EBML_NEST,
EBML_PASS,
EBML_STOP,
+ EBML_TYPE_COUNT
} EbmlType;
typedef const struct EbmlSyntax {
double time_scale;
uint64_t default_duration;
uint64_t flag_default;
+ uint64_t flag_forced;
MatroskaTrackVideo video;
MatroskaTrackAudio audio;
EbmlList encodings;
int num_levels;
MatroskaLevel levels[EBML_MAX_DEPTH];
int level_up;
+ uint32_t current_id;
uint64_t time_scale;
double duration;
AVPacket *prev_pkt;
int done;
- int has_cluster_id;
/* What to skip before effectively reading a packet. */
int skip_to_keyframe;
{ MATROSKA_ID_TRACKDEFAULTDURATION, EBML_UINT, 0, offsetof(MatroskaTrack,default_duration) },
{ MATROSKA_ID_TRACKTIMECODESCALE, EBML_FLOAT,0, offsetof(MatroskaTrack,time_scale), {.f=1.0} },
{ MATROSKA_ID_TRACKFLAGDEFAULT, EBML_UINT, 0, offsetof(MatroskaTrack,flag_default), {.u=1} },
+ { MATROSKA_ID_TRACKFLAGFORCED, EBML_UINT, 0, offsetof(MatroskaTrack,flag_forced), {.u=0} },
{ MATROSKA_ID_TRACKVIDEO, EBML_NEST, 0, offsetof(MatroskaTrack,video), {.n=matroska_track_video} },
{ MATROSKA_ID_TRACKAUDIO, EBML_NEST, 0, offsetof(MatroskaTrack,audio), {.n=matroska_track_audio} },
{ MATROSKA_ID_TRACKCONTENTENCODINGS,EBML_NEST, 0, 0, {.n=matroska_track_encodings} },
{ MATROSKA_ID_TRACKFLAGENABLED, EBML_NONE },
- { MATROSKA_ID_TRACKFLAGFORCED, EBML_NONE },
{ MATROSKA_ID_TRACKFLAGLACING, EBML_NONE },
{ MATROSKA_ID_CODECNAME, EBML_NONE },
{ MATROSKA_ID_CODECDECODEALL, EBML_NONE },
{ MATROSKA_ID_TAGSTRING, EBML_UTF8, 0, offsetof(MatroskaTag,string) },
{ MATROSKA_ID_TAGLANG, EBML_STR, 0, offsetof(MatroskaTag,lang), {.s="und"} },
{ MATROSKA_ID_TAGDEFAULT, EBML_UINT, 0, offsetof(MatroskaTag,def) },
+ { MATROSKA_ID_TAGDEFAULT_BUG, EBML_UINT, 0, offsetof(MatroskaTag,def) },
{ MATROSKA_ID_SIMPLETAG, EBML_NEST, sizeof(MatroskaTag), offsetof(MatroskaTag,sub), {.n=matroska_simpletag} },
{ 0 }
};
{ MATROSKA_ID_CUES, EBML_NEST, 0, 0, {.n=matroska_index } },
{ MATROSKA_ID_TAGS, EBML_NEST, 0, 0, {.n=matroska_tags } },
{ MATROSKA_ID_SEEKHEAD, EBML_NEST, 0, 0, {.n=matroska_seekhead } },
- { MATROSKA_ID_CLUSTER, EBML_STOP, 0, offsetof(MatroskaDemuxContext,has_cluster_id) },
+ { MATROSKA_ID_CLUSTER, EBML_STOP },
{ 0 }
};
{ 0 }
};
+static const char *matroska_doctypes[] = { "matroska", "webm" };
+
/*
* Return: Whether we reached the end of a level in the hierarchy or not.
*/
if (matroska->num_levels > 0) {
MatroskaLevel *level = &matroska->levels[matroska->num_levels - 1];
- if (pos - level->start >= level->length) {
+ if (pos - level->start >= level->length || matroska->current_id) {
matroska->num_levels--;
return 1;
}
static int ebml_read_num(MatroskaDemuxContext *matroska, ByteIOContext *pb,
int max_size, uint64_t *number)
{
- int len_mask = 0x80, read = 1, n = 1;
- int64_t total = 0;
+ int read = 1, n = 1;
+ uint64_t total = 0;
/* The first byte tells us the length in bytes - get_byte() can normally
* return 0, but since that's not a valid first ebmlID byte, we can
}
/* get the length of the EBML number */
- while (read <= max_size && !(total & len_mask)) {
- read++;
- len_mask >>= 1;
- }
+ read = 8 - ff_log2_tab[total];
if (read > max_size) {
int64_t pos = url_ftell(pb) - 1;
av_log(matroska->ctx, AV_LOG_ERROR,
}
/* read out length */
- total &= ~len_mask;
+ total ^= 1 << ff_log2_tab[total];
while (n++ < read)
total = (total << 8) | get_byte(pb);
return read;
}
+/**
+ * Read a EBML length value.
+ * This needs special handling for the "unknown length" case which has multiple
+ * encodings.
+ */
+static int ebml_read_length(MatroskaDemuxContext *matroska, ByteIOContext *pb,
+ uint64_t *number)
+{
+ int res = ebml_read_num(matroska, pb, 8, number);
+ if (res > 0 && *number + 1 == 1ULL << (7 * res))
+ *number = 0xffffffffffffffULL;
+ return res;
+}
+
/*
* Read the next element as an unsigned int.
* 0 is success, < 0 is failure.
{
int n = 0;
- if (size < 1 || size > 8)
+ if (size > 8)
return AVERROR_INVALIDDATA;
/* big-endian ordering; build up number */
*/
static int ebml_read_float(ByteIOContext *pb, int size, double *num)
{
- if (size == 4) {
+ if (size == 0) {
+ *num = 0;
+ } else if (size == 4) {
*num= av_int2flt(get_be32(pb));
} else if(size==8){
*num= av_int2dbl(get_be64(pb));
if (!(*str = av_malloc(size + 1)))
return AVERROR(ENOMEM);
if (get_buffer(pb, (uint8_t *) *str, size) != size) {
- av_free(*str);
+ av_freep(str);
return AVERROR(EIO);
}
(*str)[size] = '\0';
bin->size = length;
bin->pos = url_ftell(pb);
- if (get_buffer(pb, bin->data, length) != length)
+ if (get_buffer(pb, bin->data, length) != length) {
+ av_freep(&bin->data);
return AVERROR(EIO);
+ }
return 0;
}
* are supposed to be sub-elements which can be read separately.
* 0 is success, < 0 is failure.
*/
-static int ebml_read_master(MatroskaDemuxContext *matroska, int length)
+static int ebml_read_master(MatroskaDemuxContext *matroska, uint64_t length)
{
ByteIOContext *pb = matroska->ctx->pb;
MatroskaLevel *level;
{
ByteIOContext pb;
init_put_byte(&pb, data, size, 0, NULL, NULL, NULL, NULL);
- return ebml_read_num(matroska, &pb, 8, num);
+ return ebml_read_num(matroska, &pb, FFMIN(size, 8), num);
}
/*
for (i=0; syntax[i].id; i++)
if (id == syntax[i].id)
break;
+ if (!syntax[i].id && id == MATROSKA_ID_CLUSTER &&
+ matroska->num_levels > 0 &&
+ matroska->levels[matroska->num_levels-1].length == 0xffffffffffffff)
+ return 0; // we reached the end of an unknown size cluster
if (!syntax[i].id && id != EBML_ID_VOID && id != EBML_ID_CRC32)
av_log(matroska->ctx, AV_LOG_INFO, "Unknown entry 0x%X\n", id);
return ebml_parse_elem(matroska, &syntax[i], data);
static int ebml_parse(MatroskaDemuxContext *matroska, EbmlSyntax *syntax,
void *data)
{
- uint64_t id;
- int res = ebml_read_num(matroska, matroska->ctx->pb, 4, &id);
- id |= 1 << 7*res;
- return res < 0 ? res : ebml_parse_id(matroska, syntax, id, data);
+ if (!matroska->current_id) {
+ uint64_t id;
+ int res = ebml_read_num(matroska, matroska->ctx->pb, 4, &id);
+ if (res < 0)
+ return res;
+ matroska->current_id = id | 1 << 7*res;
+ }
+ return ebml_parse_id(matroska, syntax, matroska->current_id, data);
}
static int ebml_parse_nest(MatroskaDemuxContext *matroska, EbmlSyntax *syntax,
static int ebml_parse_elem(MatroskaDemuxContext *matroska,
EbmlSyntax *syntax, void *data)
{
+ static const uint64_t max_lengths[EBML_TYPE_COUNT] = {
+ [EBML_UINT] = 8,
+ [EBML_FLOAT] = 8,
+ // max. 16 MB for strings
+ [EBML_STR] = 0x1000000,
+ [EBML_UTF8] = 0x1000000,
+ // max. 256 MB for binary data
+ [EBML_BIN] = 0x10000000,
+ // no limits for anything else
+ };
ByteIOContext *pb = matroska->ctx->pb;
uint32_t id = syntax->id;
uint64_t length;
list->nb_elem++;
}
- if (syntax->type != EBML_PASS && syntax->type != EBML_STOP)
- if ((res = ebml_read_num(matroska, pb, 8, &length)) < 0)
+ if (syntax->type != EBML_PASS && syntax->type != EBML_STOP) {
+ matroska->current_id = 0;
+ if ((res = ebml_read_length(matroska, pb, &length)) < 0)
return res;
+ if (max_lengths[syntax->type] && length > max_lengths[syntax->type]) {
+ av_log(matroska->ctx, AV_LOG_ERROR,
+ "Invalid length 0x%"PRIx64" > 0x%"PRIx64" for syntax element %i\n",
+ length, max_lengths[syntax->type], syntax->type);
+ return AVERROR_INVALIDDATA;
+ }
+ }
switch (syntax->type) {
case EBML_UINT: res = ebml_read_uint (pb, length, data); break;
matroska->segment_start = url_ftell(matroska->ctx->pb);
return ebml_parse_nest(matroska, syntax->def.n, data);
case EBML_PASS: return ebml_parse_id(matroska, syntax->def.n, id, data);
- case EBML_STOP: *(int *)data = 1; return 1;
+ case EBML_STOP: return 1;
default: return url_fseek(pb,length,SEEK_CUR)<0 ? AVERROR(EIO) : 0;
}
if (res == AVERROR_INVALIDDATA)
static int matroska_probe(AVProbeData *p)
{
uint64_t total = 0;
- int len_mask = 0x80, size = 1, n = 1;
- static const char probe_data[] = "matroska";
+ int len_mask = 0x80, size = 1, n = 1, i;
/* EBML header? */
if (AV_RB32(p->buf) != EBML_ID_HEADER)
if (p->buf_size < 4 + size + total)
return 0;
- /* The header must contain the document type 'matroska'. For now,
+ /* The header should contain a known document type. For now,
* we don't parse the whole header but simply check for the
* availability of that array of characters inside the header.
* Not fully fool-proof, but good enough. */
- for (n = 4+size; n <= 4+size+total-(sizeof(probe_data)-1); n++)
- if (!memcmp(p->buf+n, probe_data, sizeof(probe_data)-1))
- return AVPROBE_SCORE_MAX;
+ for (i = 0; i < FF_ARRAY_ELEMS(matroska_doctypes); i++) {
+ int probelen = strlen(matroska_doctypes[i]);
+ for (n = 4+size; n <= 4+size+total-probelen; n++)
+ if (!memcmp(p->buf+n, matroska_doctypes[i], probelen))
+ return AVPROBE_SCORE_MAX;
+ }
- return 0;
+ // probably valid EBML header but no recognized doctype
+ return AVPROBE_SCORE_MAX/2;
}
static MatroskaTrack *matroska_find_track_by_num(MatroskaDemuxContext *matroska,
int result = 0;
int olen;
+ if (pkt_size >= 10000000)
+ return -1;
+
switch (encodings[0].compression.algo) {
case MATROSKA_TRACK_ENCODING_COMP_HEADERSTRIP:
return encodings[0].compression.settings.size;
for (i=0; i < list->nb_elem; i++) {
const char *lang = strcmp(tags[i].lang, "und") ? tags[i].lang : NULL;
+
+ if (!tags[i].name) {
+ av_log(s, AV_LOG_WARNING, "Skipping invalid tag with no TagName.\n");
+ continue;
+ }
if (prefix) snprintf(key, sizeof(key), "%s/%s", prefix, tags[i].name);
else av_strlcpy(key, tags[i].name, sizeof(key));
if (tags[i].def || !lang) {
- av_metadata_set(metadata, key, tags[i].string);
+ av_metadata_set2(metadata, key, tags[i].string, 0);
if (tags[i].sub.nb_elem)
matroska_convert_tag(s, &tags[i].sub, metadata, key);
}
if (lang) {
av_strlcat(key, "-", sizeof(key));
av_strlcat(key, lang, sizeof(key));
- av_metadata_set(metadata, key, tags[i].string);
+ av_metadata_set2(metadata, key, tags[i].string, 0);
if (tags[i].sub.nb_elem)
matroska_convert_tag(s, &tags[i].sub, metadata, key);
}
}
+ ff_metadata_conv(metadata, NULL, ff_mkv_metadata_conv);
}
static void matroska_convert_tags(AVFormatContext *s)
MatroskaSeekhead *seekhead = seekhead_list->elem;
uint32_t level_up = matroska->level_up;
int64_t before_pos = url_ftell(matroska->ctx->pb);
+ uint32_t saved_id = matroska->current_id;
MatroskaLevel level;
int i;
+ // we should not do any seeking in the streaming case
+ if (url_is_streamed(matroska->ctx->pb) ||
+ (matroska->ctx->flags & AVFMT_FLAG_IGNIDX))
+ return;
+
for (i=0; i<seekhead_list->nb_elem; i++) {
int64_t offset = seekhead[i].pos + matroska->segment_start;
level.length = (uint64_t)-1;
matroska->levels[matroska->num_levels] = level;
matroska->num_levels++;
+ matroska->current_id = 0;
ebml_parse(matroska, matroska_segment, matroska);
/* seek back */
url_fseek(matroska->ctx->pb, before_pos, SEEK_SET);
matroska->level_up = level_up;
+ matroska->current_id = saved_id;
}
static int matroska_aac_profile(char *codec_id)
uint64_t max_start = 0;
Ebml ebml = { 0 };
AVStream *st;
- int i, j;
+ int i, j, res;
matroska->ctx = s;
/* First read the EBML header. */
if (ebml_parse(matroska, ebml_syntax, &ebml)
|| ebml.version > EBML_VERSION || ebml.max_size > sizeof(uint64_t)
- || ebml.id_length > sizeof(uint32_t) || strcmp(ebml.doctype, "matroska")
- || ebml.doctype_version > 2) {
+ || ebml.id_length > sizeof(uint32_t) || ebml.doctype_version > 2) {
av_log(matroska->ctx, AV_LOG_ERROR,
"EBML header using unsupported features\n"
"(EBML version %"PRIu64", doctype %s, doc version %"PRIu64")\n",
ebml.version, ebml.doctype, ebml.doctype_version);
+ ebml_free(ebml_syntax, &ebml);
return AVERROR_PATCHWELCOME;
}
+ for (i = 0; i < FF_ARRAY_ELEMS(matroska_doctypes); i++)
+ if (!strcmp(ebml.doctype, matroska_doctypes[i]))
+ break;
+ if (i >= FF_ARRAY_ELEMS(matroska_doctypes)) {
+ av_log(s, AV_LOG_WARNING, "Unknown EBML doctype '%s'\n", ebml.doctype);
+ }
ebml_free(ebml_syntax, &ebml);
/* The next thing is a segment. */
- if (ebml_parse(matroska, matroska_segments, matroska) < 0)
- return -1;
+ if ((res = ebml_parse(matroska, matroska_segments, matroska)) < 0)
+ return res;
matroska_execute_seekhead(matroska);
+ if (!matroska->time_scale)
+ matroska->time_scale = 1000000;
if (matroska->duration)
matroska->ctx->duration = matroska->duration * matroska->time_scale
* 1000 / AV_TIME_BASE;
- av_metadata_set(&s->metadata, "title", matroska->title);
+ av_metadata_set2(&s->metadata, "title", matroska->title, 0);
tracks = matroska->tracks.elem;
for (i=0; i < matroska->tracks.nb_elem; i++) {
st->codec->codec_id = codec_id;
st->start_time = 0;
if (strcmp(track->language, "und"))
- av_metadata_set(&st->metadata, "language", track->language);
- av_metadata_set(&st->metadata, "title", track->name);
+ av_metadata_set2(&st->metadata, "language", track->language, 0);
+ av_metadata_set2(&st->metadata, "title", track->name, 0);
if (track->flag_default)
st->disposition |= AV_DISPOSITION_DEFAULT;
+ if (track->flag_forced)
+ st->disposition |= AV_DISPOSITION_FORCED;
if (track->default_duration)
av_reduce(&st->codec->time_base.num, &st->codec->time_base.den,
255);
if (st->codec->codec_id != CODEC_ID_H264)
st->need_parsing = AVSTREAM_PARSE_HEADERS;
+ if (track->default_duration)
+ st->avg_frame_rate = av_d2q(1000000000.0/track->default_duration, INT_MAX);
} else if (track->type == MATROSKA_TRACK_TYPE_AUDIO) {
st->codec->codec_type = AVMEDIA_TYPE_AUDIO;
st->codec->sample_rate = track->audio.out_samplerate;
st->codec->channels = track->audio.channels;
+ if (st->codec->codec_id != CODEC_ID_AAC)
+ st->need_parsing = AVSTREAM_PARSE_HEADERS;
} else if (track->type == MATROSKA_TRACK_TYPE_SUBTITLE) {
st->codec->codec_type = AVMEDIA_TYPE_SUBTITLE;
}
AVStream *st = av_new_stream(s, 0);
if (st == NULL)
break;
- av_metadata_set(&st->metadata, "filename",attachements[j].filename);
+ av_metadata_set2(&st->metadata, "filename",attachements[j].filename, 0);
st->codec->codec_id = CODEC_ID_NONE;
st->codec->codec_type = AVMEDIA_TYPE_ATTACHMENT;
st->codec->extradata = av_malloc(attachements[j].bin.size);
ff_new_chapter(s, chapters[i].uid, (AVRational){1, 1000000000},
chapters[i].start, chapters[i].end,
chapters[i].title);
- av_metadata_set(&chapters[i].chapter->metadata,
- "title", chapters[i].title);
+ av_metadata_set2(&chapters[i].chapter->metadata,
+ "title", chapters[i].title, 0);
max_start = chapters[i].start;
}
int offset = 0, pkt_size = lace_size[n];
uint8_t *pkt_data = data;
- if (lace_size[n] > size) {
+ if (pkt_size > size) {
av_log(matroska->ctx, AV_LOG_ERROR, "Invalid packet size\n");
break;
}
if (matroska->prev_pkt &&
timecode != AV_NOPTS_VALUE &&
matroska->prev_pkt->pts == timecode &&
- matroska->prev_pkt->stream_index == st->index)
+ matroska->prev_pkt->stream_index == st->index &&
+ st->codec->codec_id == CODEC_ID_SSA)
matroska_merge_packets(matroska->prev_pkt, pkt);
else {
dynarray_add(&matroska->packets,&matroska->num_packets,pkt);
int i, res;
int64_t pos = url_ftell(matroska->ctx->pb);
matroska->prev_pkt = NULL;
- if (matroska->has_cluster_id){
- /* For the first cluster we parse, its ID was already read as
- part of matroska_read_header(), so don't read it again */
- res = ebml_parse_id(matroska, matroska_clusters,
- MATROSKA_ID_CLUSTER, &cluster);
+ if (matroska->current_id)
pos -= 4; /* sizeof the ID which was already read */
- matroska->has_cluster_id = 0;
- } else
- res = ebml_parse(matroska, matroska_clusters, &cluster);
+ res = ebml_parse(matroska, matroska_clusters, &cluster);
blocks_list = &cluster.blocks;
blocks = blocks_list->elem;
for (i=0; i<blocks_list->nb_elem; i++)
- if (blocks[i].bin.size > 0) {
+ if (blocks[i].bin.size > 0 && blocks[i].bin.data) {
int is_keyframe = blocks[i].non_simple ? !blocks[i].reference : -1;
res=matroska_parse_block(matroska,
blocks[i].bin.data, blocks[i].bin.size,
return 0;
}
-AVInputFormat matroska_demuxer = {
- "matroska",
- NULL_IF_CONFIG_SMALL("Matroska file format"),
+AVInputFormat ff_matroska_demuxer = {
+ "matroska,webm",
+ NULL_IF_CONFIG_SMALL("Matroska/WebM file format"),
sizeof(MatroskaDemuxContext),
matroska_probe,
matroska_read_header,
matroska_read_packet,
matroska_read_close,
matroska_read_seek,
- .metadata_conv = ff_mkv_metadata_conv,
};
projects / ffmpeg / blobdiff