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.

1974 lines
69 KiB

/*
* AVI 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 <inttypes.h>
#include "libavutil/avassert.h"
#include "libavutil/avstring.h"
#include "libavutil/opt.h"
#include "libavutil/dict.h"
#include "libavutil/internal.h"
#include "libavutil/intreadwrite.h"
#include "libavutil/mathematics.h"
#include "avformat.h"
#include "avi.h"
#include "dv.h"
#include "internal.h"
#include "isom.h"
#include "riff.h"
#include "libavcodec/bytestream.h"
#include "libavcodec/exif.h"
#include "libavcodec/internal.h"
typedef struct AVIStream {
int64_t frame_offset; /* current frame (video) or byte (audio) counter
* (used to compute the pts) */
int remaining;
int packet_size;
uint32_t handler;
uint32_t scale;
uint32_t rate;
int sample_size; /* size of one sample (or packet)
* (in the rate/scale sense) in bytes */
int64_t cum_len; /* temporary storage (used during seek) */
int prefix; /* normally 'd'<<8 + 'c' or 'w'<<8 + 'b' */
int prefix_count;
uint32_t pal[256];
int has_pal;
int dshow_block_align; /* block align variable used to emulate bugs in
* the MS dshow demuxer */
AVFormatContext *sub_ctx;
AVPacket sub_pkt;
AVBufferRef *sub_buffer;
int64_t seek_pos;
} AVIStream;
typedef struct AVIContext {
const AVClass *class;
int64_t riff_end;
int64_t movi_end;
int64_t fsize;
int64_t io_fsize;
int64_t movi_list;
int64_t last_pkt_pos;
int index_loaded;
int is_odml;
int non_interleaved;
int stream_index;
DVDemuxContext *dv_demux;
int odml_depth;
int use_odml;
#define MAX_ODML_DEPTH 1000
int64_t dts_max;
} AVIContext;
static const AVOption options[] = {
{ "use_odml", "use odml index", offsetof(AVIContext, use_odml), AV_OPT_TYPE_BOOL, {.i64 = 1}, -1, 1, AV_OPT_FLAG_DECODING_PARAM},
{ NULL },
};
static const AVClass demuxer_class = {
.class_name = "avi",
.item_name = av_default_item_name,
.option = options,
.version = LIBAVUTIL_VERSION_INT,
.category = AV_CLASS_CATEGORY_DEMUXER,
};
static const char avi_headers[][8] = {
{ 'R', 'I', 'F', 'F', 'A', 'V', 'I', ' ' },
{ 'R', 'I', 'F', 'F', 'A', 'V', 'I', 'X' },
{ 'R', 'I', 'F', 'F', 'A', 'V', 'I', 0x19 },
{ 'O', 'N', '2', ' ', 'O', 'N', '2', 'f' },
{ 'R', 'I', 'F', 'F', 'A', 'M', 'V', ' ' },
{ 0 }
};
static const AVMetadataConv avi_metadata_conv[] = {
{ "strn", "title" },
{ "isbj", "subject" },
{ "inam", "title" },
{ "iart", "artist" },
{ "icop", "copyright" },
{ "icmt", "comment" },
{ "ignr", "genre" },
{ "iprd", "product" },
{ "isft", "software" },
{ 0 },
};
static int avi_read_close(AVFormatContext *s);
static int avi_load_index(AVFormatContext *s);
static int guess_ni_flag(AVFormatContext *s);
#define print_tag(s, str, tag, size) \
av_log(s, AV_LOG_TRACE, "pos:%"PRIX64" %s: tag=%s size=0x%x\n", \
avio_tell(pb), str, av_fourcc2str(tag), size) \
static inline int get_duration(AVIStream *ast, int len)
{
if (ast->sample_size)
return len;
else if (ast->dshow_block_align)
return (len + (int64_t)ast->dshow_block_align - 1) / ast->dshow_block_align;
else
return 1;
}
static int get_riff(AVFormatContext *s, AVIOContext *pb)
{
AVIContext *avi = s->priv_data;
char header[8] = {0};
int i;
/* check RIFF header */
avio_read(pb, header, 4);
avi->riff_end = avio_rl32(pb); /* RIFF chunk size */
avi->riff_end += avio_tell(pb); /* RIFF chunk end */
avio_read(pb, header + 4, 4);
for (i = 0; avi_headers[i][0]; i++)
if (!memcmp(header, avi_headers[i], 8))
break;
if (!avi_headers[i][0])
return AVERROR_INVALIDDATA;
if (header[7] == 0x19)
av_log(s, AV_LOG_INFO,
"This file has been generated by a totally broken muxer.\n");
return 0;
}
static int read_odml_index(AVFormatContext *s, int frame_num)
{
AVIContext *avi = s->priv_data;
AVIOContext *pb = s->pb;
int longs_per_entry = avio_rl16(pb);
int index_sub_type = avio_r8(pb);
int index_type = avio_r8(pb);
int entries_in_use = avio_rl32(pb);
int chunk_id = avio_rl32(pb);
int64_t base = avio_rl64(pb);
int stream_id = ((chunk_id & 0xFF) - '0') * 10 +
((chunk_id >> 8 & 0xFF) - '0');
AVStream *st;
AVIStream *ast;
int i;
int64_t last_pos = -1;
int64_t filesize = avi->fsize;
av_log(s, AV_LOG_TRACE,
"longs_per_entry:%d index_type:%d entries_in_use:%d "
"chunk_id:%X base:%16"PRIX64" frame_num:%d\n",
longs_per_entry,
index_type,
entries_in_use,
chunk_id,
base,
frame_num);
if (stream_id >= s->nb_streams || stream_id < 0)
return AVERROR_INVALIDDATA;
st = s->streams[stream_id];
ast = st->priv_data;
if (index_sub_type)
return AVERROR_INVALIDDATA;
avio_rl32(pb);
if (index_type && longs_per_entry != 2)
return AVERROR_INVALIDDATA;
if (index_type > 1)
return AVERROR_INVALIDDATA;
if (filesize > 0 && base >= filesize) {
av_log(s, AV_LOG_ERROR, "ODML index invalid\n");
if (base >> 32 == (base & 0xFFFFFFFF) &&
(base & 0xFFFFFFFF) < filesize &&
filesize <= 0xFFFFFFFF)
base &= 0xFFFFFFFF;
else
return AVERROR_INVALIDDATA;
}
for (i = 0; i < entries_in_use; i++) {
if (index_type) {
int64_t pos = avio_rl32(pb) + base - 8;
int len = avio_rl32(pb);
int key = len >= 0;
len &= 0x7FFFFFFF;
av_log(s, AV_LOG_TRACE, "pos:%"PRId64", len:%X\n", pos, len);
if (avio_feof(pb))
return AVERROR_INVALIDDATA;
if (last_pos == pos || pos == base - 8)
avi->non_interleaved = 1;
if (last_pos != pos && len)
av_add_index_entry(st, pos, ast->cum_len, len, 0,
key ? AVINDEX_KEYFRAME : 0);
ast->cum_len += get_duration(ast, len);
last_pos = pos;
} else {
int64_t offset, pos;
int duration;
offset = avio_rl64(pb);
avio_rl32(pb); /* size */
duration = avio_rl32(pb);
if (avio_feof(pb))
return AVERROR_INVALIDDATA;
pos = avio_tell(pb);
if (avi->odml_depth > MAX_ODML_DEPTH) {
av_log(s, AV_LOG_ERROR, "Too deeply nested ODML indexes\n");
return AVERROR_INVALIDDATA;
}
if (avio_seek(pb, offset + 8, SEEK_SET) < 0)
return -1;
avi->odml_depth++;
read_odml_index(s, frame_num);
avi->odml_depth--;
frame_num += duration;
if (avio_seek(pb, pos, SEEK_SET) < 0) {
av_log(s, AV_LOG_ERROR, "Failed to restore position after reading index\n");
return -1;
}
}
}
avi->index_loaded = 2;
return 0;
}
static void clean_index(AVFormatContext *s)
{
int i;
int64_t j;
for (i = 0; i < s->nb_streams; i++) {
AVStream *st = s->streams[i];
AVIStream *ast = st->priv_data;
int n = st->internal->nb_index_entries;
int max = ast->sample_size;
int64_t pos, size, ts;
if (n != 1 || ast->sample_size == 0)
continue;
while (max < 1024)
max += max;
pos = st->internal->index_entries[0].pos;
size = st->internal->index_entries[0].size;
ts = st->internal->index_entries[0].timestamp;
for (j = 0; j < size; j += max)
av_add_index_entry(st, pos + j, ts + j, FFMIN(max, size - j), 0,
AVINDEX_KEYFRAME);
}
}
static int avi_read_tag(AVFormatContext *s, AVStream *st, uint32_t tag,
uint32_t size)
{
AVIOContext *pb = s->pb;
char key[5] = { 0 };
char *value;
size += (size & 1);
if (size == UINT_MAX)
return AVERROR(EINVAL);
value = av_malloc(size + 1);
if (!value)
return AVERROR(ENOMEM);
if (avio_read(pb, value, size) != size) {
av_freep(&value);
return AVERROR_INVALIDDATA;
}
value[size] = 0;
AV_WL32(key, tag);
return av_dict_set(st ? &st->metadata : &s->metadata, key, value,
AV_DICT_DONT_STRDUP_VAL);
}
static const char months[12][4] = { "Jan", "Feb", "Mar", "Apr", "May", "Jun",
"Jul", "Aug", "Sep", "Oct", "Nov", "Dec" };
static void avi_metadata_creation_time(AVDictionary **metadata, char *date)
{
char month[4], time[9], buffer[64];
int i, day, year;
/* parse standard AVI date format (ie. "Mon Mar 10 15:04:43 2003") */
if (sscanf(date, "%*3s%*[ ]%3s%*[ ]%2d%*[ ]%8s%*[ ]%4d",
month, &day, time, &year) == 4) {
for (i = 0; i < 12; i++)
if (!av_strcasecmp(month, months[i])) {
snprintf(buffer, sizeof(buffer), "%.4d-%.2d-%.2d %s",
year, i + 1, day, time);
av_dict_set(metadata, "creation_time", buffer, 0);
}
} else if (date[4] == '/' && date[7] == '/') {
date[4] = date[7] = '-';
av_dict_set(metadata, "creation_time", date, 0);
}
}
static void avi_read_nikon(AVFormatContext *s, uint64_t end)
{
while (avio_tell(s->pb) < end && !avio_feof(s->pb)) {
uint32_t tag = avio_rl32(s->pb);
uint32_t size = avio_rl32(s->pb);
switch (tag) {
case MKTAG('n', 'c', 't', 'g'): /* Nikon Tags */
{
uint64_t tag_end = avio_tell(s->pb) + size;
while (avio_tell(s->pb) < tag_end && !avio_feof(s->pb)) {
uint16_t tag = avio_rl16(s->pb);
uint16_t size = avio_rl16(s->pb);
const char *name = NULL;
char buffer[64] = { 0 };
size = FFMIN(size, tag_end - avio_tell(s->pb));
size -= avio_read(s->pb, buffer,
FFMIN(size, sizeof(buffer) - 1));
switch (tag) {
case 0x03:
name = "maker";
break;
case 0x04:
name = "model";
break;
case 0x13:
name = "creation_time";
if (buffer[4] == ':' && buffer[7] == ':')
buffer[4] = buffer[7] = '-';
break;
}
if (name)
av_dict_set(&s->metadata, name, buffer, 0);
avio_skip(s->pb, size);
}
break;
}
default:
avio_skip(s->pb, size);
break;
}
}
}
static int avi_extract_stream_metadata(AVFormatContext *s, AVStream *st)
{
GetByteContext gb;
uint8_t *data = st->codecpar->extradata;
int data_size = st->codecpar->extradata_size;
int tag, offset;
if (!data || data_size < 8) {
return AVERROR_INVALIDDATA;
}
bytestream2_init(&gb, data, data_size);
tag = bytestream2_get_le32(&gb);
switch (tag) {
case MKTAG('A', 'V', 'I', 'F'):
// skip 4 byte padding
bytestream2_skip(&gb, 4);
offset = bytestream2_tell(&gb);
// decode EXIF tags from IFD, AVI is always little-endian
return avpriv_exif_decode_ifd(s, data + offset, data_size - offset,
1, 0, &st->metadata);
break;
case MKTAG('C', 'A', 'S', 'I'):
avpriv_request_sample(s, "RIFF stream data tag type CASI (%u)", tag);
break;
case MKTAG('Z', 'o', 'r', 'a'):
avpriv_request_sample(s, "RIFF stream data tag type Zora (%u)", tag);
break;
default:
break;
}
return 0;
}
static int calculate_bitrate(AVFormatContext *s)
{
AVIContext *avi = s->priv_data;
int i, j;
int64_t lensum = 0;
int64_t maxpos = 0;
for (i = 0; i<s->nb_streams; i++) {
int64_t len = 0;
AVStream *st = s->streams[i];
if (!st->internal->nb_index_entries)
continue;
for (j = 0; j < st->internal->nb_index_entries; j++)
len += st->internal->index_entries[j].size;
maxpos = FFMAX(maxpos, st->internal->index_entries[j-1].pos);
lensum += len;
}
if (maxpos < av_rescale(avi->io_fsize, 9, 10)) // index does not cover the whole file
return 0;
if (lensum*9/10 > maxpos || lensum < maxpos*9/10) // frame sum and filesize mismatch
return 0;
for (i = 0; i<s->nb_streams; i++) {
int64_t len = 0;
AVStream *st = s->streams[i];
int64_t duration;
int64_t bitrate;
for (j = 0; j < st->internal->nb_index_entries; j++)
len += st->internal->index_entries[j].size;
if (st->internal->nb_index_entries < 2 || st->codecpar->bit_rate > 0)
continue;
duration = st->internal->index_entries[j-1].timestamp - st->internal->index_entries[0].timestamp;
bitrate = av_rescale(8*len, st->time_base.den, duration * st->time_base.num);
if (bitrate > 0) {
st->codecpar->bit_rate = bitrate;
}
}
return 1;
}
#define RETURN_ERROR(code) do { ret = (code); goto fail; } while (0)
static int avi_read_header(AVFormatContext *s)
{
AVIContext *avi = s->priv_data;
AVIOContext *pb = s->pb;
unsigned int tag, tag1, handler;
int codec_type, stream_index, frame_period;
unsigned int size;
int i;
AVStream *st;
AVIStream *ast = NULL;
int avih_width = 0, avih_height = 0;
int amv_file_format = 0;
uint64_t list_end = 0;
int64_t pos;
int ret;
AVDictionaryEntry *dict_entry;
avi->stream_index = -1;
ret = get_riff(s, pb);
if (ret < 0)
return ret;
av_log(avi, AV_LOG_DEBUG, "use odml:%d\n", avi->use_odml);
avi->io_fsize = avi->fsize = avio_size(pb);
if (avi->fsize <= 0 || avi->fsize < avi->riff_end)
avi->fsize = avi->riff_end == 8 ? INT64_MAX : avi->riff_end;
/* first list tag */
stream_index = -1;
codec_type = -1;
frame_period = 0;
for (;;) {
if (avio_feof(pb))
RETURN_ERROR(AVERROR_INVALIDDATA);
tag = avio_rl32(pb);
size = avio_rl32(pb);
print_tag(s, "tag", tag, size);
switch (tag) {
case MKTAG('L', 'I', 'S', 'T'):
list_end = avio_tell(pb) + size;
/* Ignored, except at start of video packets. */
tag1 = avio_rl32(pb);
print_tag(s, "list", tag1, 0);
if (tag1 == MKTAG('m', 'o', 'v', 'i')) {
avi->movi_list = avio_tell(pb) - 4;
if (size)
avi->movi_end = avi->movi_list + size + (size & 1);
else
avi->movi_end = avi->fsize;
av_log(s, AV_LOG_TRACE, "movi end=%"PRIx64"\n", avi->movi_end);
goto end_of_header;
} else if (tag1 == MKTAG('I', 'N', 'F', 'O'))
ff_read_riff_info(s, size - 4);
else if (tag1 == MKTAG('n', 'c', 'd', 't'))
avi_read_nikon(s, list_end);
break;
case MKTAG('I', 'D', 'I', 'T'):
{
unsigned char date[64] = { 0 };
size += (size & 1);
size -= avio_read(pb, date, FFMIN(size, sizeof(date) - 1));
avio_skip(pb, size);
avi_metadata_creation_time(&s->metadata, date);
break;
}
case MKTAG('d', 'm', 'l', 'h'):
avi->is_odml = 1;
avio_skip(pb, size + (size & 1));
break;
case MKTAG('a', 'm', 'v', 'h'):
amv_file_format = 1;
case MKTAG('a', 'v', 'i', 'h'):
/* AVI header */
/* using frame_period is bad idea */
frame_period = avio_rl32(pb);
avio_rl32(pb); /* max. bytes per second */
avio_rl32(pb);
avi->non_interleaved |= avio_rl32(pb) & AVIF_MUSTUSEINDEX;
avio_skip(pb, 2 * 4);
avio_rl32(pb);
avio_rl32(pb);
avih_width = avio_rl32(pb);
avih_height = avio_rl32(pb);
avio_skip(pb, size - 10 * 4);
break;
case MKTAG('s', 't', 'r', 'h'):
/* stream header */
tag1 = avio_rl32(pb);
handler = avio_rl32(pb); /* codec tag */
if (tag1 == MKTAG('p', 'a', 'd', 's')) {
avio_skip(pb, size - 8);
break;
} else {
stream_index++;
st = avformat_new_stream(s, NULL);
if (!st)
RETURN_ERROR(AVERROR(ENOMEM));
st->id = stream_index;
ast = av_mallocz(sizeof(AVIStream));
if (!ast)
RETURN_ERROR(AVERROR(ENOMEM));
st->priv_data = ast;
}
if (amv_file_format)
tag1 = stream_index ? MKTAG('a', 'u', 'd', 's')
: MKTAG('v', 'i', 'd', 's');
print_tag(s, "strh", tag1, -1);
if (tag1 == MKTAG('i', 'a', 'v', 's') ||
tag1 == MKTAG('i', 'v', 'a', 's')) {
int64_t dv_dur;
/* After some consideration -- I don't think we
* have to support anything but DV in type1 AVIs. */
if (s->nb_streams != 1)
RETURN_ERROR(AVERROR_INVALIDDATA);
if (handler != MKTAG('d', 'v', 's', 'd') &&
handler != MKTAG('d', 'v', 'h', 'd') &&
handler != MKTAG('d', 'v', 's', 'l'))
return AVERROR_INVALIDDATA;
if (!CONFIG_DV_DEMUXER)
return AVERROR_DEMUXER_NOT_FOUND;
ast = s->streams[0]->priv_data;
st->priv_data = NULL;
ff_free_stream(s, st);
avi->dv_demux = avpriv_dv_init_demux(s);
if (!avi->dv_demux) {
av_free(ast);
return AVERROR(ENOMEM);
}
s->streams[0]->priv_data = ast;
avio_skip(pb, 3 * 4);
ast->scale = avio_rl32(pb);
ast->rate = avio_rl32(pb);
avio_skip(pb, 4); /* start time */
dv_dur = avio_rl32(pb);
if (ast->scale > 0 && ast->rate > 0 && dv_dur > 0) {
dv_dur *= AV_TIME_BASE;
s->duration = av_rescale(dv_dur, ast->scale, ast->rate);
}
/* else, leave duration alone; timing estimation in utils.c
* will make a guess based on bitrate. */
stream_index = s->nb_streams - 1;
avio_skip(pb, size - 9 * 4);
break;
}
av_assert0(stream_index < s->nb_streams);
ast->handler = handler;
avio_rl32(pb); /* flags */
avio_rl16(pb); /* priority */
avio_rl16(pb); /* language */
avio_rl32(pb); /* initial frame */
ast->scale = avio_rl32(pb);
ast->rate = avio_rl32(pb);
if (!(ast->scale && ast->rate)) {
av_log(s, AV_LOG_WARNING,
"scale/rate is %"PRIu32"/%"PRIu32" which is invalid. "
"(This file has been generated by broken software.)\n",
ast->scale,
ast->rate);
if (frame_period) {
ast->rate = 1000000;
ast->scale = frame_period;
} else {
ast->rate = 25;
ast->scale = 1;
}
}
avpriv_set_pts_info(st, 64, ast->scale, ast->rate);
ast->cum_len = avio_rl32(pb); /* start */
st->nb_frames = avio_rl32(pb);
st->start_time = 0;
avio_rl32(pb); /* buffer size */
avio_rl32(pb); /* quality */
if (ast->cum_len > 3600LL * ast->rate / ast->scale) {
av_log(s, AV_LOG_ERROR, "crazy start time, iam scared, giving up\n");
ast->cum_len = 0;
}
ast->sample_size = avio_rl32(pb);
ast->cum_len *= FFMAX(1, ast->sample_size);
av_log(s, AV_LOG_TRACE, "%"PRIu32" %"PRIu32" %d\n",
ast->rate, ast->scale, ast->sample_size);
switch (tag1) {
case MKTAG('v', 'i', 'd', 's'):
codec_type = AVMEDIA_TYPE_VIDEO;
ast->sample_size = 0;
st->avg_frame_rate = av_inv_q(st->time_base);
break;
case MKTAG('a', 'u', 'd', 's'):
codec_type = AVMEDIA_TYPE_AUDIO;
break;
case MKTAG('t', 'x', 't', 's'):
codec_type = AVMEDIA_TYPE_SUBTITLE;
break;
case MKTAG('d', 'a', 't', 's'):
codec_type = AVMEDIA_TYPE_DATA;
break;
default:
av_log(s, AV_LOG_INFO, "unknown stream type %X\n", tag1);
}
if (ast->sample_size < 0) {
if (s->error_recognition & AV_EF_EXPLODE) {
av_log(s, AV_LOG_ERROR,
"Invalid sample_size %d at stream %d\n",
ast->sample_size,
stream_index);
RETURN_ERROR(AVERROR_INVALIDDATA);
}
av_log(s, AV_LOG_WARNING,
"Invalid sample_size %d at stream %d "
"setting it to 0\n",
ast->sample_size,
stream_index);
ast->sample_size = 0;
}
if (ast->sample_size == 0) {
st->duration = st->nb_frames;
if (st->duration > 0 && avi->io_fsize > 0 && avi->riff_end > avi->io_fsize) {
av_log(s, AV_LOG_DEBUG, "File is truncated adjusting duration\n");
st->duration = av_rescale(st->duration, avi->io_fsize, avi->riff_end);
}
}
ast->frame_offset = ast->cum_len;
avio_skip(pb, size - 12 * 4);
break;
case MKTAG('s', 't', 'r', 'f'):
/* stream header */
if (!size && (codec_type == AVMEDIA_TYPE_AUDIO ||
codec_type == AVMEDIA_TYPE_VIDEO))
break;
if (stream_index >= (unsigned)s->nb_streams || avi->dv_demux) {
avio_skip(pb, size);
} else {
uint64_t cur_pos = avio_tell(pb);
unsigned esize;
if (cur_pos < list_end)
size = FFMIN(size, list_end - cur_pos);
st = s->streams[stream_index];
if (st->codecpar->codec_type != AVMEDIA_TYPE_UNKNOWN) {
avio_skip(pb, size);
break;
}
switch (codec_type) {
case AVMEDIA_TYPE_VIDEO:
if (amv_file_format) {
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
st->codecpar->width = avih_width;
st->codecpar->height = avih_height;
st->codecpar->codec_type = AVMEDIA_TYPE_VIDEO;
st->codecpar->codec_id = AV_CODEC_ID_AMV;
avio_skip(pb, size);
break;
}
tag1 = ff_get_bmp_header(pb, st, &esize);
if (tag1 == MKTAG('D', 'X', 'S', 'B') ||
tag1 == MKTAG('D', 'X', 'S', 'A')) {
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
st->codecpar->codec_type = AVMEDIA_TYPE_SUBTITLE;
st->codecpar->codec_tag = tag1;
st->codecpar->codec_id = AV_CODEC_ID_XSUB;
break;
}
if (size > 10 * 4 && size < (1 << 30) && size < avi->fsize) {
if (esize == size-1 && (esize&1)) {
st->codecpar->extradata_size = esize - 10 * 4;
} else
st->codecpar->extradata_size = size - 10 * 4;
if (st->codecpar->extradata) {
av_log(s, AV_LOG_WARNING, "New extradata in strf chunk, freeing previous one.\n");
}
ret = ff_get_extradata(s, st->codecpar, pb,
st->codecpar->extradata_size);
if (ret < 0)
return ret;
}
// FIXME: check if the encoder really did this correctly
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
if (st->codecpar->extradata_size & 1)
avio_r8(pb);
/* Extract palette from extradata if bpp <= 8.
* This code assumes that extradata contains only palette.
* This is true for all paletted codecs implemented in
* FFmpeg. */
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
if (st->codecpar->extradata_size &&
(st->codecpar->bits_per_coded_sample <= 8)) {
int pal_size = (1 << st->codecpar->bits_per_coded_sample) << 2;
const uint8_t *pal_src;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
pal_size = FFMIN(pal_size, st->codecpar->extradata_size);
pal_src = st->codecpar->extradata +
st->codecpar->extradata_size - pal_size;
/* Exclude the "BottomUp" field from the palette */
if (pal_src - st->codecpar->extradata >= 9 &&
!memcmp(st->codecpar->extradata + st->codecpar->extradata_size - 9, "BottomUp", 9))
pal_src -= 9;
for (i = 0; i < pal_size / 4; i++)
ast->pal[i] = 0xFFU<<24 | AV_RL32(pal_src + 4 * i);
ast->has_pal = 1;
}
print_tag(s, "video", tag1, 0);
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
st->codecpar->codec_type = AVMEDIA_TYPE_VIDEO;
st->codecpar->codec_tag = tag1;
st->codecpar->codec_id = ff_codec_get_id(ff_codec_bmp_tags,
tag1);
/* If codec is not found yet, try with the mov tags. */
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
if (!st->codecpar->codec_id) {
st->codecpar->codec_id =
ff_codec_get_id(ff_codec_movvideo_tags, tag1);
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
if (st->codecpar->codec_id)
av_log(s, AV_LOG_WARNING,
"mov tag found in avi (fourcc %s)\n",
av_fourcc2str(tag1));
}
if (!st->codecpar->codec_id)
st->codecpar->codec_id = ff_codec_get_id(ff_codec_bmp_tags_unofficial, tag1);
/* This is needed to get the pict type which is necessary
* for generating correct pts. */
st->need_parsing = AVSTREAM_PARSE_HEADERS;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
if (st->codecpar->codec_id == AV_CODEC_ID_MPEG4 &&
ast->handler == MKTAG('X', 'V', 'I', 'D'))
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
st->codecpar->codec_tag = MKTAG('X', 'V', 'I', 'D');
if (st->codecpar->codec_tag == MKTAG('V', 'S', 'S', 'H'))
st->need_parsing = AVSTREAM_PARSE_FULL;
if (st->codecpar->codec_id == AV_CODEC_ID_RV40)
st->need_parsing = AVSTREAM_PARSE_NONE;
if (st->codecpar->codec_id == AV_CODEC_ID_HEVC &&
st->codecpar->codec_tag == MKTAG('H', '2', '6', '5'))
st->need_parsing = AVSTREAM_PARSE_FULL;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
if (st->codecpar->codec_tag == 0 && st->codecpar->height > 0 &&
st->codecpar->extradata_size < 1U << 30) {
st->codecpar->extradata_size += 9;
if ((ret = av_reallocp(&st->codecpar->extradata,
st->codecpar->extradata_size +
AV_INPUT_BUFFER_PADDING_SIZE)) < 0) {
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
st->codecpar->extradata_size = 0;
return ret;
} else
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
memcpy(st->codecpar->extradata + st->codecpar->extradata_size - 9,
"BottomUp", 9);
}
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
st->codecpar->height = FFABS(st->codecpar->height);
// avio_skip(pb, size - 5 * 4);
break;
case AVMEDIA_TYPE_AUDIO:
ret = ff_get_wav_header(s, pb, st->codecpar, size, 0);
if (ret < 0)
return ret;
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
ast->dshow_block_align = st->codecpar->block_align;
if (ast->sample_size && st->codecpar->block_align &&
ast->sample_size != st->codecpar->block_align) {
av_log(s,
AV_LOG_WARNING,
"sample size (%d) != block align (%d)\n",
ast->sample_size,
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
st->codecpar->block_align);
ast->sample_size = st->codecpar->block_align;
}
/* 2-aligned
* (fix for Stargate SG-1 - 3x18 - Shades of Grey.avi) */
if (size & 1)
avio_skip(pb, 1);
/* Force parsing as several audio frames can be in
* one packet and timestamps refer to packet start. */
st->need_parsing = AVSTREAM_PARSE_TIMESTAMPS;
/* ADTS header is in extradata, AAC without header must be
* stored as exact frames. Parser not needed and it will
* fail. */
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
if (st->codecpar->codec_id == AV_CODEC_ID_AAC &&
st->codecpar->extradata_size)
st->need_parsing = AVSTREAM_PARSE_NONE;
// The flac parser does not work with AVSTREAM_PARSE_TIMESTAMPS
if (st->codecpar->codec_id == AV_CODEC_ID_FLAC)
st->need_parsing = AVSTREAM_PARSE_NONE;
/* AVI files with Xan DPCM audio (wrongly) declare PCM
* audio in the header but have Axan as stream_code_tag. */
if (ast->handler == AV_RL32("Axan")) {
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
st->codecpar->codec_id = AV_CODEC_ID_XAN_DPCM;
st->codecpar->codec_tag = 0;
ast->dshow_block_align = 0;
}
if (amv_file_format) {
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
st->codecpar->codec_id = AV_CODEC_ID_ADPCM_IMA_AMV;
ast->dshow_block_align = 0;
}
if ((st->codecpar->codec_id == AV_CODEC_ID_AAC ||
st->codecpar->codec_id == AV_CODEC_ID_FLAC ||
st->codecpar->codec_id == AV_CODEC_ID_MP2 ) && ast->dshow_block_align <= 4 && ast->dshow_block_align) {
av_log(s, AV_LOG_DEBUG, "overriding invalid dshow_block_align of %d\n", ast->dshow_block_align);
ast->dshow_block_align = 0;
}
if (st->codecpar->codec_id == AV_CODEC_ID_AAC && ast->dshow_block_align == 1024 && ast->sample_size == 1024 ||
st->codecpar->codec_id == AV_CODEC_ID_AAC && ast->dshow_block_align == 4096 && ast->sample_size == 4096 ||
st->codecpar->codec_id == AV_CODEC_ID_MP3 && ast->dshow_block_align == 1152 && ast->sample_size == 1152) {
av_log(s, AV_LOG_DEBUG, "overriding sample_size\n");
ast->sample_size = 0;
}
break;
case AVMEDIA_TYPE_SUBTITLE:
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
st->codecpar->codec_type = AVMEDIA_TYPE_SUBTITLE;
st->internal->request_probe= 1;
avio_skip(pb, size);
break;
default:
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
st->codecpar->codec_type = AVMEDIA_TYPE_DATA;
st->codecpar->codec_id = AV_CODEC_ID_NONE;
st->codecpar->codec_tag = 0;
avio_skip(pb, size);
break;
}
}
break;
case MKTAG('s', 't', 'r', 'd'):
if (stream_index >= (unsigned)s->nb_streams
|| s->streams[stream_index]->codecpar->extradata_size
|| s->streams[stream_index]->codecpar->codec_tag == MKTAG('H','2','6','4')) {
avio_skip(pb, size);
} else {
uint64_t cur_pos = avio_tell(pb);
if (cur_pos < list_end)
size = FFMIN(size, list_end - cur_pos);
st = s->streams[stream_index];
if (size<(1<<30)) {
if (st->codecpar->extradata) {
av_log(s, AV_LOG_WARNING, "New extradata in strd chunk, freeing previous one.\n");
}
if ((ret = ff_get_extradata(s, st->codecpar, pb, size)) < 0)
goto fail;
}
if (st->codecpar->extradata_size & 1) //FIXME check if the encoder really did this correctly
avio_r8(pb);
ret = avi_extract_stream_metadata(s, st);
if (ret < 0) {
av_log(s, AV_LOG_WARNING, "could not decoding EXIF data in stream header.\n");
}
}
break;
case MKTAG('i', 'n', 'd', 'x'):
pos = avio_tell(pb);
if ((pb->seekable & AVIO_SEEKABLE_NORMAL) && !(s->flags & AVFMT_FLAG_IGNIDX) &&
avi->use_odml &&
read_odml_index(s, 0) < 0 &&
(s->error_recognition & AV_EF_EXPLODE))
RETURN_ERROR(AVERROR_INVALIDDATA);
avio_seek(pb, pos + size, SEEK_SET);
break;
case MKTAG('v', 'p', 'r', 'p'):
if (stream_index < (unsigned)s->nb_streams && size > 9 * 4) {
AVRational active, active_aspect;
st = s->streams[stream_index];
avio_rl32(pb);
avio_rl32(pb);
avio_rl32(pb);
avio_rl32(pb);
avio_rl32(pb);
active_aspect.den = avio_rl16(pb);
active_aspect.num = avio_rl16(pb);
active.num = avio_rl32(pb);
active.den = avio_rl32(pb);
avio_rl32(pb); // nbFieldsPerFrame
if (active_aspect.num && active_aspect.den &&
active.num && active.den) {
st->sample_aspect_ratio = av_div_q(active_aspect, active);
av_log(s, AV_LOG_TRACE, "vprp %d/%d %d/%d\n",
active_aspect.num, active_aspect.den,
active.num, active.den);
}
size -= 9 * 4;
}
avio_skip(pb, size);
break;
case MKTAG('s', 't', 'r', 'n'):
case MKTAG('i', 's', 'b', 'j'):
case MKTAG('i', 'n', 'a', 'm'):
case MKTAG('i', 'a', 'r', 't'):
case MKTAG('i', 'c', 'o', 'p'):
case MKTAG('i', 'c', 'm', 't'):
case MKTAG('i', 'g', 'n', 'r'):
case MKTAG('i', 'p', 'o', 'd'):
case MKTAG('i', 's', 'o', 'f'):
if (s->nb_streams) {
ret = avi_read_tag(s, s->streams[s->nb_streams - 1], tag, size);
if (ret < 0)
goto fail;
break;
}
default:
if (size > 1000000) {
av_log(s, AV_LOG_ERROR,
"Something went wrong during header parsing, "
"tag %s has size %u, "
"I will ignore it and try to continue anyway.\n",
av_fourcc2str(tag), size);
if (s->error_recognition & AV_EF_EXPLODE)
RETURN_ERROR(AVERROR_INVALIDDATA);
avi->movi_list = avio_tell(pb) - 4;
avi->movi_end = avi->fsize;
goto end_of_header;
}
/* Do not fail for very large idx1 tags */
case MKTAG('i', 'd', 'x', '1'):
/* skip tag */
size += (size & 1);
avio_skip(pb, size);
break;
}
}
end_of_header:
/* check stream number */
if (stream_index != s->nb_streams - 1) {
RETURN_ERROR(AVERROR_INVALIDDATA);
}
if (!avi->index_loaded && (pb->seekable & AVIO_SEEKABLE_NORMAL))
avi_load_index(s);
calculate_bitrate(s);
avi->index_loaded |= 1;
if ((ret = guess_ni_flag(s)) < 0)
goto fail;
avi->non_interleaved |= ret | (s->flags & AVFMT_FLAG_SORT_DTS);
dict_entry = av_dict_get(s->metadata, "ISFT", NULL, 0);
if (dict_entry && !strcmp(dict_entry->value, "PotEncoder"))
for (i = 0; i < s->nb_streams; i++) {
AVStream *st = s->streams[i];
if ( st->codecpar->codec_id == AV_CODEC_ID_MPEG1VIDEO
|| st->codecpar->codec_id == AV_CODEC_ID_MPEG2VIDEO)
st->need_parsing = AVSTREAM_PARSE_FULL;
}
for (i = 0; i < s->nb_streams; i++) {
AVStream *st = s->streams[i];
if (st->internal->nb_index_entries)
break;
}
// DV-in-AVI cannot be non-interleaved, if set this must be
// a mis-detection.
if (avi->dv_demux)
avi->non_interleaved = 0;
if (i == s->nb_streams && avi->non_interleaved) {
av_log(s, AV_LOG_WARNING,
"Non-interleaved AVI without index, switching to interleaved\n");
avi->non_interleaved = 0;
}
if (avi->non_interleaved) {
av_log(s, AV_LOG_INFO, "non-interleaved AVI\n");
clean_index(s);
}
ff_metadata_conv_ctx(s, NULL, avi_metadata_conv);
ff_metadata_conv_ctx(s, NULL, ff_riff_info_conv);
return 0;
fail:
avi_read_close(s);
return ret;
}
static int read_gab2_sub(AVFormatContext *s, AVStream *st, AVPacket *pkt)
{
if (pkt->size >= 7 &&
pkt->size < INT_MAX - AVPROBE_PADDING_SIZE &&
!strcmp(pkt->data, "GAB2") && AV_RL16(pkt->data + 5) == 2) {
uint8_t desc[256];
int score = AVPROBE_SCORE_EXTENSION, ret;
AVIStream *ast = st->priv_data;
ff_const59 AVInputFormat *sub_demuxer;
AVRational time_base;
int size;
AVProbeData pd;
unsigned int desc_len;
AVIOContext *pb = avio_alloc_context(pkt->data + 7,
pkt->size - 7,
0, NULL, NULL, NULL, NULL);
if (!pb)
goto error;
desc_len = avio_rl32(pb);
if (desc_len > pb->buf_end - pb->buf_ptr)
goto error;
ret = avio_get_str16le(pb, desc_len, desc, sizeof(desc));
avio_skip(pb, desc_len - ret);
if (*desc)
av_dict_set(&st->metadata, "title", desc, 0);
avio_rl16(pb); /* flags? */
avio_rl32(pb); /* data size */
size = pb->buf_end - pb->buf_ptr;
pd = (AVProbeData) { .buf = av_mallocz(size + AVPROBE_PADDING_SIZE),
.buf_size = size };
if (!pd.buf)
goto error;
memcpy(pd.buf, pb->buf_ptr, size);
sub_demuxer = av_probe_input_format2(&pd, 1, &score);
av_freep(&pd.buf);
if (!sub_demuxer)
goto error;
if (strcmp(sub_demuxer->name, "srt") && strcmp(sub_demuxer->name, "ass"))
goto error;
if (!(ast->sub_ctx = avformat_alloc_context()))
goto error;
ast->sub_ctx->pb = pb;
if (ff_copy_whiteblacklists(ast->sub_ctx, s) < 0)
goto error;
if (!avformat_open_input(&ast->sub_ctx, "", sub_demuxer, NULL)) {
if (ast->sub_ctx->nb_streams != 1)
goto error;
ff_read_packet(ast->sub_ctx, &ast->sub_pkt);
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
avcodec_parameters_copy(st->codecpar, ast->sub_ctx->streams[0]->codecpar);
time_base = ast->sub_ctx->streams[0]->time_base;
avpriv_set_pts_info(st, 64, time_base.num, time_base.den);
}
ast->sub_buffer = pkt->buf;
pkt->buf = NULL;
av_packet_unref(pkt);
return 1;
error:
av_freep(&ast->sub_ctx);
avio_context_free(&pb);
}
return 0;
}
static AVStream *get_subtitle_pkt(AVFormatContext *s, AVStream *next_st,
AVPacket *pkt)
{
AVIStream *ast, *next_ast = next_st->priv_data;
int64_t ts, next_ts, ts_min = INT64_MAX;
AVStream *st, *sub_st = NULL;
int i;
next_ts = av_rescale_q(next_ast->frame_offset, next_st->time_base,
AV_TIME_BASE_Q);
for (i = 0; i < s->nb_streams; i++) {
st = s->streams[i];
ast = st->priv_data;
if (st->discard < AVDISCARD_ALL && ast && ast->sub_pkt.data) {
ts = av_rescale_q(ast->sub_pkt.dts, st->time_base, AV_TIME_BASE_Q);
if (ts <= next_ts && ts < ts_min) {
ts_min = ts;
sub_st = st;
}
}
}
if (sub_st) {
ast = sub_st->priv_data;
*pkt = ast->sub_pkt;
pkt->stream_index = sub_st->index;
if (ff_read_packet(ast->sub_ctx, &ast->sub_pkt) < 0)
ast->sub_pkt.data = NULL;
}
return sub_st;
}
static int get_stream_idx(const unsigned *d)
{
if (d[0] >= '0' && d[0] <= '9' &&
d[1] >= '0' && d[1] <= '9') {
return (d[0] - '0') * 10 + (d[1] - '0');
} else {
return 100; // invalid stream ID
}
}
/**
*
* @param exit_early set to 1 to just gather packet position without making the changes needed to actually read & return the packet
*/
static int avi_sync(AVFormatContext *s, int exit_early)
{
AVIContext *avi = s->priv_data;
AVIOContext *pb = s->pb;
int n;
unsigned int d[8];
unsigned int size;
int64_t i, sync;
start_sync:
memset(d, -1, sizeof(d));
for (i = sync = avio_tell(pb); !avio_feof(pb); i++) {
int j;
for (j = 0; j < 7; j++)
d[j] = d[j + 1];
d[7] = avio_r8(pb);
size = d[4] + (d[5] << 8) + (d[6] << 16) + (d[7] << 24);
n = get_stream_idx(d + 2);
ff_tlog(s, "%X %X %X %X %X %X %X %X %"PRId64" %u %d\n",
d[0], d[1], d[2], d[3], d[4], d[5], d[6], d[7], i, size, n);
if (i*(avi->io_fsize>0) + (uint64_t)size > avi->fsize || d[0] > 127)
continue;
// parse ix##
if ((d[0] == 'i' && d[1] == 'x' && n < s->nb_streams) ||
// parse JUNK
(d[0] == 'J' && d[1] == 'U' && d[2] == 'N' && d[3] == 'K') ||
(d[0] == 'i' && d[1] == 'd' && d[2] == 'x' && d[3] == '1') ||
(d[0] == 'i' && d[1] == 'n' && d[2] == 'd' && d[3] == 'x')) {
avio_skip(pb, size);
goto start_sync;
}
// parse stray LIST
if (d[0] == 'L' && d[1] == 'I' && d[2] == 'S' && d[3] == 'T') {
avio_skip(pb, 4);
goto start_sync;
}
n = get_stream_idx(d);
if (!((i - avi->last_pkt_pos) & 1) &&
get_stream_idx(d + 1) < s->nb_streams)
continue;
// detect ##ix chunk and skip
if (d[2] == 'i' && d[3] == 'x' && n < s->nb_streams) {
avio_skip(pb, size);
goto start_sync;
}
if (d[2] == 'w' && d[3] == 'c' && n < s->nb_streams) {
avio_skip(pb, 16 * 3 + 8);
goto start_sync;
}
if (avi->dv_demux && n != 0)
continue;
// parse ##dc/##wb
if (n < s->nb_streams) {
AVStream *st;
AVIStream *ast;
st = s->streams[n];
ast = st->priv_data;
if (!ast) {
av_log(s, AV_LOG_WARNING, "Skipping foreign stream %d packet\n", n);
continue;
}
if (s->nb_streams >= 2) {
AVStream *st1 = s->streams[1];
AVIStream *ast1 = st1->priv_data;
// workaround for broken small-file-bug402.avi
if ( d[2] == 'w' && d[3] == 'b'
&& n == 0
&& st ->codecpar->codec_type == AVMEDIA_TYPE_VIDEO
&& st1->codecpar->codec_type == AVMEDIA_TYPE_AUDIO
&& ast->prefix == 'd'*256+'c'
&& (d[2]*256+d[3] == ast1->prefix || !ast1->prefix_count)
) {
n = 1;
st = st1;
ast = ast1;
av_log(s, AV_LOG_WARNING,
"Invalid stream + prefix combination, assuming audio.\n");
}
}
if (d[2] == 'p' && d[3] == 'c' && size <= 4 * 256 + 4) {
int k = avio_r8(pb);
int last = (k + avio_r8(pb) - 1) & 0xFF;
avio_rl16(pb); // flags
// b + (g << 8) + (r << 16);
for (; k <= last; k++)
ast->pal[k] = 0xFFU<<24 | avio_rb32(pb)>>8;
ast->has_pal = 1;
goto start_sync;
} else if (((ast->prefix_count < 5 || sync + 9 > i) &&
d[2] < 128 && d[3] < 128) ||
d[2] * 256 + d[3] == ast->prefix /* ||
(d[2] == 'd' && d[3] == 'c') ||
(d[2] == 'w' && d[3] == 'b') */) {
if (exit_early)
return 0;
if (d[2] * 256 + d[3] == ast->prefix)
ast->prefix_count++;
else {
ast->prefix = d[2] * 256 + d[3];
ast->prefix_count = 0;
}
if (!avi->dv_demux &&
((st->discard >= AVDISCARD_DEFAULT && size == 0) /* ||
// FIXME: needs a little reordering
(st->discard >= AVDISCARD_NONKEY &&
!(pkt->flags & AV_PKT_FLAG_KEY)) */
|| st->discard >= AVDISCARD_ALL)) {
ast->frame_offset += get_duration(ast, size);
avio_skip(pb, size);
goto start_sync;
}
avi->stream_index = n;
ast->packet_size = size + 8;
ast->remaining = size;
if (size) {
uint64_t pos = avio_tell(pb) - 8;
if (!st->internal->index_entries || !st->internal->nb_index_entries ||
st->internal->index_entries[st->internal->nb_index_entries - 1].pos < pos) {
av_add_index_entry(st, pos, ast->frame_offset, size,
0, AVINDEX_KEYFRAME);
}
}
return 0;
}
}
}
if (pb->error)
return pb->error;
return AVERROR_EOF;
}
static int ni_prepare_read(AVFormatContext *s)
{
AVIContext *avi = s->priv_data;
int best_stream_index = 0;
AVStream *best_st = NULL;
AVIStream *best_ast;
int64_t best_ts = INT64_MAX;
int i;
for (i = 0; i < s->nb_streams; i++) {
AVStream *st = s->streams[i];
AVIStream *ast = st->priv_data;
int64_t ts = ast->frame_offset;
int64_t last_ts;
if (!st->internal->nb_index_entries)
continue;
last_ts = st->internal->index_entries[st->internal->nb_index_entries - 1].timestamp;
if (!ast->remaining && ts > last_ts)
continue;
ts = av_rescale_q(ts, st->time_base,
(AVRational) { FFMAX(1, ast->sample_size),
AV_TIME_BASE });
av_log(s, AV_LOG_TRACE, "%"PRId64" %d/%d %"PRId64"\n", ts,
st->time_base.num, st->time_base.den, ast->frame_offset);
if (ts < best_ts) {
best_ts = ts;
best_st = st;
best_stream_index = i;
}
}
if (!best_st)
return AVERROR_EOF;
best_ast = best_st->priv_data;
best_ts = best_ast->frame_offset;
if (best_ast->remaining) {
i = av_index_search_timestamp(best_st,
best_ts,
AVSEEK_FLAG_ANY |
AVSEEK_FLAG_BACKWARD);
} else {
i = av_index_search_timestamp(best_st, best_ts, AVSEEK_FLAG_ANY);
if (i >= 0)
best_ast->frame_offset = best_st->internal->index_entries[i].timestamp;
}
if (i >= 0) {
int64_t pos = best_st->internal->index_entries[i].pos;
pos += best_ast->packet_size - best_ast->remaining;
if (avio_seek(s->pb, pos + 8, SEEK_SET) < 0)
return AVERROR_EOF;
av_assert0(best_ast->remaining <= best_ast->packet_size);
avi->stream_index = best_stream_index;
if (!best_ast->remaining)
best_ast->packet_size =
best_ast->remaining = best_st->internal->index_entries[i].size;
}
else
return AVERROR_EOF;
return 0;
}
static int avi_read_packet(AVFormatContext *s, AVPacket *pkt)
{
AVIContext *avi = s->priv_data;
AVIOContext *pb = s->pb;
int err;
if (CONFIG_DV_DEMUXER && avi->dv_demux) {
int size = avpriv_dv_get_packet(avi->dv_demux, pkt);
if (size >= 0)
return size;
else
goto resync;
}
if (avi->non_interleaved) {
err = ni_prepare_read(s);
if (err < 0)
return err;
}
resync:
if (avi->stream_index >= 0) {
AVStream *st = s->streams[avi->stream_index];
AVIStream *ast = st->priv_data;
int dv_demux = CONFIG_DV_DEMUXER && avi->dv_demux;
int size, err;
if (get_subtitle_pkt(s, st, pkt))
return 0;
// minorityreport.AVI block_align=1024 sample_size=1 IMA-ADPCM
if (ast->sample_size <= 1)
size = INT_MAX;
else if (ast->sample_size < 32)
// arbitrary multiplier to avoid tiny packets for raw PCM data
size = 1024 * ast->sample_size;
else
size = ast->sample_size;
if (size > ast->remaining)
size = ast->remaining;
avi->last_pkt_pos = avio_tell(pb);
err = av_get_packet(pb, pkt, size);
if (err < 0)
return err;
size = err;
if (ast->has_pal && pkt->size < (unsigned)INT_MAX / 2 && !dv_demux) {
uint8_t *pal;
pal = av_packet_new_side_data(pkt,
AV_PKT_DATA_PALETTE,
AVPALETTE_SIZE);
if (!pal) {
av_log(s, AV_LOG_ERROR,
"Failed to allocate data for palette\n");
} else {
memcpy(pal, ast->pal, AVPALETTE_SIZE);
ast->has_pal = 0;
}
}
if (CONFIG_DV_DEMUXER && dv_demux) {
AVBufferRef *avbuf = pkt->buf;
size = avpriv_dv_produce_packet(avi->dv_demux, pkt,
pkt->data, pkt->size, pkt->pos);
pkt->buf = avbuf;
pkt->flags |= AV_PKT_FLAG_KEY;
if (size < 0)
av_packet_unref(pkt);
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
} else if (st->codecpar->codec_type == AVMEDIA_TYPE_SUBTITLE &&
!st->codecpar->codec_tag && read_gab2_sub(s, st, pkt)) {
ast->frame_offset++;
avi->stream_index = -1;
ast->remaining = 0;
goto resync;
} else {
/* XXX: How to handle B-frames in AVI? */
pkt->dts = ast->frame_offset;
// pkt->dts += ast->start;
if (ast->sample_size)
pkt->dts /= ast->sample_size;
pkt->stream_index = avi->stream_index;
if (st->codecpar->codec_type == AVMEDIA_TYPE_VIDEO && st->internal->index_entries) {
AVIndexEntry *e;
int index;
index = av_index_search_timestamp(st, ast->frame_offset, AVSEEK_FLAG_ANY);
e = &st->internal->index_entries[index];
if (index >= 0 && e->timestamp == ast->frame_offset) {
if (index == st->internal->nb_index_entries-1) {
int key=1;
uint32_t state=-1;
if (st->codecpar->codec_id == AV_CODEC_ID_MPEG4) {
const uint8_t *ptr = pkt->data, *end = ptr + FFMIN(size, 256);
while (ptr < end) {
ptr = avpriv_find_start_code(ptr, end, &state);
if (state == 0x1B6 && ptr < end) {
key = !(*ptr & 0xC0);
break;
}
}
}
if (!key)
e->flags &= ~AVINDEX_KEYFRAME;
}
if (e->flags & AVINDEX_KEYFRAME)
pkt->flags |= AV_PKT_FLAG_KEY;
}
} else {
pkt->flags |= AV_PKT_FLAG_KEY;
}
ast->frame_offset += get_duration(ast, pkt->size);
}
ast->remaining -= err;
if (!ast->remaining) {
avi->stream_index = -1;
ast->packet_size = 0;
}
if (!avi->non_interleaved && pkt->pos >= 0 && ast->seek_pos > pkt->pos) {
av_packet_unref(pkt);
goto resync;
}
ast->seek_pos= 0;
if (!avi->non_interleaved && st->internal->nb_index_entries>1 && avi->index_loaded>1) {
int64_t dts= av_rescale_q(pkt->dts, st->time_base, AV_TIME_BASE_Q);
if (avi->dts_max < dts) {
avi->dts_max = dts;
} else if (avi->dts_max - (uint64_t)dts > 2*AV_TIME_BASE) {
avi->non_interleaved= 1;
av_log(s, AV_LOG_INFO, "Switching to NI mode, due to poor interleaving\n");
}
}
return 0;
}
if ((err = avi_sync(s, 0)) < 0)
return err;
goto resync;
}
/* XXX: We make the implicit supposition that the positions are sorted
* for each stream. */
static int avi_read_idx1(AVFormatContext *s, int size)
{
AVIContext *avi = s->priv_data;
AVIOContext *pb = s->pb;
int nb_index_entries, i;
AVStream *st;
AVIStream *ast;
int64_t pos;
unsigned int index, tag, flags, len, first_packet = 1;
int64_t last_pos = -1;
unsigned last_idx = -1;
int64_t idx1_pos, first_packet_pos = 0, data_offset = 0;
int anykey = 0;
nb_index_entries = size / 16;
if (nb_index_entries <= 0)
return AVERROR_INVALIDDATA;
idx1_pos = avio_tell(pb);
avio_seek(pb, avi->movi_list + 4, SEEK_SET);
if (avi_sync(s, 1) == 0)
first_packet_pos = avio_tell(pb) - 8;
avi->stream_index = -1;
avio_seek(pb, idx1_pos, SEEK_SET);
if (s->nb_streams == 1 && s->streams[0]->codecpar->codec_tag == AV_RL32("MMES")) {
first_packet_pos = 0;
data_offset = avi->movi_list;
}
/* Read the entries and sort them in each stream component. */
for (i = 0; i < nb_index_entries; i++) {
if (avio_feof(pb))
return -1;
tag = avio_rl32(pb);
flags = avio_rl32(pb);
pos = avio_rl32(pb);
len = avio_rl32(pb);
av_log(s, AV_LOG_TRACE, "%d: tag=0x%x flags=0x%x pos=0x%"PRIx64" len=%d/",
i, tag, flags, pos, len);
index = ((tag & 0xff) - '0') * 10;
index += (tag >> 8 & 0xff) - '0';
if (index >= s->nb_streams)
continue;
st = s->streams[index];
ast = st->priv_data;
/* Skip 'xxpc' palette change entries in the index until a logic
* to process these is properly implemented. */
if ((tag >> 16 & 0xff) == 'p' && (tag >> 24 & 0xff) == 'c')
continue;
if (first_packet && first_packet_pos) {
if (avi->movi_list + 4 != pos || pos + 500 > first_packet_pos)
data_offset = first_packet_pos - pos;
first_packet = 0;
}
pos += data_offset;
av_log(s, AV_LOG_TRACE, "%d cum_len=%"PRId64"\n", len, ast->cum_len);
// even if we have only a single stream, we should
// switch to non-interleaved to get correct timestamps
if (last_pos == pos)
avi->non_interleaved = 1;
if (last_idx != pos && len) {
av_add_index_entry(st, pos, ast->cum_len, len, 0,
(flags & AVIIF_INDEX) ? AVINDEX_KEYFRAME : 0);
last_idx= pos;
}
ast->cum_len += get_duration(ast, len);
last_pos = pos;
anykey |= flags&AVIIF_INDEX;
}
if (!anykey) {
for (index = 0; index < s->nb_streams; index++) {
st = s->streams[index];
if (st->internal->nb_index_entries)
st->internal->index_entries[0].flags |= AVINDEX_KEYFRAME;
}
}
return 0;
}
/* Scan the index and consider any file with streams more than
* 2 seconds or 64MB apart non-interleaved. */
static int check_stream_max_drift(AVFormatContext *s)
{
int64_t min_pos, pos;
int i;
int *idx = av_mallocz_array(s->nb_streams, sizeof(*idx));
if (!idx)
return AVERROR(ENOMEM);
for (min_pos = pos = 0; min_pos != INT64_MAX; pos = min_pos + 1LU) {
int64_t max_dts = INT64_MIN / 2;
int64_t min_dts = INT64_MAX / 2;
int64_t max_buffer = 0;
min_pos = INT64_MAX;
for (i = 0; i < s->nb_streams; i++) {
AVStream *st = s->streams[i];
AVIStream *ast = st->priv_data;
int n = st->internal->nb_index_entries;
while (idx[i] < n && st->internal->index_entries[idx[i]].pos < pos)
idx[i]++;
if (idx[i] < n) {
int64_t dts;
dts = av_rescale_q(st->internal->index_entries[idx[i]].timestamp /
FFMAX(ast->sample_size, 1),
st->time_base, AV_TIME_BASE_Q);
min_dts = FFMIN(min_dts, dts);
min_pos = FFMIN(min_pos, st->internal->index_entries[idx[i]].pos);
}
}
for (i = 0; i < s->nb_streams; i++) {
AVStream *st = s->streams[i];
AVIStream *ast = st->priv_data;
if (idx[i] && min_dts != INT64_MAX / 2) {
int64_t dts, delta_dts;
dts = av_rescale_q(st->internal->index_entries[idx[i] - 1].timestamp /
FFMAX(ast->sample_size, 1),
st->time_base, AV_TIME_BASE_Q);
delta_dts = av_sat_sub64(dts, min_dts);
max_dts = FFMAX(max_dts, dts);
max_buffer = FFMAX(max_buffer,
av_rescale(delta_dts,
lavf: replace AVStream.codec with AVStream.codecpar Currently, AVStream contains an embedded AVCodecContext instance, which is used by demuxers to export stream parameters to the caller and by muxers to receive stream parameters from the caller. It is also used internally as the codec context that is passed to parsers. In addition, it is also widely used by the callers as the decoding (when demuxer) or encoding (when muxing) context, though this has been officially discouraged since Libav 11. There are multiple important problems with this approach: - the fields in AVCodecContext are in general one of * stream parameters * codec options * codec state However, it's not clear which ones are which. It is consequently unclear which fields are a demuxer allowed to set or a muxer allowed to read. This leads to erratic behaviour depending on whether decoding or encoding is being performed or not (and whether it uses the AVStream embedded codec context). - various synchronization issues arising from the fact that the same context is used by several different APIs (muxers/demuxers, parsers, bitstream filters and encoders/decoders) simultaneously, with there being no clear rules for who can modify what and the different processes being typically delayed with respect to each other. - avformat_find_stream_info() making it necessary to support opening and closing a single codec context multiple times, thus complicating the semantics of freeing various allocated objects in the codec context. Those problems are resolved by replacing the AVStream embedded codec context with a newly added AVCodecParameters instance, which stores only the stream parameters exported by the demuxers or read by the muxers.
11 years ago
st->codecpar->bit_rate,
AV_TIME_BASE));
}
}
if (av_sat_sub64(max_dts, min_dts) > 2 * AV_TIME_BASE ||
max_buffer > 1024 * 1024 * 8 * 8) {
av_free(idx);
return 1;
}
}
av_free(idx);
return 0;
}
static int guess_ni_flag(AVFormatContext *s)
{
int i;
int64_t last_start = 0;
int64_t first_end = INT64_MAX;
int64_t oldpos = avio_tell(s->pb);
for (i = 0; i < s->nb_streams; i++) {
AVStream *st = s->streams[i];
int n = st->internal->nb_index_entries;
unsigned int size;
if (n <= 0)
continue;
if (n >= 2) {
int64_t pos = st->internal->index_entries[0].pos;
unsigned tag[2];
avio_seek(s->pb, pos, SEEK_SET);
tag[0] = avio_r8(s->pb);
tag[1] = avio_r8(s->pb);
avio_rl16(s->pb);
size = avio_rl32(s->pb);
if (get_stream_idx(tag) == i && pos + size > st->internal->index_entries[1].pos)
last_start = INT64_MAX;
if (get_stream_idx(tag) == i && size == st->internal->index_entries[0].size + 8)
last_start = INT64_MAX;
}
if (st->internal->index_entries[0].pos > last_start)
last_start = st->internal->index_entries[0].pos;
if (st->internal->index_entries[n - 1].pos < first_end)
first_end = st->internal->index_entries[n - 1].pos;
}
avio_seek(s->pb, oldpos, SEEK_SET);
if (last_start > first_end)
return 1;
return check_stream_max_drift(s);
}
static int avi_load_index(AVFormatContext *s)
{
AVIContext *avi = s->priv_data;
AVIOContext *pb = s->pb;
uint32_t tag, size;
int64_t pos = avio_tell(pb);
int64_t next;
int ret = -1;
if (avio_seek(pb, avi->movi_end, SEEK_SET) < 0)
goto the_end; // maybe truncated file
av_log(s, AV_LOG_TRACE, "movi_end=0x%"PRIx64"\n", avi->movi_end);
for (;;) {
tag = avio_rl32(pb);
size = avio_rl32(pb);
if (avio_feof(pb))
break;
next = avio_tell(pb) + size + (size & 1);
if (tag == MKTAG('i', 'd', 'x', '1') &&
avi_read_idx1(s, size) >= 0) {
avi->index_loaded=2;
ret = 0;
}else if (tag == MKTAG('L', 'I', 'S', 'T')) {
uint32_t tag1 = avio_rl32(pb);
if (tag1 == MKTAG('I', 'N', 'F', 'O'))
ff_read_riff_info(s, size - 4);
}else if (!ret)
break;
if (avio_seek(pb, next, SEEK_SET) < 0)
break; // something is wrong here
}
the_end:
avio_seek(pb, pos, SEEK_SET);
return ret;
}
static void seek_subtitle(AVStream *st, AVStream *st2, int64_t timestamp)
{
AVIStream *ast2 = st2->priv_data;
int64_t ts2 = av_rescale_q(timestamp, st->time_base, st2->time_base);
av_packet_unref(&ast2->sub_pkt);
if (avformat_seek_file(ast2->sub_ctx, 0, INT64_MIN, ts2, ts2, 0) >= 0 ||
avformat_seek_file(ast2->sub_ctx, 0, ts2, ts2, INT64_MAX, 0) >= 0)
ff_read_packet(ast2->sub_ctx, &ast2->sub_pkt);
}
static int avi_read_seek(AVFormatContext *s, int stream_index,
int64_t timestamp, int flags)
{
AVIContext *avi = s->priv_data;
AVStream *st;
int i, index;
int64_t pos, pos_min;
AVIStream *ast;
/* Does not matter which stream is requested dv in avi has the
* stream information in the first video stream.
*/
if (avi->dv_demux)
stream_index = 0;
if (!avi->index_loaded) {
/* we only load the index on demand */
avi_load_index(s);
avi->index_loaded |= 1;
}
av_assert0(stream_index >= 0);
st = s->streams[stream_index];
ast = st->priv_data;
index = av_index_search_timestamp(st,
timestamp * FFMAX(ast->sample_size, 1),
flags);
if (index < 0) {
if (st->internal->nb_index_entries > 0)
av_log(s, AV_LOG_DEBUG, "Failed to find timestamp %"PRId64 " in index %"PRId64 " .. %"PRId64 "\n",
timestamp * FFMAX(ast->sample_size, 1),
st->internal->index_entries[0].timestamp,
st->internal->index_entries[st->internal->nb_index_entries - 1].timestamp);
return AVERROR_INVALIDDATA;
}
/* find the position */
pos = st->internal->index_entries[index].pos;
timestamp = st->internal->index_entries[index].timestamp / FFMAX(ast->sample_size, 1);
av_log(s, AV_LOG_TRACE, "XX %"PRId64" %d %"PRId64"\n",
timestamp, index, st->internal->index_entries[index].timestamp);
if (CONFIG_DV_DEMUXER && avi->dv_demux) {
/* One and only one real stream for DV in AVI, and it has video */
/* offsets. Calling with other stream indexes should have failed */
/* the av_index_search_timestamp call above. */
if (avio_seek(s->pb, pos, SEEK_SET) < 0)
return -1;
/* Feed the DV video stream version of the timestamp to the */
/* DV demux so it can synthesize correct timestamps. */
ff_dv_offset_reset(avi->dv_demux, timestamp);
avi->stream_index = -1;
return 0;
}
pos_min = pos;
for (i = 0; i < s->nb_streams; i++) {
AVStream *st2 = s->streams[i];
AVIStream *ast2 = st2->priv_data;
ast2->packet_size =
ast2->remaining = 0;
if (ast2->sub_ctx) {
seek_subtitle(st, st2, timestamp);
continue;
}
if (st2->internal->nb_index_entries <= 0)
continue;
// av_assert1(st2->codecpar->block_align);
index = av_index_search_timestamp(st2,
av_rescale_q(timestamp,
st->time_base,
st2->time_base) *
FFMAX(ast2->sample_size, 1),
flags |
AVSEEK_FLAG_BACKWARD |
(st2->codecpar->codec_type != AVMEDIA_TYPE_VIDEO ? AVSEEK_FLAG_ANY : 0));
if (index < 0)
index = 0;
ast2->seek_pos = st2->internal->index_entries[index].pos;
pos_min = FFMIN(pos_min,ast2->seek_pos);
}
for (i = 0; i < s->nb_streams; i++) {
AVStream *st2 = s->streams[i];
AVIStream *ast2 = st2->priv_data;
if (ast2->sub_ctx || st2->internal->nb_index_entries <= 0)
continue;
index = av_index_search_timestamp(
st2,
av_rescale_q(timestamp, st->time_base, st2->time_base) * FFMAX(ast2->sample_size, 1),
flags | AVSEEK_FLAG_BACKWARD | (st2->codecpar->codec_type != AVMEDIA_TYPE_VIDEO ? AVSEEK_FLAG_ANY : 0));
if (index < 0)
index = 0;
while (!avi->non_interleaved && index>0 && st2->internal->index_entries[index-1].pos >= pos_min)
index--;
ast2->frame_offset = st2->internal->index_entries[index].timestamp;
}
/* do the seek */
if (avio_seek(s->pb, pos_min, SEEK_SET) < 0) {
av_log(s, AV_LOG_ERROR, "Seek failed\n");
return -1;
}
avi->stream_index = -1;
avi->dts_max = INT_MIN;
return 0;
}
static int avi_read_close(AVFormatContext *s)
{
int i;
AVIContext *avi = s->priv_data;
for (i = 0; i < s->nb_streams; i++) {
AVStream *st = s->streams[i];
AVIStream *ast = st->priv_data;
if (ast) {
if (ast->sub_ctx) {
av_freep(&ast->sub_ctx->pb);
avformat_close_input(&ast->sub_ctx);
}
av_buffer_unref(&ast->sub_buffer);
av_packet_unref(&ast->sub_pkt);
}
}
av_freep(&avi->dv_demux);
return 0;
}
static int avi_probe(const AVProbeData *p)
{
int i;
/* check file header */
for (i = 0; avi_headers[i][0]; i++)
if (AV_RL32(p->buf ) == AV_RL32(avi_headers[i] ) &&
AV_RL32(p->buf + 8) == AV_RL32(avi_headers[i] + 4))
return AVPROBE_SCORE_MAX;
return 0;
}
AVInputFormat ff_avi_demuxer = {
.name = "avi",
.long_name = NULL_IF_CONFIG_SMALL("AVI (Audio Video Interleaved)"),
.priv_data_size = sizeof(AVIContext),
.extensions = "avi",
.read_probe = avi_probe,
.read_header = avi_read_header,
.read_packet = avi_read_packet,
.read_close = avi_read_close,
.read_seek = avi_read_seek,
.priv_class = &demuxer_class,
};