The handling of audio stream numbering was handled in the stream
selection property was a total mess. The most important issue was
confusion between values used as index for demuxer->audio_streams[]
array (consistently stored in demuxer->audio->id) and values stored
in sh_audio->aid and used as "-aid N" option values. Now demuxer audio
switch control functions and demuxer_switch_audio() are supposed to
return the new value for the "-aid" option (internal MPEG demuxers
still don't; the demuxer requirement could perhaps be dropped as it
can be easily calculated afterwards). That is also the value
returned for the "switch_audio" property. The main changes are:
- Make command.c mp_property_audio() consistently use and return the
"-aid" values. Before it used that as input but the array index as
output, with extra mess related to demuxer_switch_audio() return
value. Don't modify the audio_id option field any more.
- Make demuxer_switch_audio() always return "-aid" values (like it
takes as input). There are two changes for this: picking this
return value in case the demuxer doesn't support switching, and
overriding demuxer return value (for internal MPEG demuxers).
- Make demux_lavf return "-aid" values from DEMUXER_CTRL_SWITCH_AUDIO
code. This isn't actually necessary because of the override part
above.
Here's some history of the relevant behavior that I looked up:
* For most demuxers array index and "-aid" values are the same. At
least demux_mkv, (some of?) the internal MPEG demuxers and demux_ogg
have differed for a long time. demux_ogg doesn't matter because it
doesn't support stream switching.
* Old code seemed to assume that demuxer_switch_audio() return value was
array index, but this wasn't true at least for demux_mkv.
* In svn r19951 reimar mostly removed use of the return value.
* In r20162 ptt added mp_property_audio(). This set the global
audio_id variable (-aid option value) to the return value of
demuxer_switch_audio() and treated the global as the persistent
value of the property, apparently assuming that it would be set to
the "-aid" value, not array index. This was false for internal
MPEG.
* In r30124 reimar changed the property code so that even though it
still modified the option value it didn't use that as the value of
the property any more; instead it incorrectly used the array index.
This meant that for demux_mkv the return value didn't match -aid any
more (though input still did, so setting the property and querying
it didn't match as they used different value systems).
* In r31129 aurel made demux_lavf changes that resulted in its -aid
and array index values no longer matching either. He didn't change
the return value from audio switch when changing -aid, so it now
matched array index only. The latter part didn't cause additional
problems from r20162 though because either choice would have been
broken anyway after r30124 as long as they weren't the same value.
The option field corresponding to -slang is now called "sub_lang"
instead of the old misleading global name "dvdsub_lang". The code
handling -slang in subreader.c looks rather broken; disable it instead
of converting it to use the option field.
Move "struct bstr" definition from ebml.h to its own header and add
some utility functions/macros. Change length field type from int to
size_t and adjust using code accordingly.
Partially based on a patch from Anton Khirnov.
Nowadays FFmpeg is faster than liba52 and external liba52 is well supported.
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@31147 b3059339-0415-0410-9bf9-f77b7e298cf2
libmpdemux/muxer_mpeg.c:2121: warning: implicit declaration of function 'a52_syncinfo'
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@31146 b3059339-0415-0410-9bf9-f77b7e298cf2
aid and vid are now 0-based, instead of being a globally unique id.
This matches the way sid is handled and the way other demuxers manage aid.
As a side effect, it slightly simplifies demux_lavf.
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@31129 b3059339-0415-0410-9bf9-f77b7e298cf2
Playback of nonseekable y4m streams was broken by a change merged from
svn, r30970 which added support for multiple yuv4mpeg files
concatenated together. The change included a stream_skip(stream, -1)
call which only works for seekable files. Work around this by
disabling the concatenated-file check for nonseekable streams. This
means concatenated files are only supported if the stream is seekable,
but this is an improvement compared to the svn commit which caused a
regression making _all_ files fail in the nonseekable case.
It would be reasonably easy to make the concatenation support work in
all cases, either by modifying demux_y4m or (probably better) adding a
general stream API to allow pushing back the read byte. However for
now I'm only fixing the regression with the above trivial change.
Before "-chapter 1" did nothing even if the first chapter didn't start
at the beginning of file. Fix it.
Before all chapter property commands (including chapter seek keys)
failed if the current playback position was before the start of the
first chapter. Now they'll work. Relative chapter seeks will go to the
first chapter (even if that's in the wrong direction for backward
seeks).
Use lavf's flv demuxer for rtmp/rtmps/... stream types. Letting
generic format probing handle this could work, but with the current
probing implementation it'd at least depend on not-really-guaranteed
details of the stream layer (probing different formats and then
decoding depends on seeking back in between; rtmp streams don't
support such seeking directly so would need to rely on details of
caching behavior).
Only try to use the dvd/dvdnav stream seek hack with those stream
types. Generally demuxers can not be expected to cope with the stream
suddenly seeking under them. In principle it would be more correct to
make the test demuxer-based (instead of assuming that using stream
seeks in this manner is OK with whatever demuxer that will be used
with these streams), but that'd be more work.
Move code for resetting decoders after seeks, chapter seeks and angle
changes out of demuxer.c. This functionality belongs on a higher
level; the demux layer can't always know what kind of reinitialization
is required.
to restart without seeking even without cache for easily detectable formats.
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@30946 b3059339-0415-0410-9bf9-f77b7e298cf2
Convert demux_mkv_decode() to allocate possible new storage with
talloc and fix a talloc/malloc conflict in demux_mkv_open_sub() that
broke decoding of files which had a subtitle track with compressed
private data.
Fixes elementary streams with -vc ffodivxvdpau and the native demuxer.
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@30904 b3059339-0415-0410-9bf9-f77b7e298cf2
FFmpeg increased the amount of padding that must be readable beyond
input buffers without SIGSEGV from 8 to 64, and the MPlayer value must
be changed accordingly.
Tweak some code parts that used to rely on string literals from
translation macros being concatenated with other adjacent string
literals. Break up the resulting string into independently translated
parts, so that the existing translations for those parts can still be
used.
For some reason commit e306174952, which
replaced translation macro names with the corresponding English
strings, also collapsed multiple consecutive space characters into
one. Change most of these back. In a couple of cases the amount of
whitespace is important for alignment, and for the rest it at least
keeps the strings closer to the existing translations.
These functions return void*, which is compatible with any pointer,
so there is no need for casts.
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@30744 b3059339-0415-0410-9bf9-f77b7e298cf2
-identify still does not show the right values though.
Fixes bug #1636.
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@30681 b3059339-0415-0410-9bf9-f77b7e298cf2
format autodetection with -nocache and non-seekable streams.
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@30668 b3059339-0415-0410-9bf9-f77b7e298cf2
This ensures that function declarations in both files always match.
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@30596 b3059339-0415-0410-9bf9-f77b7e298cf2
This will read progressively more data if we still did not detect
the format. The lavfpref demuxer is unaffected to avoid hanging
for a long time in case of a slow network stream that some native
demuxer may be able to handle.
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@30547 b3059339-0415-0410-9bf9-f77b7e298cf2
Add support for compression algorithm 3 (header stripping). Rewrite
some of the code related to handling manyfold compression, it was just
completely broken (I don't have samples to test whether it actually
works now).
The decompression step wasn't run at all for subtitle types other than
vobsub. Fix that. Remove a "!mkv_d->v_skip_to_keyframe" test from the
subtitle handling - for properly timed subtitles unnecessary packets
do little harm, and the subtitles could stay visible.
Allow decoding a 0-sized buffer with zlib algorithm to produce 0-sized
output. Fixes spurious errors reported with subtitle tracks marked to
use compression for track private data without having any such data.
Change the demuxer_add_attachment() and demuxer_add_chapter()
functions to take a length argument for various name strings, so those
strings do not need to be 0-terminated. This will make it easier to
directly pass demuxed data without first making a copy just to add
0-termination. Also allocate the struct demuxer data structures for
attachments and chapters with talloc.
The main demuxing code signaled EOF and stopped playback if it hit a
top-level element other than Cluster. There are files with other
elements between Cluster ones, at least repeated copies of Track
headers. Change the code to skip any non-Cluster element and only stop
searching on real file EOF.
Rewrite Cues parsing code using the new EBML parser. The new version
fixes a hang in some cases of incomplete files and supports a cuepoint
specifying multiple tracks per timecode (the previous code added an
index entry for the track mentioned last only).
Restructure the code reading toplevel header elements and rewrite the
SeekHead parsing code using the new EBML parser. Now every type of
header element is read anywhere in the file if there's a SeekHead
entry pointing to it. The new SeekHead parsing code has more
diagnostic output in case of errors.
Add a new EBML parser implementation that should allow significant
improvements to the Matroska demuxer. The new parsing code is not
actually used yet by the demuxer. The only changes to existing code in
this commit are to generate the MATROSKA_ID_* / EBML_ID_* macro
definitions from the new implementation and to rename some of them
(the new implementation uses names matching the official Matroska spec).
The main parser implementation is added in ebml.c. There are two new
generated files, ebml_defs.c and ebml_types.h, that contain
definitions of EBML elements. Those are generated by the new script
TOOLS/matroska.py. There's a new Makefile target "generated_ebml" that
run the script to refresh the content of the generated files.
the ASF demuxer (seek seems to end up right after the keyframe?) and seem to have
no purpose anyway.
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@30438 b3059339-0415-0410-9bf9-f77b7e298cf2
This is necessary to use the ffmp2 decoder with dvr-ms files.
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@30437 b3059339-0415-0410-9bf9-f77b7e298cf2
based only on substream id.
Works with all available DTS and TrueHD samples available (2 each).
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@30429 b3059339-0415-0410-9bf9-f77b7e298cf2
AC3 is still broken due to the libavcodec parser being broken.
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@30421 b3059339-0415-0410-9bf9-f77b7e298cf2
scattered all over the place with half of it forgotten in some places.
git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@30420 b3059339-0415-0410-9bf9-f77b7e298cf2