id3v2: restructure compressed and unsync code

This should fix the interaction between the 2.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

ffv1dec: remove incorrect assert()

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

rmdec: fix av_log level and context

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

truemotion2: Fix av_log level and context

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

mpeg12: fix av_log level and context

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

dcadec: fix av_log level

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

avidec: try to support oddly muxed MMES stream

Fixes ticket1804

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

vqavideo: Reinitialise the actually used partial codebook bytestream-reader.

Fixes ticket #1942.

xmv: signal EOF

Fixes #1946.

Signed-off-by: Paul B Mahol <onemda@gmail.com>

movenc: fix assert failure caused by rounding.

Fixes Ticket1820

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

ADPCM AFC decoder

Signed-off-by: Paul B Mahol <onemda@gmail.com>

id3v2: print ver/flags/len on debug level.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

tests: prefer ffprobe -show_entries to -show_format_entry

The latter option was deprecated.

Merge remote-tracking branch 'qatar/master'

* qatar/master:
  id3v2: fix reading unsynchronized frames.
  cdgraphics: fix incorrect vertical offset mask in cdg_scroll()
  apetag: fix error handling in ff_ape_parse_tag()

Conflicts:
libavformat/id3v2.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>

ffprobe: implement -sections option

ffprobe: implement subsection field selection through the -show_entries option

id3v2: fix reading unsynchronized frames.

Current code would incorrectly process e.g. 'ff 00 ff 00 ff' to
'ff ff ff', while it should be 'ff ff 00 ff'.

Fixes Bug 395.

CC: libav-stable@libav.org

lavfi/avfiltergraph: improve error message in graph_check_validity()

Fix trac ticket #1944.

lavfi/drawtext: add support for printing frame numbers

Fix trac ticket #1949.

ffmpeg: add setting of field_order flag

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

frwu: do not set interlacing flags.

Our fate frwu sample is definitely not interlaced.
If someone has a interlaced frwu file, please contact us.

Without this change the field order patch would cause
incorrect interlacing information to be written when
using frwu input.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

brender: move declaration out of for() to make old compilers happy.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

brender_pix: a new image decoder

PIX is an image file format that was used by the BRender 3d engine.

Signed-off-by: Aleksi Nurmi <aleksi.nurmi@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Fix the path to the C99-to-C89 Converter & Wrapper to the only existing repository.

Ronald deleted his public c99-to-c89 repository today.

ffmpeg: fix weird reindent

ffmpeg: print debug_ts off and off_time values with timestamp helper routines

mips: Fixing HAVE_MIPSFPU dependencies

Signed-off-by: Bojan Zivkovic <bojan@mips.com>
Reviewed-by: Vitor Sessak <vitor1001@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

use correct doxygen parameter direction syntax

Signed-off-by: Peter Ross <pross@xvid.org>
Reviewed-by: Stefano Sabatini <stefasab@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

ensure comment blocks that contain doxygen commands start with double asterix

Reveiwed-by: Stefano Sabatini <stefasab@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

siff: signal EOF

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

segafilm: signal EOF

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

bethsoftvid: signal EOF

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

bfi: signal EOF

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

yop: fix 10l typo

Signed-off-by: Paul B Mahol <onemda@gmail.com>

imgconvert-test: count the number of unused pixel format values.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

imgconvert-test: skip pix formats without name

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

ffprobe: fix typo in a comment

ffprobe: add "," at the end of enum list

imgconvert-test: test alpha flags

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

pixdesc: fix alpha flags

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

imgconvert: remove avg_bits_per_pixel(), its redundant

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

bink: return AVERROR_EOF upon reaching end of file

Fixes ticket #1935.

Signed-off-by: Peter Ross <pross@xvid.org>
Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

codec_descriptors: fix typo in mpeg 2 video

Found-by: Ramiro
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Merge remote-tracking branch 'qatar/master'

* qatar/master:
  flashsv: Drop unused function and struct parameters
  pcm: fix decoding of pcm_s16le_planar on big-endian

Conflicts:
libavcodec/pcm.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit 'ae3822bca16f1cdb2460a35b16f8ef636a04314e'

* commit 'ae3822bca16f1cdb2460a35b16f8ef636a04314e':
  imgconvert: remove PixFmtInfo

Conflicts:
libavcodec/imgconvert.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>

imgconvert: fix color type for non normal pix_fmts like HW stuff and unused entries.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

avcodec_get_pix_fmt_loss: remove PixFmtInfo use

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

imgconvert: print color type too

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

cdxl: use url_feof()

Signed-off-by: Paul B Mahol <onemda@gmail.com>

imgconvert: remove PixFmtInfo use from avg_bits_per_pixel()

This fixes several inconsistencies with YUV >8bit formats caused
due to missing information in the PixFmtInfo table.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

avcodec_find_best_pix_fmt_of_2: favor formats with fewer components if it does not incur a loss.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

pixdesc: add av_get_padded_bits_per_pixel()

This will be needed to allow removing the PixFmtInfo table

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

imgconvert-test: add avg bits per pixel

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

is_yuv_planar: remove use of PixFmtInfo

This fixes the behavior for a few yuva 16bit formats

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

imgconvert: add self test code

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

rawenc: cosmetics: reindent

Signed-off-by: Paul B Mahol <onemda@gmail.com>

Merge commit 'd1d9efaae6c7e8466b06c30ca21c6b569dd2e480'

* commit 'd1d9efaae6c7e8466b06c30ca21c6b569dd2e480':
  avcodec: split avpicture from imgconvert
  pixdesc: add PIX_FMT_ALPHA flag

Conflicts:
libavcodec/imgconvert.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>

cdgraphics: fix incorrect vertical offset mask in cdg_scroll()

The vertical offset mask 0x07 is suspicious.

    v_off = FFMIN(data[2] & 0x07, CDG_BORDER_HEIGHT - 1);

Note that v_off is up to 11 (CDG_BORDER_HEIGHT - 1), the correct mask
should be 0x0F.

Signed-off-by: Xi Wang <xi.wang@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>

apetag: fix error handling in ff_ape_parse_tag()

The following error handling is broken due to signedness.

    int file_size;
    uint32_t tag_bytes;
    int64_t tag_start;
    ...
    tag_start = file_size - tag_bytes - APE_TAG_FOOTER_BYTES;
    if (tag_start < 0) { ... }

Note that tag_bytes is unsigned, which makes the right-hand side of
`tag_start = ...' unsigned, too.  The 32-bit unsigned value is then
zero-extended to 64 bits.  Therefore, tag_start must be non-negative,
and the check (tag_start < 0) is always false, which breaks the error
handling.  This patch fixes the check.

Signed-off-by: Xi Wang <xi.wang@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>

flashsv: Drop unused function and struct parameters

sierravmd: signal EOF

cafdec: return right code if EOF is reached

Signed-off-by: Paul B Mahol <onemda@gmail.com>

thp: signal EOF

rpl: return AVERROR_EOF instead of -1

lavf/hls: whitespace cosmetics after 23db5418.

lavf: move libmodplug registration with the other ext lib.

Also remove the duplicated #if.

hls: create an AVProgram for each variant

Without the information, application may choose audio from one variant
and video from another variant, which leads to fetch two variants from
network. This enables av_find_best_stream() to find matching audio and
video streams, so that only one variant is fetched from network.

Signed-off-by: LYF <yefei.li@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

lavf/segment: fix value for the M3U8 EXT-X-MEDIA

From the M3U8 specification:
|The EXT-X-MEDIA-SEQUENCE tag indicates the sequence number of the first
|URI that appears in a Playlist file.

Previously it was using the list index number. Also remove now unused
list_count field.

lavf/segment: do not pre-increment segment_idx value

Increment the value just when a new segment with the given index is
started. Simplify logic, improve readability.

lavf/segment: unbreak behavior for segment muxer

The segment muxer presumes the use of AVFMT_GLOBALHEADER, ssegment should
be use in case of formats which requires no global headers.

Regression introduced when merging
0826d8513d14f54da0b519360813327d5836b7fc.

lavf/segment: consistently use "seg" in segment_start()

ffmpeg: rework debugging timestamp logs in process_input()

Print timestamps before and after ffmpeg timestamp postprocessing.
Help to debug how ffmpeg processes the input timestamps.

lavf/srtdec: trim line break event separators from packet.

The muxer add them automatically, so this avoid having a bunch of line
breaks all over the output files. One '\n' is still kept/added because
the lavc subrip decoder seems to have trouble with line ending abruptly
(it doesn't flush correctly the tags). This bug is harmless but should
be fixed; though, this doesn't look like a trivial change. When this bug
gets fixed, we can consider removing the '\n' at the end of the packet.

The 2048B buffer limit was also removed while moving to AVBPrint API.
Note that this doesn't really matter since the decoder is limited as
well.

Add -skip_initial_bytes option.

Fixes ticket #1909.

Correctly signal EOF when demuxing caf files.

swr: include stdint.h instead of inttypes.h.

stdint.h is for the [u]int*_t types, which is the only thing we need for
the prototypes. inttypes.h includes stdint.h and defines more thing we
don't need here.

Bump micro in case a user app was relying on this include for its own
code.

rc: fix 10l typo in  rc_max_available_vbv_use calculation

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

fate: increase pcm coverage

Signed-off-by: Paul B Mahol <onemda@gmail.com>

swr: fix a few typo in the public header.

swr-doxy: elaborate on swr_get_delay() and the timebase

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

swr: general doxy text about swr and example code.

Based on doxy from avr
Reviewed-by: Clément Bœsch
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

lavf/segment: do not copy codec_tag when not available

Some muxers do not allow stream if codec_tag is incompatible.

Sometimes the passed input codec's codec_tag is not compatible with the
output muxer.

Because the codec_tag field of the segment muxer cannot be set, ffmpeg.c
doesn't know how to handle these cases.

Signed-off-by: ChanMin Kim <kcm1700@gmail.com>
Signed-off-by: Stefano Sabatini <stefasab@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

the hyperlink to the Developer's Certificate of Origin no longer works; use Linux kernel hyperlink

Signed-off-by: Peter Ross <pross@xvid.org>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

add some planar PCM encoders

Signed-off-by: Paul B Mahol <onemda@gmail.com>

pcmdec: use memcpy() when possible for planar decoders

This is faster (at least with clang).

Signed-off-by: Paul B Mahol <onemda@gmail.com>

pcm: fix decoding of pcm_s16le_planar on big-endian

The sample count is decremented by the DECODE() macro and needs to be reset
in each loop iteration. Also, DECODE() increments the src pointer so that does
not need to be done separately.

oggdec: fix memleak on header parsing failure

Fixes Ticket1931

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

doc/platform: replace link to site arrozcru.org with link to ffmpeg.zeranoe.com

ffmpeg.arrozcru.org is currently unmaintained, and it redirects to
ffmpeg.zeranoe.com.

Merge remote-tracking branch 'qatar/master'

* qatar/master:
  FATE: add 24-bit ALAC tests
  FATE: rename ALAC tests from alac-* to alac-16-*

Merged-by: Michael Niedermayer <michaelni@gmx.at>

pcmenc: use ENCODE macro for pcm_s8

Signed-off-by: Paul B Mahol <onemda@gmail.com>

pcm: revert from libavs planar code to durandals.

The libav code does not work on big endian.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '7c278d2ae410a64bdd89f1777026b4b963c30a1a'

* commit '7c278d2ae410a64bdd89f1777026b4b963c30a1a':
  alacenc: support 24-bit encoding
  pcmdec: use planar sample format for pcm_s16le_planar
  vorbisdec: use float planar sample format

Conflicts:
libavcodec/pcm.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>

imgconvert: remove PixFmtInfo

It is pleonastic and was used in stale functions pending replacement.

avcodec: split avpicture from imgconvert

All the non deprecated functions are in avpicture.c now.

pixdesc: add PIX_FMT_ALPHA flag

8svx: move pcm_s8_planar decoder to pcm.c

Removes limitation of max 2 channels for pcm_s8_planar decoder by moving
it to more natural place.

AV_CODEC_ID_8SVX_RAW is not used by anything anymore and is going to be
removed.

Signed-off-by: Paul B Mahol <onemda@gmail.com>

4xm: return error code if decode_init() failed

Signed-off-by: Paul B Mahol <onemda@gmail.com>

FATE: add 24-bit ALAC tests

alacenc: support 24-bit encoding

FATE: rename ALAC tests from alac-* to alac-16-*

This will distinguish ALAC 16-bit tests from ALAC tests with other bit
depths.