You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

1807 lines
63 KiB

/*
* MOV demuxer
* Copyright (c) 2001 Fabrice Bellard.
*
* This file is part of FFmpeg.
*
* FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* FFmpeg 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
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
#include <limits.h>
//#define DEBUG
#include "avformat.h"
#include "riff.h"
#include "isom.h"
#include "dv.h"
#ifdef CONFIG_ZLIB
#include <zlib.h>
#endif
/*
* First version by Francois Revol revol@free.fr
* Seek function by Gael Chardon gael.dev@4now.net
*
* Features and limitations:
* - reads most of the QT files I have (at least the structure),
* the exceptions are .mov with zlib compressed headers ('cmov' section). It shouldn't be hard to implement.
* FIXED, Francois Revol, 07/17/2002
* - ffmpeg has nearly none of the usual QuickTime codecs,
* although I succesfully dumped raw and mp3 audio tracks off .mov files.
* Sample QuickTime files with mp3 audio can be found at: http://www.3ivx.com/showcase.html
* - .mp4 parsing is still hazardous, although the format really is QuickTime with some minor changes
* (to make .mov parser crash maybe ?), despite what they say in the MPEG FAQ at
* http://mpeg.telecomitalialab.com/faq.htm
* - the code is quite ugly... maybe I won't do it recursive next time :-)
* - seek is not supported with files that contain edit list
*
* Funny I didn't know about http://sourceforge.net/projects/qt-ffmpeg/
* when coding this :) (it's a writer anyway)
*
* Reference documents:
* http://www.geocities.com/xhelmboyx/quicktime/formats/qtm-layout.txt
* Apple:
* http://developer.apple.com/documentation/QuickTime/QTFF/
* http://developer.apple.com/documentation/QuickTime/QTFF/qtff.pdf
* QuickTime is a trademark of Apple (AFAIK :))
*/
#include "qtpalette.h"
#undef NDEBUG
#include <assert.h>
static const AVCodecTag mov_video_tags[] = {
/* { CODEC_ID_, MKTAG('c', 'v', 'i', 'd') }, *//* Cinepak */
/* { CODEC_ID_H263, MKTAG('r', 'a', 'w', ' ') }, *//* Uncompressed RGB */
/* { CODEC_ID_H263, MKTAG('Y', 'u', 'v', '2') }, *//* Uncompressed YUV422 */
/* { CODEC_ID_RAWVIDEO, MKTAG('A', 'V', 'U', 'I') }, *//* YUV with alpha-channel (AVID Uncompressed) */
/* Graphics */
/* Animation */
/* Apple video */
/* Kodak Photo CD */
{ CODEC_ID_MJPEG, MKTAG('j', 'p', 'e', 'g') }, /* PhotoJPEG */
{ CODEC_ID_MPEG1VIDEO, MKTAG('m', 'p', 'e', 'g') }, /* MPEG */
{ CODEC_ID_MJPEG, MKTAG('m', 'j', 'p', 'a') }, /* Motion-JPEG (format A) */
{ CODEC_ID_MJPEGB, MKTAG('m', 'j', 'p', 'b') }, /* Motion-JPEG (format B) */
{ CODEC_ID_MJPEG, MKTAG('A', 'V', 'D', 'J') }, /* MJPEG with alpha-channel (AVID JFIF meridien compressed) */
/* { CODEC_ID_MJPEG, MKTAG('A', 'V', 'R', 'n') }, *//* MJPEG with alpha-channel (AVID ABVB/Truevision NuVista) */
{ CODEC_ID_GIF, MKTAG('g', 'i', 'f', ' ') }, /* embedded gif files as frames (usually one "click to play movie" frame) */
/* Sorenson video */
{ CODEC_ID_SVQ1, MKTAG('S', 'V', 'Q', '1') }, /* Sorenson Video v1 */
{ CODEC_ID_SVQ1, MKTAG('s', 'v', 'q', '1') }, /* Sorenson Video v1 */
{ CODEC_ID_SVQ1, MKTAG('s', 'v', 'q', 'i') }, /* Sorenson Video v1 (from QT specs)*/
{ CODEC_ID_SVQ3, MKTAG('S', 'V', 'Q', '3') }, /* Sorenson Video v3 */
{ CODEC_ID_MPEG4, MKTAG('m', 'p', '4', 'v') },
{ CODEC_ID_MPEG4, MKTAG('D', 'I', 'V', 'X') }, /* OpenDiVX *//* sample files at http://heroinewarrior.com/xmovie.php3 use this tag */
{ CODEC_ID_MPEG4, MKTAG('X', 'V', 'I', 'D') },
{ CODEC_ID_MPEG4, MKTAG('3', 'I', 'V', '2') }, /* experimental: 3IVX files before ivx D4 4.5.1 */
/* { CODEC_ID_, MKTAG('I', 'V', '5', '0') }, *//* Indeo 5.0 */
{ CODEC_ID_H263, MKTAG('h', '2', '6', '3') }, /* H263 */
{ CODEC_ID_H263, MKTAG('s', '2', '6', '3') }, /* H263 ?? works */
{ CODEC_ID_DVVIDEO, MKTAG('d', 'v', 'c', ' ') }, /* DV NTSC */
{ CODEC_ID_DVVIDEO, MKTAG('d', 'v', 'c', 'p') }, /* DV PAL */
{ CODEC_ID_VP3, MKTAG('V', 'P', '3', '1') }, /* On2 VP3 */
{ CODEC_ID_RPZA, MKTAG('r', 'p', 'z', 'a') }, /* Apple Video (RPZA) */
{ CODEC_ID_CINEPAK, MKTAG('c', 'v', 'i', 'd') }, /* Cinepak */
{ CODEC_ID_8BPS, MKTAG('8', 'B', 'P', 'S') }, /* Planar RGB (8BPS) */
{ CODEC_ID_SMC, MKTAG('s', 'm', 'c', ' ') }, /* Apple Graphics (SMC) */
{ CODEC_ID_QTRLE, MKTAG('r', 'l', 'e', ' ') }, /* Apple Animation (RLE) */
{ CODEC_ID_MSRLE, MKTAG('W', 'R', 'L', 'E') },
{ CODEC_ID_QDRAW, MKTAG('q', 'd', 'r', 'w') }, /* QuickDraw */
{ CODEC_ID_H264, MKTAG('a', 'v', 'c', '1') }, /* AVC-1/H.264 */
{ CODEC_ID_MPEG2VIDEO, MKTAG('h', 'd', 'v', '2') }, /* MPEG2 produced by Sony HD camera */
{ CODEC_ID_MPEG2VIDEO, MKTAG('h', 'd', 'v', '3') }, /* HDV produced by FCP */
{ CODEC_ID_MPEG2VIDEO, MKTAG('m', 'x', '5', 'n') }, /* MPEG2 IMX NTSC 525/60 50mb/s produced by FCP */
{ CODEC_ID_MPEG2VIDEO, MKTAG('m', 'x', '5', 'p') }, /* MPEG2 IMX PAL 625/50 50mb/s produced by FCP */
{ CODEC_ID_MPEG2VIDEO, MKTAG('m', 'x', '3', 'n') }, /* MPEG2 IMX NTSC 525/60 30mb/s produced by FCP */
{ CODEC_ID_MPEG2VIDEO, MKTAG('m', 'x', '3', 'p') }, /* MPEG2 IMX PAL 625/50 30mb/s produced by FCP */
{ CODEC_ID_DVVIDEO, MKTAG('d', 'v', 'p', 'p') }, /* DVCPRO PAL produced by FCP */
//{ CODEC_ID_DVVIDEO, MKTAG('d', 'v', 'h', '5') }, /* DVCPRO HD 50i produced by FCP */
//{ CODEC_ID_DVVIDEO, MKTAG('d', 'v', 'h', '6') }, /* DVCPRO HD 60i produced by FCP */
{ CODEC_ID_DVVIDEO, MKTAG('d', 'v', '5', 'p') }, /* DVCPRO50 PAL produced by FCP */
{ CODEC_ID_DVVIDEO, MKTAG('d', 'v', '5', 'n') }, /* DVCPRO50 NTSC produced by FCP */
{ CODEC_ID_DVVIDEO, MKTAG('A', 'V', 'd', 'v') }, /* AVID DV */
//{ CODEC_ID_JPEG2000, MKTAG('m', 'j', 'p', '2') }, /* JPEG 2000 produced by FCP */
{ CODEC_ID_TARGA, MKTAG('t', 'g', 'a', ' ') }, /* Truevision Targa */
{ CODEC_ID_TIFF, MKTAG('t', 'i', 'f', 'f') }, /* TIFF embedded in MOV */
{ CODEC_ID_RAWVIDEO, MKTAG('2', 'v', 'u', 'y') }, /* UNCOMPRESSED 8BIT 4:2:2 */
{ CODEC_ID_VC1, MKTAG('v', 'c', '-', '1') }, /* SMPTE RP 2025 */
{ CODEC_ID_WMV3, MKTAG('v', 'c', '-', '1') }, /* SMPTE RP 2025 */
{ CODEC_ID_PNG, MKTAG('p', 'n', 'g', ' ') },
{ CODEC_ID_NONE, 0 },
};
static const AVCodecTag mov_audio_tags[] = {
{ CODEC_ID_PCM_S32BE, MKTAG('i', 'n', '3', '2') },
{ CODEC_ID_PCM_S24BE, MKTAG('i', 'n', '2', '4') },
{ CODEC_ID_PCM_S16BE, MKTAG('N', 'O', 'N', 'E') }, /* uncompressed */
{ CODEC_ID_PCM_S16BE, MKTAG('t', 'w', 'o', 's') }, /* 16 bits */
{ CODEC_ID_PCM_U8, MKTAG('r', 'a', 'w', ' ') }, /* 8 bits unsigned */
{ CODEC_ID_PCM_S16LE, MKTAG('s', 'o', 'w', 't') }, /* */
{ CODEC_ID_PCM_MULAW, MKTAG('u', 'l', 'a', 'w') }, /* */
{ CODEC_ID_PCM_ALAW, MKTAG('a', 'l', 'a', 'w') }, /* */
{ CODEC_ID_ADPCM_IMA_QT, MKTAG('i', 'm', 'a', '4') }, /* IMA-4 ADPCM */
{ CODEC_ID_ADPCM_MS, MKTAG('m', 's', 0x00, 0x02) }, /* MS ADPCM */
{ CODEC_ID_MACE3, MKTAG('M', 'A', 'C', '3') }, /* Macintosh Audio Compression and Expansion 3:1 */
{ CODEC_ID_MACE6, MKTAG('M', 'A', 'C', '6') }, /* Macintosh Audio Compression and Expansion 6:1 */
{ CODEC_ID_MP3, MKTAG('.', 'm', 'p', '3') }, /* MPEG layer 3 */ /* sample files at http://www.3ivx.com/showcase.html use this tag */
{ CODEC_ID_MP2, 0x6D730055 }, /* MPEG layer 3 */
{ CODEC_ID_MP2, 0x5500736D }, /* MPEG layer 3 *//* XXX: check endianness */
/* { CODEC_ID_OGG_VORBIS, MKTAG('O', 'g', 'g', 'S') }, *//* sample files at http://heroinewarrior.com/xmovie.php3 use this tag */
/* MP4 tags */
{ CODEC_ID_AAC, MKTAG('m', 'p', '4', 'a') }, /* MPEG-4 AAC */
/* The standard for mpeg4 audio is still not normalised AFAIK anyway */
{ CODEC_ID_AMR_NB, MKTAG('s', 'a', 'm', 'r') }, /* AMR-NB 3gp */
{ CODEC_ID_AMR_WB, MKTAG('s', 'a', 'w', 'b') }, /* AMR-WB 3gp */
{ CODEC_ID_AC3, MKTAG('m', 's', 0x20, 0x00) }, /* Dolby AC-3 */
{ CODEC_ID_ALAC,MKTAG('a', 'l', 'a', 'c') }, /* Apple Lossless */
{ CODEC_ID_QDM2,MKTAG('Q', 'D', 'M', '2') }, /* QDM2 */
{ CODEC_ID_DVAUDIO, MKTAG('v', 'd', 'v', 'a') },
{ CODEC_ID_DVAUDIO, MKTAG('d', 'v', 'c', 'a') },
{ CODEC_ID_PCM_S16LE, MKTAG('l', 'p', 'c', 'm') },
{ CODEC_ID_NONE, 0 },
};
/* the QuickTime file format is quite convoluted...
* it has lots of index tables, each indexing something in another one...
* Here we just use what is needed to read the chunks
*/
typedef struct MOV_sample_to_chunk_tbl {
long first;
long count;
long id;
} MOV_sample_to_chunk_tbl;
typedef struct {
uint32_t type;
int64_t offset;
int64_t size; /* total size (excluding the size and type fields) */
} MOV_atom_t;
typedef struct {
int seed;
int flags;
int size;
void* clrs;
} MOV_ctab_t;
typedef struct MOV_mdat_atom_s {
offset_t offset;
int64_t size;
} MOV_mdat_atom_t;
typedef struct {
uint8_t version;
uint32_t flags; // 24bit
/* 0x03 ESDescrTag */
uint16_t es_id;
#define MP4ODescrTag 0x01
#define MP4IODescrTag 0x02
#define MP4ESDescrTag 0x03
#define MP4DecConfigDescrTag 0x04
#define MP4DecSpecificDescrTag 0x05
#define MP4SLConfigDescrTag 0x06
#define MP4ContentIdDescrTag 0x07
#define MP4SupplContentIdDescrTag 0x08
#define MP4IPIPtrDescrTag 0x09
#define MP4IPMPPtrDescrTag 0x0A
#define MP4IPMPDescrTag 0x0B
#define MP4RegistrationDescrTag 0x0D
#define MP4ESIDIncDescrTag 0x0E
#define MP4ESIDRefDescrTag 0x0F
#define MP4FileIODescrTag 0x10
#define MP4FileODescrTag 0x11
#define MP4ExtProfileLevelDescrTag 0x13
#define MP4ExtDescrTagsStart 0x80
#define MP4ExtDescrTagsEnd 0xFE
uint8_t stream_priority;
/* 0x04 DecConfigDescrTag */
uint8_t object_type_id;
uint8_t stream_type;
/* XXX: really streamType is
* only 6bit, followed by:
* 1bit upStream
* 1bit reserved
*/
uint32_t buffer_size_db; // 24
uint32_t max_bitrate;
uint32_t avg_bitrate;
/* 0x05 DecSpecificDescrTag */
uint8_t decoder_cfg_len;
uint8_t *decoder_cfg;
/* 0x06 SLConfigDescrTag */
uint8_t sl_config_len;
uint8_t *sl_config;
} MOV_esds_t;
struct MOVParseTableEntry;
typedef struct MOVStreamContext {
int ffindex; /* the ffmpeg stream id */
long next_chunk;
unsigned int chunk_count;
int64_t *chunk_offsets;
unsigned int stts_count;
Time2Sample *stts_data;
unsigned int ctts_count;
Time2Sample *ctts_data;
unsigned int edit_count; /* number of 'edit' (elst atom) */
unsigned int sample_to_chunk_sz;
MOV_sample_to_chunk_tbl *sample_to_chunk;
int sample_to_ctime_index;
int sample_to_ctime_sample;
unsigned int sample_size;
unsigned int sample_count;
long *sample_sizes;
unsigned int keyframe_count;
long *keyframes;
int time_scale;
int time_rate;
long current_sample;
MOV_esds_t esds;
AVRational sample_size_v1;
int dv_audio_container;
} MOVStreamContext;
typedef struct MOVContext {
AVFormatContext *fc;
int time_scale;
int64_t duration; /* duration of the longest track */
int found_moov; /* when both 'moov' and 'mdat' sections has been found */
int found_mdat; /* we suppose we have enough data to read the file */
int64_t mdat_offset;
int total_streams;
MOVStreamContext *streams[MAX_STREAMS];
int ctab_size;
MOV_ctab_t **ctab; /* color tables */
const struct MOVParseTableEntry *parse_table; /* could be eventually used to change the table */
/* NOTE: for recursion save to/ restore from local variable! */
AVPaletteControl palette_control;
MOV_mdat_atom_t *mdat_list;
int mdat_count;
DVDemuxContext *dv_demux;
AVFormatContext *dv_fctx;
int isom; /* 1 if file is ISO Media (mp4/3gp) */
} MOVContext;
/* XXX: it's the first time I make a recursive parser I think... sorry if it's ugly :P */
/* those functions parse an atom */
/* return code:
1: found what I wanted, exit
0: continue to parse next atom
-1: error occured, exit
*/
typedef int (*mov_parse_function)(MOVContext *ctx, ByteIOContext *pb, MOV_atom_t atom);
/* links atom IDs to parse functions */
typedef struct MOVParseTableEntry {
uint32_t type;
mov_parse_function func;
} MOVParseTableEntry;
static int mov_read_default(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
int64_t total_size = 0;
MOV_atom_t a;
int i;
int err = 0;
a.offset = atom.offset;
if (atom.size < 0)
atom.size = 0x7fffffffffffffffLL;
while(((total_size + 8) < atom.size) && !url_feof(pb) && !err) {
a.size = atom.size;
a.type=0L;
if(atom.size >= 8) {
a.size = get_be32(pb);
a.type = get_le32(pb);
}
total_size += 8;
a.offset += 8;
dprintf("type: %08x %.4s sz: %"PRIx64" %"PRIx64" %"PRIx64"\n", a.type, (char*)&a.type, a.size, atom.size, total_size);
if (a.size == 1) { /* 64 bit extended size */
a.size = get_be64(pb) - 8;
a.offset += 8;
total_size += 8;
}
if (a.size == 0) {
a.size = atom.size - total_size;
if (a.size <= 8)
break;
}
for (i = 0; c->parse_table[i].type != 0L
&& c->parse_table[i].type != a.type; i++)
/* empty */;
a.size -= 8;
if(a.size < 0)
break;
if (c->parse_table[i].type == 0) { /* skip leaf atoms data */
url_fskip(pb, a.size);
} else {
offset_t start_pos = url_ftell(pb);
int64_t left;
err = (c->parse_table[i].func)(c, pb, a);
left = a.size - url_ftell(pb) + start_pos;
if (left > 0) /* skip garbage at atom end */
url_fskip(pb, left);
}
a.offset += a.size;
total_size += a.size;
}
if (!err && total_size < atom.size && atom.size < 0x7ffff) {
url_fskip(pb, atom.size - total_size);
}
return err;
}
static int mov_read_ctab(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
#if 1
url_fskip(pb, atom.size); // for now
#else
VERY VERY BROKEN, NEVER execute this, needs rewrite
unsigned int len;
MOV_ctab_t *t;
c->ctab = av_realloc(c->ctab, ++c->ctab_size);
t = c->ctab[c->ctab_size];
t->seed = get_be32(pb);
t->flags = get_be16(pb);
t->size = get_be16(pb) + 1;
len = 2 * t->size * 4;
if (len > 0) {
t->clrs = av_malloc(len); // 16bit A R G B
if (t->clrs)
get_buffer(pb, t->clrs, len);
}
#endif
return 0;
}
static int mov_read_hdlr(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
AVStream *st = c->fc->streams[c->fc->nb_streams-1];
uint32_t type;
uint32_t ctype;
get_byte(pb); /* version */
get_byte(pb); get_byte(pb); get_byte(pb); /* flags */
/* component type */
ctype = get_le32(pb);
type = get_le32(pb); /* component subtype */
dprintf("ctype= %c%c%c%c (0x%08lx)\n", *((char *)&ctype), ((char *)&ctype)[1], ((char *)&ctype)[2], ((char *)&ctype)[3], (long) ctype);
dprintf("stype= %c%c%c%c\n", *((char *)&type), ((char *)&type)[1], ((char *)&type)[2], ((char *)&type)[3]);
if(!ctype)
c->isom = 1;
if(type == MKTAG('v', 'i', 'd', 'e'))
st->codec->codec_type = CODEC_TYPE_VIDEO;
else if(type == MKTAG('s', 'o', 'u', 'n'))
st->codec->codec_type = CODEC_TYPE_AUDIO;
else if(type == MKTAG('m', '1', 'a', ' '))
st->codec->codec_id = CODEC_ID_MP2;
else if(type == MKTAG('s', 'u', 'b', 'p')) {
st->codec->codec_type = CODEC_TYPE_SUBTITLE;
st->codec->codec_id = CODEC_ID_DVD_SUBTITLE;
}
get_be32(pb); /* component manufacture */
get_be32(pb); /* component flags */
get_be32(pb); /* component flags mask */
if(atom.size <= 24)
return 0; /* nothing left to read */
url_fskip(pb, atom.size - (url_ftell(pb) - atom.offset));
return 0;
}
static int mov_mp4_read_descr_len(ByteIOContext *pb)
{
int len = 0;
int count = 4;
while (count--) {
int c = get_byte(pb);
len = (len << 7) | (c & 0x7f);
if (!(c & 0x80))
break;
}
return len;
}
static int mov_mp4_read_descr(ByteIOContext *pb, int *tag)
{
int len;
*tag = get_byte(pb);
len = mov_mp4_read_descr_len(pb);
dprintf("MPEG4 description: tag=0x%02x len=%d\n", *tag, len);
return len;
}
static int mov_read_esds(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
AVStream *st = c->fc->streams[c->fc->nb_streams-1];
MOVStreamContext *sc = (MOVStreamContext *)st->priv_data;
int tag, len;
/* Well, broken but suffisant for some MP4 streams */
get_be32(pb); /* version + flags */
len = mov_mp4_read_descr(pb, &tag);
if (tag == MP4ESDescrTag) {
get_be16(pb); /* ID */
get_byte(pb); /* priority */
} else
get_be16(pb); /* ID */
len = mov_mp4_read_descr(pb, &tag);
if (tag == MP4DecConfigDescrTag) {
sc->esds.object_type_id = get_byte(pb);
sc->esds.stream_type = get_byte(pb);
sc->esds.buffer_size_db = get_be24(pb);
sc->esds.max_bitrate = get_be32(pb);
sc->esds.avg_bitrate = get_be32(pb);
st->codec->codec_id= codec_get_id(ff_mov_obj_type, sc->esds.object_type_id);
dprintf("esds object type id %d\n", sc->esds.object_type_id);
len = mov_mp4_read_descr(pb, &tag);
if (tag == MP4DecSpecificDescrTag) {
dprintf("Specific MPEG4 header len=%d\n", len);
st->codec->extradata = av_mallocz(len + FF_INPUT_BUFFER_PADDING_SIZE);
if (st->codec->extradata) {
get_buffer(pb, st->codec->extradata, len);
st->codec->extradata_size = len;
/* from mplayer */
if ((*st->codec->extradata >> 3) == 29) {
st->codec->codec_id = CODEC_ID_MP3ON4;
}
}
}
}
return 0;
}
/* this atom contains actual media data */
static int mov_read_mdat(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
if(atom.size == 0) /* wrong one (MP4) */
return 0;
c->mdat_list = av_realloc(c->mdat_list, (c->mdat_count + 1) * sizeof(*c->mdat_list));
c->mdat_list[c->mdat_count].offset = atom.offset;
c->mdat_list[c->mdat_count].size = atom.size;
c->mdat_count++;
c->found_mdat=1;
c->mdat_offset = atom.offset;
if(c->found_moov)
return 1; /* found both, just go */
url_fskip(pb, atom.size);
return 0; /* now go for moov */
}
static int mov_read_ftyp(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
uint32_t type = get_le32(pb);
if (type != MKTAG('q','t',' ',' '))
c->isom = 1;
av_log(c->fc, AV_LOG_DEBUG, "ISO: File Type Major Brand: %.4s\n",(char *)&type);
get_be32(pb); /* minor version */
url_fskip(pb, atom.size - 8);
return 0;
}
/* this atom should contain all header atoms */
static int mov_read_moov(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
int err;
err = mov_read_default(c, pb, atom);
/* we parsed the 'moov' atom, we can terminate the parsing as soon as we find the 'mdat' */
/* so we don't parse the whole file if over a network */
c->found_moov=1;
if(c->found_mdat)
return 1; /* found both, just go */
return 0; /* now go for mdat */
}
static int mov_read_mdhd(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
AVStream *st = c->fc->streams[c->fc->nb_streams-1];
MOVStreamContext *sc = (MOVStreamContext *)st->priv_data;
int version = get_byte(pb);
int lang;
if (version > 1)
return 1; /* unsupported */
get_byte(pb); get_byte(pb);
get_byte(pb); /* flags */
if (version == 1) {
get_be64(pb);
get_be64(pb);
} else {
get_be32(pb); /* creation time */
get_be32(pb); /* modification time */
}
sc->time_scale = get_be32(pb);
st->duration = (version == 1) ? get_be64(pb) : get_be32(pb); /* duration */
lang = get_be16(pb); /* language */
ff_mov_lang_to_iso639(lang, st->language);
get_be16(pb); /* quality */
return 0;
}
static int mov_read_mvhd(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
int version = get_byte(pb); /* version */
get_byte(pb); get_byte(pb); get_byte(pb); /* flags */
if (version == 1) {
get_be64(pb);
get_be64(pb);
} else {
get_be32(pb); /* creation time */
get_be32(pb); /* modification time */
}
c->time_scale = get_be32(pb); /* time scale */
#ifdef DEBUG
av_log(NULL, AV_LOG_DEBUG, "time scale = %i\n", c->time_scale);
#endif
c->duration = (version == 1) ? get_be64(pb) : get_be32(pb); /* duration */
get_be32(pb); /* preferred scale */
get_be16(pb); /* preferred volume */
url_fskip(pb, 10); /* reserved */
url_fskip(pb, 36); /* display matrix */
get_be32(pb); /* preview time */
get_be32(pb); /* preview duration */
get_be32(pb); /* poster time */
get_be32(pb); /* selection time */
get_be32(pb); /* selection duration */
get_be32(pb); /* current time */
get_be32(pb); /* next track ID */
return 0;
}
static int mov_read_smi(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
AVStream *st = c->fc->streams[c->fc->nb_streams-1];
if((uint64_t)atom.size > (1<<30))
return -1;
// currently SVQ3 decoder expect full STSD header - so let's fake it
// this should be fixed and just SMI header should be passed
av_free(st->codec->extradata);
st->codec->extradata_size = 0x5a + atom.size;
st->codec->extradata = av_mallocz(st->codec->extradata_size + FF_INPUT_BUFFER_PADDING_SIZE);
if (st->codec->extradata) {
memcpy(st->codec->extradata, "SVQ3", 4); // fake
get_buffer(pb, st->codec->extradata + 0x5a, atom.size);
dprintf("Reading SMI %"PRId64" %s\n", atom.size, st->codec->extradata + 0x5a);
} else
url_fskip(pb, atom.size);
return 0;
}
static int mov_read_enda(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
AVStream *st = c->fc->streams[c->fc->nb_streams-1];
int little_endian = get_be16(pb);
if (little_endian) {
switch (st->codec->codec_id) {
case CODEC_ID_PCM_S24BE:
st->codec->codec_id = CODEC_ID_PCM_S24LE;
break;
case CODEC_ID_PCM_S32BE:
st->codec->codec_id = CODEC_ID_PCM_S32LE;
break;
default:
break;
}
}
return 0;
}
static int mov_read_alac(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
AVStream *st = c->fc->streams[c->fc->nb_streams-1];
// currently ALAC decoder expect full atom header - so let's fake it
// this should be fixed and just ALAC header should be passed
av_free(st->codec->extradata);
st->codec->extradata_size = 36;
st->codec->extradata = av_mallocz(st->codec->extradata_size + FF_INPUT_BUFFER_PADDING_SIZE);
if (st->codec->extradata) {
memcpy(st->codec->extradata + 4, "alac", 4); // fake
get_buffer(pb, st->codec->extradata + 8, 36 - 8);
dprintf("Reading alac %d %s\n", st->codec->extradata_size, st->codec->extradata);
} else
url_fskip(pb, atom.size);
return 0;
}
static int mov_read_wave(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
AVStream *st = c->fc->streams[c->fc->nb_streams-1];
if((uint64_t)atom.size > (1<<30))
return -1;
if (st->codec->codec_id == CODEC_ID_QDM2) {
// pass all frma atom to codec, needed at least for QDM2
av_free(st->codec->extradata);
st->codec->extradata_size = atom.size;
st->codec->extradata = av_mallocz(st->codec->extradata_size + FF_INPUT_BUFFER_PADDING_SIZE);
if (st->codec->extradata) {
get_buffer(pb, st->codec->extradata, atom.size);
} else
url_fskip(pb, atom.size);
} else if (atom.size > 8) { /* to read frma, esds atoms */
mov_read_default(c, pb, atom);
} else
url_fskip(pb, atom.size);
return 0;
}
static int mov_read_jp2h(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
AVStream *st = c->fc->streams[c->fc->nb_streams-1];
if((uint64_t)atom.size > (1<<30))
return -1;
av_free(st->codec->extradata);
st->codec->extradata_size = atom.size + 8;
st->codec->extradata = av_mallocz(st->codec->extradata_size + FF_INPUT_BUFFER_PADDING_SIZE);
/* pass all jp2h atom to codec */
if (st->codec->extradata) {
memcpy(st->codec->extradata + 4, "jp2h", 4);
get_buffer(pb, st->codec->extradata + 8, atom.size);
} else
url_fskip(pb, atom.size);
return 0;
}
static int mov_read_avcC(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
AVStream *st = c->fc->streams[c->fc->nb_streams-1];
if((uint64_t)atom.size > (1<<30))
return -1;
av_free(st->codec->extradata);
st->codec->extradata_size = atom.size;
st->codec->extradata = av_mallocz(st->codec->extradata_size + FF_INPUT_BUFFER_PADDING_SIZE);
if (st->codec->extradata) {
get_buffer(pb, st->codec->extradata, atom.size);
} else
url_fskip(pb, atom.size);
return 0;
}
static int mov_read_stco(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
AVStream *st = c->fc->streams[c->fc->nb_streams-1];
MOVStreamContext *sc = (MOVStreamContext *)st->priv_data;
unsigned int i, entries;
get_byte(pb); /* version */
get_byte(pb); get_byte(pb); get_byte(pb); /* flags */
entries = get_be32(pb);
if(entries >= UINT_MAX/sizeof(int64_t))
return -1;
sc->chunk_count = entries;
sc->chunk_offsets = av_malloc(entries * sizeof(int64_t));
if (!sc->chunk_offsets)
return -1;
if (atom.type == MKTAG('s', 't', 'c', 'o')) {
for(i=0; i<entries; i++) {
sc->chunk_offsets[i] = get_be32(pb);
}
} else if (atom.type == MKTAG('c', 'o', '6', '4')) {
for(i=0; i<entries; i++) {
sc->chunk_offsets[i] = get_be64(pb);
}
} else
return -1;
return 0;
}
static int mov_read_stsd(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
AVStream *st = c->fc->streams[c->fc->nb_streams-1];
MOVStreamContext *sc = (MOVStreamContext *)st->priv_data;
int entries, frames_per_sample;
uint32_t format;
uint8_t codec_name[32];
/* for palette traversal */
int color_depth;
int color_start;
int color_count;
int color_end;
int color_index;
int color_dec;
int color_greyscale;
unsigned char *color_table;
int j;
unsigned char r, g, b;
get_byte(pb); /* version */
get_byte(pb); get_byte(pb); get_byte(pb); /* flags */
entries = get_be32(pb);
while(entries--) { //Parsing Sample description table
enum CodecID id;
MOV_atom_t a = { 0, 0, 0 };
offset_t start_pos = url_ftell(pb);
int size = get_be32(pb); /* size */
format = get_le32(pb); /* data format */
get_be32(pb); /* reserved */
get_be16(pb); /* reserved */
get_be16(pb); /* index */
if (st->codec->codec_tag) {
/* multiple fourcc, just skip for now */
url_fskip(pb, size - (url_ftell(pb) - start_pos));
continue;
}
st->codec->codec_tag = format;
id = codec_get_id(mov_audio_tags, format);
if (st->codec->codec_type != CODEC_TYPE_VIDEO && id > 0) {
st->codec->codec_type = CODEC_TYPE_AUDIO;
} else if (st->codec->codec_type != CODEC_TYPE_AUDIO && /* do not overwrite codec type */
format && format != MKTAG('m', 'p', '4', 's')) { /* skip old asf mpeg4 tag */
id = codec_get_id(mov_video_tags, format);
if (id <= 0)
id = codec_get_id(codec_bmp_tags, format);
if (id > 0)
st->codec->codec_type = CODEC_TYPE_VIDEO;
}
dprintf("size=%d 4CC= %c%c%c%c codec_type=%d\n",
size,
(format >> 0) & 0xff, (format >> 8) & 0xff, (format >> 16) & 0xff, (format >> 24) & 0xff,
st->codec->codec_type);
if(st->codec->codec_type==CODEC_TYPE_VIDEO) {
st->codec->codec_id = id;
get_be16(pb); /* version */
get_be16(pb); /* revision level */
get_be32(pb); /* vendor */
get_be32(pb); /* temporal quality */
get_be32(pb); /* spacial quality */
st->codec->width = get_be16(pb); /* width */
st->codec->height = get_be16(pb); /* height */
get_be32(pb); /* horiz resolution */
get_be32(pb); /* vert resolution */
get_be32(pb); /* data size, always 0 */
frames_per_sample = get_be16(pb); /* frames per samples */
#ifdef DEBUG
av_log(NULL, AV_LOG_DEBUG, "frames/samples = %d\n", frames_per_sample);
#endif
get_buffer(pb, codec_name, 32); /* codec name, pascal string (FIXME: true for mp4?) */
if (codec_name[0] <= 31) {
memcpy(st->codec->codec_name, &codec_name[1],codec_name[0]);
st->codec->codec_name[codec_name[0]] = 0;
}
st->codec->bits_per_sample = get_be16(pb); /* depth */
st->codec->color_table_id = get_be16(pb); /* colortable id */
/* figure out the palette situation */
color_depth = st->codec->bits_per_sample & 0x1F;
color_greyscale = st->codec->bits_per_sample & 0x20;
/* if the depth is 2, 4, or 8 bpp, file is palettized */
if ((color_depth == 2) || (color_depth == 4) ||
(color_depth == 8)) {
if (color_greyscale) {
/* compute the greyscale palette */
color_count = 1 << color_depth;
color_index = 255;
color_dec = 256 / (color_count - 1);
for (j = 0; j < color_count; j++) {
r = g = b = color_index;
c->palette_control.palette[j] =
(r << 16) | (g << 8) | (b);
color_index -= color_dec;
if (color_index < 0)
color_index = 0;
}
} else if (st->codec->color_table_id & 0x08) {
/* if flag bit 3 is set, use the default palette */
color_count = 1 << color_depth;
if (color_depth == 2)
color_table = ff_qt_default_palette_4;
else if (color_depth == 4)
color_table = ff_qt_default_palette_16;
else
color_table = ff_qt_default_palette_256;
for (j = 0; j < color_count; j++) {
r = color_table[j * 4 + 0];
g = color_table[j * 4 + 1];
b = color_table[j * 4 + 2];
c->palette_control.palette[j] =
(r << 16) | (g << 8) | (b);
}
} else {
/* load the palette from the file */
color_start = get_be32(pb);
color_count = get_be16(pb);
color_end = get_be16(pb);
for (j = color_start; j <= color_end; j++) {
/* each R, G, or B component is 16 bits;
* only use the top 8 bits; skip alpha bytes
* up front */
get_byte(pb);
get_byte(pb);
r = get_byte(pb);
get_byte(pb);
g = get_byte(pb);
get_byte(pb);
b = get_byte(pb);
get_byte(pb);
c->palette_control.palette[j] =
(r << 16) | (g << 8) | (b);
}
}
st->codec->palctrl = &c->palette_control;
st->codec->palctrl->palette_changed = 1;
} else
st->codec->palctrl = NULL;
} else if(st->codec->codec_type==CODEC_TYPE_AUDIO) {
int bits_per_sample;
uint16_t version = get_be16(pb);
st->codec->codec_id = id;
get_be16(pb); /* revision level */
get_be32(pb); /* vendor */
st->codec->channels = get_be16(pb); /* channel count */
dprintf("audio channels %d\n", st->codec->channels);
st->codec->bits_per_sample = get_be16(pb); /* sample size */
/* do we need to force to 16 for AMR ? */
/* handle specific s8 codec */
get_be16(pb); /* compression id = 0*/
get_be16(pb); /* packet size = 0 */
st->codec->sample_rate = ((get_be32(pb) >> 16));
switch (st->codec->codec_id) {
case CODEC_ID_PCM_S8:
case CODEC_ID_PCM_U8:
if (st->codec->bits_per_sample == 16)
st->codec->codec_id = CODEC_ID_PCM_S16BE;
break;
case CODEC_ID_PCM_S16LE:
case CODEC_ID_PCM_S16BE:
if (st->codec->bits_per_sample == 8)
st->codec->codec_id = CODEC_ID_PCM_S8;
else if (st->codec->bits_per_sample == 24)
st->codec->codec_id = CODEC_ID_PCM_S24BE;
break;
default:
break;
}
//Read QT version 1 fields. In version 0 theese dont exist
dprintf("version =%d, isom =%d\n",version,c->isom);
if(!c->isom) {
if(version==1) {
sc->sample_size_v1.den = get_be32(pb); /* samples per packet */
get_be32(pb); /* bytes per packet */
sc->sample_size_v1.num = get_be32(pb); /* bytes per frame */
get_be32(pb); /* bytes per sample */
} else if(version==2) {
get_be32(pb); /* sizeof struct only */
st->codec->sample_rate = av_int2dbl(get_be64(pb)); /* float 64 */
st->codec->channels = get_be32(pb);
get_be32(pb); /* always 0x7F000000 */
get_be32(pb); /* bits per channel if sound is uncompressed */
get_be32(pb); /* lcpm format specific flag */
get_be32(pb); /* bytes per audio packet if constant */
get_be32(pb); /* lpcm frames per audio packet if constant */
}
}
bits_per_sample = av_get_bits_per_sample(st->codec->codec_id);
if (bits_per_sample) {
st->codec->bits_per_sample = bits_per_sample;
sc->sample_size = (bits_per_sample >> 3) * st->codec->channels;
}
} else {
/* other codec type, just skip (rtp, mp4s, tmcd ...) */
url_fskip(pb, size - (url_ftell(pb) - start_pos));
}
/* this will read extra atoms at the end (wave, alac, damr, avcC, SMI ...) */
a.size = size - (url_ftell(pb) - start_pos);
if (a.size > 8)
mov_read_default(c, pb, a);
else if (a.size > 0)
url_fskip(pb, a.size);
}
if(st->codec->codec_type==CODEC_TYPE_AUDIO && st->codec->sample_rate==0 && sc->time_scale>1) {
st->codec->sample_rate= sc->time_scale;
}
/* special codec parameters handling */
switch (st->codec->codec_id) {
#ifdef CONFIG_H261_DECODER
case CODEC_ID_H261:
#endif
#ifdef CONFIG_H263_DECODER
case CODEC_ID_H263:
#endif
#ifdef CONFIG_MPEG4_DECODER
case CODEC_ID_MPEG4:
#endif
st->codec->width= 0; /* let decoder init width/height */
st->codec->height= 0;
break;
#ifdef CONFIG_LIBFAAD
case CODEC_ID_AAC:
#endif
#ifdef CONFIG_VORBIS_DECODER
case CODEC_ID_VORBIS:
#endif
case CODEC_ID_MP3ON4:
st->codec->sample_rate= 0; /* let decoder init parameters properly */
break;
#ifdef CONFIG_DV_DEMUXER
case CODEC_ID_DVAUDIO:
c->dv_fctx = av_alloc_format_context();
c->dv_demux = dv_init_demux(c->dv_fctx);
if (!c->dv_demux) {
av_log(c->fc, AV_LOG_ERROR, "dv demux context init error\n");
return -1;
}
sc->dv_audio_container = 1;
st->codec->codec_id = CODEC_ID_PCM_S16LE;
break;
#endif
/* no ifdef since parameters are always those */
case CODEC_ID_AMR_WB:
st->codec->sample_rate= 16000;
st->codec->channels= 1; /* really needed */
break;
case CODEC_ID_AMR_NB:
st->codec->sample_rate= 8000;
st->codec->channels= 1; /* really needed */
break;
case CODEC_ID_MP2:
st->codec->codec_type = CODEC_TYPE_AUDIO; /* force type after stsd for m1a hdlr */
st->need_parsing = 1;
break;
default:
break;
}
return 0;
}
static int mov_read_stsc(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
AVStream *st = c->fc->streams[c->fc->nb_streams-1];
MOVStreamContext *sc = (MOVStreamContext *)st->priv_data;
unsigned int i, entries;
get_byte(pb); /* version */
get_byte(pb); get_byte(pb); get_byte(pb); /* flags */
entries = get_be32(pb);
if(entries >= UINT_MAX / sizeof(MOV_sample_to_chunk_tbl))
return -1;
#ifdef DEBUG
av_log(NULL, AV_LOG_DEBUG, "track[%i].stsc.entries = %i\n", c->fc->nb_streams-1, entries);
#endif
sc->sample_to_chunk_sz = entries;
sc->sample_to_chunk = av_malloc(entries * sizeof(MOV_sample_to_chunk_tbl));
if (!sc->sample_to_chunk)
return -1;
for(i=0; i<entries; i++) {
sc->sample_to_chunk[i].first = get_be32(pb);
sc->sample_to_chunk[i].count = get_be32(pb);
sc->sample_to_chunk[i].id = get_be32(pb);
}
return 0;
}
static int mov_read_stss(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
AVStream *st = c->fc->streams[c->fc->nb_streams-1];
MOVStreamContext *sc = (MOVStreamContext *)st->priv_data;
unsigned int i, entries;
get_byte(pb); /* version */
get_byte(pb); get_byte(pb); get_byte(pb); /* flags */
entries = get_be32(pb);
if(entries >= UINT_MAX / sizeof(long))
return -1;
sc->keyframe_count = entries;
#ifdef DEBUG
av_log(NULL, AV_LOG_DEBUG, "keyframe_count = %ld\n", sc->keyframe_count);
#endif
sc->keyframes = av_malloc(entries * sizeof(long));
if (!sc->keyframes)
return -1;
for(i=0; i<entries; i++) {
sc->keyframes[i] = get_be32(pb);
#ifdef DEBUG
/* av_log(NULL, AV_LOG_DEBUG, "keyframes[]=%ld\n", sc->keyframes[i]); */
#endif
}
return 0;
}
static int mov_read_stsz(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
AVStream *st = c->fc->streams[c->fc->nb_streams-1];
MOVStreamContext *sc = (MOVStreamContext *)st->priv_data;
unsigned int i, entries, sample_size;
get_byte(pb); /* version */
get_byte(pb); get_byte(pb); get_byte(pb); /* flags */
sample_size = get_be32(pb);
if (!sc->sample_size) /* do not overwrite value computed in stsd */
sc->sample_size = sample_size;
entries = get_be32(pb);
if(entries >= UINT_MAX / sizeof(long))
return -1;
sc->sample_count = entries;
if (sample_size)
return 0;
#ifdef DEBUG
av_log(NULL, AV_LOG_DEBUG, "sample_size = %ld sample_count = %ld\n", sc->sample_size, sc->sample_count);
#endif
sc->sample_sizes = av_malloc(entries * sizeof(long));
if (!sc->sample_sizes)
return -1;
for(i=0; i<entries; i++) {
sc->sample_sizes[i] = get_be32(pb);
#ifdef DEBUG
av_log(NULL, AV_LOG_DEBUG, "sample_sizes[]=%ld\n", sc->sample_sizes[i]);
#endif
}
return 0;
}
static int mov_read_stts(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
AVStream *st = c->fc->streams[c->fc->nb_streams-1];
MOVStreamContext *sc = (MOVStreamContext *)st->priv_data;
unsigned int i, entries;
int64_t duration=0;
int64_t total_sample_count=0;
get_byte(pb); /* version */
get_byte(pb); get_byte(pb); get_byte(pb); /* flags */
entries = get_be32(pb);
if(entries >= UINT_MAX / sizeof(Time2Sample))
return -1;
sc->stts_count = entries;
sc->stts_data = av_malloc(entries * sizeof(Time2Sample));
#ifdef DEBUG
av_log(NULL, AV_LOG_DEBUG, "track[%i].stts.entries = %i\n", c->fc->nb_streams-1, entries);
#endif
sc->time_rate=0;
for(i=0; i<entries; i++) {
int sample_duration;
int sample_count;
sample_count=get_be32(pb);
sample_duration = get_be32(pb);
sc->stts_data[i].count= sample_count;
sc->stts_data[i].duration= sample_duration;
sc->time_rate= ff_gcd(sc->time_rate, sample_duration);
dprintf("sample_count=%d, sample_duration=%d\n",sample_count,sample_duration);
duration+=(int64_t)sample_duration*sample_count;
total_sample_count+=sample_count;
}
st->nb_frames= total_sample_count;
if(duration)
st->duration= duration;
return 0;
}
static int mov_read_ctts(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
AVStream *st = c->fc->streams[c->fc->nb_streams-1];
MOVStreamContext *sc = (MOVStreamContext *)st->priv_data;
unsigned int i, entries;
get_byte(pb); /* version */
get_byte(pb); get_byte(pb); get_byte(pb); /* flags */
entries = get_be32(pb);
if(entries >= UINT_MAX / sizeof(Time2Sample))
return -1;
sc->ctts_count = entries;
sc->ctts_data = av_malloc(entries * sizeof(Time2Sample));
dprintf("track[%i].ctts.entries = %i\n", c->fc->nb_streams-1, entries);
for(i=0; i<entries; i++) {
int count =get_be32(pb);
int duration =get_be32(pb);
if (duration < 0) {
av_log(c->fc, AV_LOG_ERROR, "negative ctts, ignoring\n");
sc->ctts_count = 0;
url_fskip(pb, 8 * (entries - i - 1));
break;
}
sc->ctts_data[i].count = count;
sc->ctts_data[i].duration= duration;
sc->time_rate= ff_gcd(sc->time_rate, duration);
}
return 0;
}
static int mov_read_trak(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
AVStream *st;
MOVStreamContext *sc;
st = av_new_stream(c->fc, c->fc->nb_streams);
if (!st) return -2;
sc = av_mallocz(sizeof(MOVStreamContext));
if (!sc) {
av_free(st);
return -1;
}
st->priv_data = sc;
st->codec->codec_type = CODEC_TYPE_DATA;
st->start_time = 0; /* XXX: check */
c->streams[c->fc->nb_streams-1] = sc;
return mov_read_default(c, pb, atom);
}
static int mov_read_tkhd(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
AVStream *st = c->fc->streams[c->fc->nb_streams-1];
int version = get_byte(pb);
get_byte(pb); get_byte(pb);
get_byte(pb); /* flags */
/*
MOV_TRACK_ENABLED 0x0001
MOV_TRACK_IN_MOVIE 0x0002
MOV_TRACK_IN_PREVIEW 0x0004
MOV_TRACK_IN_POSTER 0x0008
*/
if (version == 1) {
get_be64(pb);
get_be64(pb);
} else {
get_be32(pb); /* creation time */
get_be32(pb); /* modification time */
}
st->id = (int)get_be32(pb); /* track id (NOT 0 !)*/
get_be32(pb); /* reserved */
st->start_time = 0; /* check */
(version == 1) ? get_be64(pb) : get_be32(pb); /* highlevel (considering edits) duration in movie timebase */
get_be32(pb); /* reserved */
get_be32(pb); /* reserved */
get_be16(pb); /* layer */
get_be16(pb); /* alternate group */
get_be16(pb); /* volume */
get_be16(pb); /* reserved */
url_fskip(pb, 36); /* display matrix */
/* those are fixed-point */
get_be32(pb); /* track width */
get_be32(pb); /* track height */
return 0;
}
/* this atom should be null (from specs), but some buggy files put the 'moov' atom inside it... */
/* like the files created with Adobe Premiere 5.0, for samples see */
/* http://graphics.tudelft.nl/~wouter/publications/soundtests/ */
static int mov_read_wide(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
int err;
if (atom.size < 8)
return 0; /* continue */
if (get_be32(pb) != 0) { /* 0 sized mdat atom... use the 'wide' atom size */
url_fskip(pb, atom.size - 4);
return 0;
}
atom.type = get_le32(pb);
atom.offset += 8;
atom.size -= 8;
if (atom.type != MKTAG('m', 'd', 'a', 't')) {
url_fskip(pb, atom.size);
return 0;
}
err = mov_read_mdat(c, pb, atom);
return err;
}
static int mov_read_cmov(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
#ifdef CONFIG_ZLIB
ByteIOContext ctx;
uint8_t *cmov_data;
uint8_t *moov_data; /* uncompressed data */
long cmov_len, moov_len;
int ret;
get_be32(pb); /* dcom atom */
if (get_le32(pb) != MKTAG( 'd', 'c', 'o', 'm' ))
return -1;
if (get_le32(pb) != MKTAG( 'z', 'l', 'i', 'b' )) {
av_log(NULL, AV_LOG_ERROR, "unknown compression for cmov atom !");
return -1;
}
get_be32(pb); /* cmvd atom */
if (get_le32(pb) != MKTAG( 'c', 'm', 'v', 'd' ))
return -1;
moov_len = get_be32(pb); /* uncompressed size */
cmov_len = atom.size - 6 * 4;
cmov_data = av_malloc(cmov_len);
if (!cmov_data)
return -1;
moov_data = av_malloc(moov_len);
if (!moov_data) {
av_free(cmov_data);
return -1;
}
get_buffer(pb, cmov_data, cmov_len);
if(uncompress (moov_data, (uLongf *) &moov_len, (const Bytef *)cmov_data, cmov_len) != Z_OK)
return -1;
if(init_put_byte(&ctx, moov_data, moov_len, 0, NULL, NULL, NULL, NULL) != 0)
return -1;
atom.type = MKTAG( 'm', 'o', 'o', 'v' );
atom.offset = 0;
atom.size = moov_len;
#ifdef DEBUG
// { int fd = open("/tmp/uncompheader.mov", O_WRONLY | O_CREAT); write(fd, moov_data, moov_len); close(fd); }
#endif
ret = mov_read_default(c, &ctx, atom);
av_free(moov_data);
av_free(cmov_data);
return ret;
#else
av_log(c->fc, AV_LOG_ERROR, "this file requires zlib support compiled in\n");
return -1;
#endif
}
/* edit list atom */
static int mov_read_elst(MOVContext *c, ByteIOContext *pb, MOV_atom_t atom)
{
int i, edit_count;
get_byte(pb); /* version */
get_byte(pb); get_byte(pb); get_byte(pb); /* flags */
edit_count= c->streams[c->fc->nb_streams-1]->edit_count = get_be32(pb); /* entries */
for(i=0; i<edit_count; i++){
get_be32(pb); /* Track duration */
get_be32(pb); /* Media time */
get_be32(pb); /* Media rate */
}
dprintf("track[%i].edit_count = %i\n", c->fc->nb_streams-1, c->streams[c->fc->nb_streams-1]->edit_count);
return 0;
}
static const MOVParseTableEntry mov_default_parse_table[] = {
/* mp4 atoms */
{ MKTAG( 'c', 'o', '6', '4' ), mov_read_stco },
{ MKTAG( 'c', 't', 't', 's' ), mov_read_ctts }, /* composition time to sample */
{ MKTAG( 'e', 'd', 't', 's' ), mov_read_default },
{ MKTAG( 'e', 'l', 's', 't' ), mov_read_elst },
{ MKTAG( 'e', 'n', 'd', 'a' ), mov_read_enda },
{ MKTAG( 'f', 't', 'y', 'p' ), mov_read_ftyp },
{ MKTAG( 'h', 'd', 'l', 'r' ), mov_read_hdlr },
{ MKTAG( 'j', 'p', '2', 'h' ), mov_read_jp2h },
{ MKTAG( 'm', 'd', 'a', 't' ), mov_read_mdat },
{ MKTAG( 'm', 'd', 'h', 'd' ), mov_read_mdhd },
{ MKTAG( 'm', 'd', 'i', 'a' ), mov_read_default },
{ MKTAG( 'm', 'i', 'n', 'f' ), mov_read_default },
{ MKTAG( 'm', 'o', 'o', 'v' ), mov_read_moov },
{ MKTAG( 'm', 'v', 'h', 'd' ), mov_read_mvhd },
{ MKTAG( 'S', 'M', 'I', ' ' ), mov_read_smi }, /* Sorenson extension ??? */
{ MKTAG( 'a', 'l', 'a', 'c' ), mov_read_alac }, /* alac specific atom */
{ MKTAG( 'a', 'v', 'c', 'C' ), mov_read_avcC },
{ MKTAG( 's', 't', 'b', 'l' ), mov_read_default },
{ MKTAG( 's', 't', 'c', 'o' ), mov_read_stco },
{ MKTAG( 's', 't', 's', 'c' ), mov_read_stsc },
{ MKTAG( 's', 't', 's', 'd' ), mov_read_stsd }, /* sample description */
{ MKTAG( 's', 't', 's', 's' ), mov_read_stss }, /* sync sample */
{ MKTAG( 's', 't', 's', 'z' ), mov_read_stsz }, /* sample size */
{ MKTAG( 's', 't', 't', 's' ), mov_read_stts },
{ MKTAG( 't', 'k', 'h', 'd' ), mov_read_tkhd }, /* track header */
{ MKTAG( 't', 'r', 'a', 'k' ), mov_read_trak },
{ MKTAG( 'w', 'a', 'v', 'e' ), mov_read_wave },
{ MKTAG( 'c', 't', 'a', 'b' ), mov_read_ctab },
{ MKTAG( 'e', 's', 'd', 's' ), mov_read_esds },
{ MKTAG( 'w', 'i', 'd', 'e' ), mov_read_wide }, /* place holder */
{ MKTAG( 'c', 'm', 'o', 'v' ), mov_read_cmov },
{ 0L, NULL }
};
static void mov_free_stream_context(MOVStreamContext *sc)
{
if(sc) {
av_freep(&sc->ctts_data);
av_freep(&sc);
}
}
/* XXX: is it sufficient ? */
static int mov_probe(AVProbeData *p)
{
unsigned int offset;
uint32_t tag;
int score = 0;
/* check file header */
if (p->buf_size <= 12)
return 0;
offset = 0;
for(;;) {
/* ignore invalid offset */
if ((offset + 8) > (unsigned int)p->buf_size)
return score;
tag = AV_RL32(p->buf + offset + 4);
switch(tag) {
/* check for obvious tags */
case MKTAG( 'j', 'P', ' ', ' ' ): /* jpeg 2000 signature */
case MKTAG( 'm', 'o', 'o', 'v' ):
case MKTAG( 'm', 'd', 'a', 't' ):
case MKTAG( 'p', 'n', 'o', 't' ): /* detect movs with preview pics like ew.mov and april.mov */
case MKTAG( 'u', 'd', 't', 'a' ): /* Packet Video PVAuthor adds this and a lot of more junk */
return AVPROBE_SCORE_MAX;
/* those are more common words, so rate then a bit less */
case MKTAG( 'w', 'i', 'd', 'e' ):
case MKTAG( 'f', 'r', 'e', 'e' ):
case MKTAG( 'j', 'u', 'n', 'k' ):
case MKTAG( 'p', 'i', 'c', 't' ):
return AVPROBE_SCORE_MAX - 5;
case MKTAG( 'f', 't', 'y', 'p' ):
case MKTAG( 's', 'k', 'i', 'p' ):
case MKTAG( 'u', 'u', 'i', 'd' ):
offset = AV_RB32(p->buf+offset) + offset;
/* if we only find those cause probedata is too small at least rate them */
score = AVPROBE_SCORE_MAX - 50;
break;
default:
/* unrecognized tag */
return score;
}
}
return score;
}
static void mov_build_index(MOVContext *mov, AVStream *st)
{
MOVStreamContext *sc = st->priv_data;
offset_t current_offset;
int64_t current_dts = 0;
int stts_index = 0;
int stsc_index = 0;
int stss_index = 0;
int i, j, k;
if (sc->sample_sizes || st->codec->codec_type == CODEC_TYPE_VIDEO || sc->dv_audio_container) {
int keyframe, sample_size;
int current_sample = 0;
int stts_sample = 0;
int distance = 0;
st->nb_frames = sc->sample_count;
for (i = 0; i < sc->chunk_count; i++) {
current_offset = sc->chunk_offsets[i];
if (stsc_index + 1 < sc->sample_to_chunk_sz && i + 1 == sc->sample_to_chunk[stsc_index + 1].first)
stsc_index++;
for (j = 0; j < sc->sample_to_chunk[stsc_index].count; j++) {
if (current_sample >= sc->sample_count) {
av_log(mov->fc, AV_LOG_ERROR, "wrong sample count\n");
goto out;
}
keyframe = !sc->keyframe_count || current_sample + 1 == sc->keyframes[stss_index];
if (keyframe) {
distance = 0;
if (stss_index + 1 < sc->keyframe_count)
stss_index++;
}
sample_size = sc->sample_size > 0 ? sc->sample_size : sc->sample_sizes[current_sample];
dprintf("AVIndex stream %d, sample %d, offset %"PRIx64", dts %"PRId64", size %d, distance %d, keyframe %d\n",
st->index, current_sample, current_offset, current_dts, sample_size, distance, keyframe);
av_add_index_entry(st, current_offset, current_dts, sample_size, distance, keyframe ? AVINDEX_KEYFRAME : 0);
current_offset += sample_size;
assert(sc->stts_data[stts_index].duration % sc->time_rate == 0);
current_dts += sc->stts_data[stts_index].duration / sc->time_rate;
distance++;
stts_sample++;
current_sample++;
if (stts_index + 1 < sc->stts_count && stts_sample == sc->stts_data[stts_index].count) {
stts_sample = 0;
stts_index++;
}
}
}
} else { /* read whole chunk */
int chunk_samples, chunk_size, chunk_duration;
for (i = 0; i < sc->chunk_count; i++) {
current_offset = sc->chunk_offsets[i];
if (stsc_index + 1 < sc->sample_to_chunk_sz && i + 1 == sc->sample_to_chunk[stsc_index + 1].first)
stsc_index++;
chunk_samples = sc->sample_to_chunk[stsc_index].count;
/* get chunk size */
if (sc->sample_size > 1 || st->codec->codec_id == CODEC_ID_PCM_U8 || st->codec->codec_id == CODEC_ID_PCM_S8)
chunk_size = chunk_samples * sc->sample_size;
else if (sc->sample_size_v1.den > 0 && (chunk_samples * sc->sample_size_v1.num % sc->sample_size_v1.den == 0))
chunk_size = chunk_samples * sc->sample_size_v1.num / sc->sample_size_v1.den;
else { /* workaround to find nearest next chunk offset */
chunk_size = INT_MAX;
for (j = 0; j < mov->total_streams; j++) {
MOVStreamContext *msc = mov->streams[j];
for (k = msc->next_chunk; k < msc->chunk_count; k++) {
if (msc->chunk_offsets[k] > current_offset && msc->chunk_offsets[k] - current_offset < chunk_size) {
chunk_size = msc->chunk_offsets[k] - current_offset;
msc->next_chunk = k;
break;
}
}
}
/* check for last chunk */
if (chunk_size == INT_MAX)
for (j = 0; j < mov->mdat_count; j++) {
dprintf("mdat %d, offset %"PRIx64", size %"PRId64", current offset %"PRIx64"\n",
j, mov->mdat_list[j].offset, mov->mdat_list[j].size, current_offset);
if (mov->mdat_list[j].offset <= current_offset && mov->mdat_list[j].offset + mov->mdat_list[j].size > current_offset)
chunk_size = mov->mdat_list[j].offset + mov->mdat_list[j].size - current_offset;
}
assert(chunk_size != INT_MAX);
for (j = 0; j < mov->total_streams; j++) {
mov->streams[j]->next_chunk = 0;
}
}
av_add_index_entry(st, current_offset, current_dts, chunk_size, 0, AVINDEX_KEYFRAME);
/* get chunk duration */
chunk_duration = 0;
while (chunk_samples > 0) {
if (chunk_samples < sc->stts_data[stts_index].count) {
chunk_duration += sc->stts_data[stts_index].duration * chunk_samples;
sc->stts_data[stts_index].count -= chunk_samples;
break;
} else {
chunk_duration += sc->stts_data[stts_index].duration * chunk_samples;
chunk_samples -= sc->stts_data[stts_index].count;
if (stts_index + 1 < sc->stts_count) {
stts_index++;
}
}
}
dprintf("AVIndex stream %d, chunk %d, offset %"PRIx64", dts %"PRId64", size %d, duration %d\n",
st->index, i, current_offset, current_dts, chunk_size, chunk_duration);
assert(chunk_duration % sc->time_rate == 0);
current_dts += chunk_duration / sc->time_rate;
}
}
out:
/* adjust sample count to avindex entries */
sc->sample_count = st->nb_index_entries;
}
static int mov_read_header(AVFormatContext *s, AVFormatParameters *ap)
{
MOVContext *mov = (MOVContext *) s->priv_data;
ByteIOContext *pb = &s->pb;
int i, err;
MOV_atom_t atom = { 0, 0, 0 };
mov->fc = s;
mov->parse_table = mov_default_parse_table;
if(!url_is_streamed(pb)) /* .mov and .mp4 aren't streamable anyway (only progressive download if moov is before mdat) */
atom.size = url_fsize(pb);
else
atom.size = 0x7FFFFFFFFFFFFFFFLL;
/* check MOV header */
err = mov_read_default(mov, pb, atom);
if (err<0 || (!mov->found_moov && !mov->found_mdat)) {
av_log(s, AV_LOG_ERROR, "mov: header not found !!! (err:%d, moov:%d, mdat:%d) pos:%"PRId64"\n",
err, mov->found_moov, mov->found_mdat, url_ftell(pb));
return -1;
}
dprintf("on_parse_exit_offset=%d\n", (int) url_ftell(pb));
/* some cleanup : make sure we are on the mdat atom */
if(!url_is_streamed(pb) && (url_ftell(pb) != mov->mdat_offset))
url_fseek(pb, mov->mdat_offset, SEEK_SET);
mov->total_streams = s->nb_streams;
for(i=0; i<mov->total_streams; i++) {
MOVStreamContext *sc = mov->streams[i];
/* sanity checks */
if(!sc->stts_count || !sc->chunk_count || !sc->sample_to_chunk_sz ||
(!sc->sample_size && !sc->sample_count)){
av_log(s, AV_LOG_ERROR, "missing mandatory atoms, broken header\n");
return -1;
}
if(!sc->time_rate)
sc->time_rate=1;
if(!sc->time_scale)
sc->time_scale= mov->time_scale;
av_set_pts_info(s->streams[i], 64, sc->time_rate, sc->time_scale);
if(s->streams[i]->duration != AV_NOPTS_VALUE){
assert(s->streams[i]->duration % sc->time_rate == 0);
s->streams[i]->duration /= sc->time_rate;
}
sc->ffindex = i;
mov_build_index(mov, s->streams[i]);
}
for(i=0; i<mov->total_streams; i++) {
/* dont need those anymore */
av_freep(&mov->streams[i]->chunk_offsets);
av_freep(&mov->streams[i]->sample_to_chunk);
av_freep(&mov->streams[i]->sample_sizes);
av_freep(&mov->streams[i]->keyframes);
av_freep(&mov->streams[i]->stts_data);
}
av_freep(&mov->mdat_list);
return 0;
}
static int mov_read_packet(AVFormatContext *s, AVPacket *pkt)
{
MOVContext *mov = s->priv_data;
MOVStreamContext *sc = 0;
AVIndexEntry *sample = 0;
int64_t best_dts = INT64_MAX;
int i;
for (i = 0; i < mov->total_streams; i++) {
MOVStreamContext *msc = mov->streams[i];
if (s->streams[i]->discard != AVDISCARD_ALL && msc->current_sample < msc->sample_count) {
AVIndexEntry *current_sample = &s->streams[i]->index_entries[msc->current_sample];
int64_t dts = av_rescale(current_sample->timestamp * (int64_t)msc->time_rate, AV_TIME_BASE, msc->time_scale);
dprintf("stream %d, sample %ld, dts %"PRId64"\n", i, msc->current_sample, dts);
if (dts < best_dts) {
sample = current_sample;
best_dts = dts;
sc = msc;
}
}
}
if (!sample)
return -1;
/* must be done just before reading, to avoid infinite loop on sample */
sc->current_sample++;
if (sample->pos >= url_fsize(&s->pb)) {
av_log(mov->fc, AV_LOG_ERROR, "stream %d, offset 0x%"PRIx64": partial file\n", sc->ffindex, sample->pos);
return -1;
}
#ifdef CONFIG_DV_DEMUXER
if (sc->dv_audio_container) {
dv_get_packet(mov->dv_demux, pkt);
dprintf("dv audio pkt size %d\n", pkt->size);
} else {
#endif
url_fseek(&s->pb, sample->pos, SEEK_SET);
av_get_packet(&s->pb, pkt, sample->size);
#ifdef CONFIG_DV_DEMUXER
if (mov->dv_demux) {
void *pkt_destruct_func = pkt->destruct;
dv_produce_packet(mov->dv_demux, pkt, pkt->data, pkt->size);
pkt->destruct = pkt_destruct_func;
}
}
#endif
pkt->stream_index = sc->ffindex;
pkt->dts = sample->timestamp;
if (sc->ctts_data) {
assert(sc->ctts_data[sc->sample_to_ctime_index].duration % sc->time_rate == 0);
pkt->pts = pkt->dts + sc->ctts_data[sc->sample_to_ctime_index].duration / sc->time_rate;
/* update ctts context */
sc->sample_to_ctime_sample++;
if (sc->sample_to_ctime_index < sc->ctts_count && sc->ctts_data[sc->sample_to_ctime_index].count == sc->sample_to_ctime_sample) {
sc->sample_to_ctime_index++;
sc->sample_to_ctime_sample = 0;
}
} else {
pkt->pts = pkt->dts;
}
pkt->flags |= sample->flags & AVINDEX_KEYFRAME ? PKT_FLAG_KEY : 0;
pkt->pos = sample->pos;
dprintf("stream %d, pts %"PRId64", dts %"PRId64", pos 0x%"PRIx64", duration %d\n", pkt->stream_index, pkt->pts, pkt->dts, pkt->pos, pkt->duration);
return 0;
}
static int mov_seek_stream(AVStream *st, int64_t timestamp, int flags)
{
MOVStreamContext *sc = st->priv_data;
int sample, time_sample;
int i;
sample = av_index_search_timestamp(st, timestamp, flags);
dprintf("stream %d, timestamp %"PRId64", sample %d\n", st->index, timestamp, sample);
if (sample < 0) /* not sure what to do */
return -1;
sc->current_sample = sample;
dprintf("stream %d, found sample %ld\n", st->index, sc->current_sample);
/* adjust ctts index */
if (sc->ctts_data) {
time_sample = 0;
for (i = 0; i < sc->ctts_count; i++) {
time_sample += sc->ctts_data[i].count;
if (time_sample >= sc->current_sample) {
sc->sample_to_ctime_index = i;
sc->sample_to_ctime_sample = time_sample - sc->current_sample;
break;
}
}
}
return sample;
}
static int mov_read_seek(AVFormatContext *s, int stream_index, int64_t sample_time, int flags)
{
AVStream *st;
int64_t seek_timestamp, timestamp;
int sample;
int i;
if (stream_index >= s->nb_streams)
return -1;
st = s->streams[stream_index];
sample = mov_seek_stream(st, sample_time, flags);
if (sample < 0)
return -1;
/* adjust seek timestamp to found sample timestamp */
seek_timestamp = st->index_entries[sample].timestamp;
for (i = 0; i < s->nb_streams; i++) {
st = s->streams[i];
if (stream_index == i || st->discard == AVDISCARD_ALL)
continue;
timestamp = av_rescale_q(seek_timestamp, s->streams[stream_index]->time_base, st->time_base);
mov_seek_stream(st, timestamp, flags);
}
return 0;
}
static int mov_read_close(AVFormatContext *s)
{
int i;
MOVContext *mov = (MOVContext *) s->priv_data;
for(i=0; i<mov->total_streams; i++)
mov_free_stream_context(mov->streams[i]);
/* free color tabs */
for(i=0; i<mov->ctab_size; i++)
av_freep(&mov->ctab[i]);
if(mov->dv_demux){
for(i=0; i<mov->dv_fctx->nb_streams; i++){
av_freep(&mov->dv_fctx->streams[i]->codec);
av_freep(&mov->dv_fctx->streams[i]);
}
av_freep(&mov->dv_fctx);
av_freep(&mov->dv_demux);
}
av_freep(&mov->ctab);
return 0;
}
AVInputFormat mov_demuxer = {
"mov,mp4,m4a,3gp,3g2,mj2",
"QuickTime/MPEG4/Motion JPEG 2000 format",
sizeof(MOVContext),
mov_probe,
mov_read_header,
mov_read_packet,
mov_read_close,
mov_read_seek,
};