml->pos - pos;
else
ebml->handled += pos - ebml->pos;
ebml->pos = pos;
} else {
GST_LOG ("Seek outside cache range. Clearing...");
gst_ebml_write_flush_cache (ebml);
}
}
seek = gst_event_new_seek (GST_FORMAT_BYTES |
GST_SEEK_METHOD_SET,
pos);
gst_pad_push (ebml->srcpad, GST_DATA (seek));
ebml->pos = pos;
}
/*
* Get no. bytes needed to write a uint.
*/
static guint
gst_ebml_write_get_uint_size (guint64 num)
{
guint size = 1;
/* get size */
while (num >= (1LLU << (size * 8)) && size < 8) {
size++;
}
return size;
}
/*
* Write an uint into a buffer.
*/
static void
gst_ebml_write_set_uint (GstBuffer *buf,
guint64 num,
guint size)
{
guint8 *data;
data = GST_BUFFER_DATA (buf) + GST_BUFFER_SIZE (buf);
GST_BUFFER_SIZE (buf) += size;
while (size-- > 0) {
data[size] = num & 0xff;
num >>= 8;
}
}
/*
* Data type wrappers.
*/
void
gst_ebml_write_uint (GstEbmlWrite *ebml,
guint32 id,
guint64 num)
{
GstBuffer *buf = gst_ebml_write_element_new (ebml, sizeof (num));
guint size = gst_ebml_write_get_uint_size (num);
/* write */
gst_ebml_write_element_id (buf, id);
gst_ebml_write_element_size (buf, size);
gst_ebml_write_set_uint (buf, num, size);
gst_ebml_write_element_push (ebml, buf);
}
void
gst_ebml_write_sint (GstEbmlWrite *ebml,
guint32 id,
gint64 num)
{
GstBuffer *buf = gst_ebml_write_element_new (ebml, sizeof (num));
/* if the signed number is on the edge of a extra-byte,
* then we'll fall over when detecting it. Example: if I
* have a number (-)0x8000 (G_MINSHORT), then my abs()<<1
* will be 0x10000; this is G_MAXUSHORT+1! So: if (<0) -1. */
guint64 unum = (num < 0 ? (-num - 1) << 1 : num << 1);
guint size = gst_ebml_write_get_uint_size (unum);
/* make unsigned */
unum = (num < 0 ? -num : num) + (1LLU << ((8 * size) - 1));
/* write */
gst_ebml_write_element_id (buf, id);
gst_ebml_write_element_size (buf, size);
gst_ebml_write_set_uint (buf, unum, size);
gst_ebml_write_element_push (ebml, buf);
}
void
gst_ebml_write_float (GstEbmlWrite *ebml,
guint32 id,
gdouble num)
{
#if (G_BYTE_ORDER == G_LITTLE_ENDIAN)
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b... Original commit message from CVS: Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first broad testing. I've revived rifflib. Rifflib used to be a bytestream-for-riff, which just dup'ed bytestream. I've rewritten rifflib to be a modern riff- chunk parser that uses bytestream fully, plus adds some extra functions so that riff file parsing becomes extremely easy. It also contains some small usability functions for strh/strf and metadata parsing. Note that it doesn't use the new tagging yet, that's a TODO. Avidemux has been rewritten to use this. I think we all agreed that avidemux was pretty much a big mess, which is because it used all sort of bytestream magic all around the place. It was just ugly. This is a lot nicer, very complete and safe. I think this is far more robust than what the old avidemux could ever have been. Of course, it might contain bugs, please let me know. EBML writing has also been implemented. This is useful for matroska. I'm intending to modify avidemux (with a riffwriter) similarly. Maybe I'll change wavparse/-enc too to use rifflib. Lastly, several plugins have been modified to use rifflib's fourcc parsing instead of their own. this puts fourcc parsing in one central place, which should make it a lot simpler to add new fourccs. We might want to move this to its own lib instead of rifflib. Enjoy!
2003-12-07 20:00:41 +00:00
gint n;
#endif
Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first b... Original commit message from CVS: Riff, EBML, fourcc etc. work. Not fully finished, but better than what we used to have and definately worth a first broad testing. I've revived rifflib. Rifflib used to be a bytestream-for-riff, which just dup'ed bytestream. I've rewritten rifflib to be a modern riff- chunk parser that uses bytestream fully, plus adds some extra functions so that riff file parsing becomes extremely easy. It also contains some small usability functions for strh/strf and metadata parsing. Note that it doesn't use the new tagging yet, that's a TODO. Avidemux has been rewritten to use this. I think we all agreed that avidemux was pretty much a big mess, which is because it used all sort of bytestream magic all around the place. It was just ugly. This is a lot nicer, very complete and safe. I think this is far more robust than what the old avidemux could ever have been. Of course, it might contain bugs, please let me know. EBML writing has also been implemented. This is useful for matroska. I'm intending to modify avidemux (with a riffwriter) similarly. Maybe I'll change wavparse/-enc too to use rifflib. Lastly, several plugins have been modified to use rifflib's fourcc parsing instead of their own. this puts fourcc parsing in one central place, which should make it a lot simpler to add new fourccs. We might want to move this to its own lib instead of rifflib. Enjoy!
2003-12-07 20:00:41 +00:00
GstBuffer *buf = gst_ebml_write_element_new (ebml, sizeof (num));
gst_ebml_write_element_id (buf, id);
gst_ebml_write_element_size (buf, 8);
#if (G_BYTE_ORDER == G_LITTLE_ENDIAN)
for (n = 0; n < 8; n++)
GST_BUFFER_DATA (buf)[GST_BUFFER_SIZE (buf)] = ((guint8 *) &num)[7-n];
GST_BUFFER_SIZE (buf) += 8;
#else
gst_ebml_write_element_data (buf, (guint8 *) &num, 8);
#endif
gst_ebml_write_element_push (ebml, buf);
}
void
gst_ebml_write_ascii (GstEbmlWrite *ebml,
guint32 id,
const gchar *str)
{
gint len = strlen (str) + 1; /* add trailing '\0' */
GstBuffer *buf = gst_ebml_write_element_new (ebml, len);
gst_ebml_write_element_id (buf, id);
gst_ebml_write_element_size (buf, len);
gst_ebml_write_element_data (buf, (guint8 *) str, len);
gst_ebml_write_element_push (ebml, buf);
}
void
gst_ebml_write_utf8 (GstEbmlWrite *ebml,
guint32 id,
const gchar *str)
{
gst_ebml_write_ascii (ebml, id, str);
}
void
gst_ebml_write_date (GstEbmlWrite *ebml,
guint32 id,
gint64 date)
{
gst_ebml_write_sint (ebml, id, date);
}
/*
* Master writing is annoying. We use a size marker of
* the max. allowed length, so that we can later fill it
* in validly.
*/
guint64
gst_ebml_write_master_start (GstEbmlWrite *ebml,
guint32 id)
{
guint64 pos = ebml->pos, t;
GstBuffer *buf = gst_ebml_write_element_new (ebml, 0);
t = GST_BUFFER_SIZE (buf);
gst_ebml_write_element_id (buf, id);
pos += GST_BUFFER_SIZE (buf) - t;
gst_ebml_write_element_size (buf, -1);
gst_ebml_write_element_push (ebml, buf);
return pos;
}
void
gst_ebml_write_master_finish (GstEbmlWrite *ebml,
guint64 startpos)
{
guint64 pos = ebml->pos;
GstBuffer *buf;
gst_ebml_write_seek (ebml, startpos);
buf = gst_ebml_write_element_new (ebml, 0);
startpos = GUINT64_TO_BE ((1LLU << 56) | (pos - startpos - 8));
memcpy (GST_BUFFER_DATA (buf) + GST_BUFFER_SIZE (buf),
(guint8 *) &startpos, 8);
GST_BUFFER_SIZE (buf) += 8;
gst_ebml_write_element_push (ebml, buf);
gst_ebml_write_seek (ebml, pos);
}
void
gst_ebml_write_binary (GstEbmlWrite *ebml,
guint32 id,
guint8 *binary,
guint64 length)
{
GstBuffer *buf = gst_ebml_write_element_new (ebml, length);
gst_ebml_write_element_id (buf, id);
gst_ebml_write_element_size (buf, length);
gst_ebml_write_element_data (buf, binary, length);
gst_ebml_write_element_push (ebml, buf);
}
/*
* For things like video frames and audio samples,
* you want to use this function, as it doesn't have
* the overhead of memcpy() that other functions
* such as write_binary() do have.
*/
void
gst_ebml_write_buffer_header (GstEbmlWrite *ebml,
guint32 id,
guint64 length)
{
GstBuffer *buf = gst_ebml_write_element_new (ebml, 0);
gst_ebml_write_element_id (buf, id);
gst_ebml_write_element_size (buf, length);
gst_ebml_write_element_push (ebml, buf);
}
void
gst_ebml_write_buffer (GstEbmlWrite *ebml,
GstBuffer *data)
{
gst_ebml_write_element_push (ebml, data);
}
/*
* When replacing a uint, we assume that it is *always*
* 8-byte, since that's the safest guess we can do. This
* is just for simplicity.
*
* FIXME: this function needs to be replaced with something
* proper. This is a crude hack.
*/
void
gst_ebml_replace_uint (GstEbmlWrite *ebml,
guint64 pos,
guint64 num)
{
guint64 oldpos = ebml->pos;
GstBuffer *buf = gst_buffer_new_and_alloc (8);
gst_ebml_write_seek (ebml, pos);
GST_BUFFER_SIZE (buf) = 0;
gst_ebml_write_set_uint (buf, num, 8);
gst_ebml_write_element_push (ebml, buf);
gst_ebml_write_seek (ebml, oldpos);
}
/*
* Write EBML header.
*/
void
gst_ebml_write_header (GstEbmlWrite *ebml,
gchar *doctype,
guint version)
{
guint64 pos;
/* write the basic EBML header */
gst_ebml_write_set_cache (ebml, 0x40);
pos = gst_ebml_write_master_start (ebml, GST_EBML_ID_HEADER);
#if (GST_EBML_VERSION != 1)
gst_ebml_write_uint (ebml, GST_EBML_ID_EBMLVERSION, GST_EBML_VERSION);
gst_ebml_write_uint (ebml, GST_EBML_ID_EBMLREADVERSION, GST_EBML_VERSION);
#endif
#if 0
/* we don't write these until they're "non-default" (never!) */
gst_ebml_write_uint (ebml, GST_EBML_ID_EBMLMAXIDLENGTH, sizeof (guint32));
gst_ebml_write_uint (ebml, GST_EBML_ID_EBMLMAXSIZELENGTH, sizeof (guint64));
#endif
gst_ebml_write_ascii (ebml, GST_EBML_ID_DOCTYPE, doctype);
gst_ebml_write_uint (ebml, GST_EBML_ID_DOCTYPEVERSION, version);
gst_ebml_write_uint (ebml, GST_EBML_ID_DOCTYPEREADVERSION, version);
gst_ebml_write_master_finish (ebml, pos);
gst_ebml_write_flush_cache (ebml);
}