/*
* Matroska file demuxer
- * Copyright (c) 2003-2008 The FFmpeg Project
+ * Copyright (c) 2003-2008 The Libav Project
*
- * This file is part of FFmpeg.
+ * This file is part of Libav.
*
- * FFmpeg is free software; you can redistribute it and/or
+ * Libav 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,
+ * Libav 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
+ * License along with Libav; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
/**
* @file
* Matroska file demuxer
- * by Ronald Bultje <rbultje@ronald.bitfreak.net>
- * with a little help from Moritz Bunkus <moritz@bunkus.org>
- * totally reworked by Aurelien Jacobs <aurel@gnuage.org>
- * Specs available on the Matroska project page: http://www.matroska.org/.
+ * @author Ronald Bultje <rbultje@ronald.bitfreak.net>
+ * @author with a little help from Moritz Bunkus <moritz@bunkus.org>
+ *
@author totally reworked by Aurelien Jacobs <aurel@gnuage.org>
+ * @see specs available on the Matroska project page: http://www.matroska.org/
*/
#include <stdio.h>
#include "avformat.h"
#include "internal.h"
+#include "avio_internal.h"
/* For ff_codec_get_id(). */
#include "riff.h"
#include "isom.h"
#include "rm.h"
#include "matroska.h"
#include "libavcodec/mpeg4audio.h"
-#include "libavutil/intfloat_readwrite.h"
+#include "libavutil/intfloat.h"
#include "libavutil/intreadwrite.h"
#include "libavutil/avstring.h"
#include "libavutil/lzo.h"
+#include "libavutil/dict.h"
#if CONFIG_ZLIB
#include <zlib.h>
#endif
EBML_NEST,
EBML_PASS,
EBML_STOP,
+ EBML_TYPE_COUNT
} EbmlType;
typedef const struct EbmlSyntax {
int sub_packet_size;
int sub_packet_cnt;
int pkt_cnt;
+ uint64_t buf_timecode;
uint8_t *buf;
} MatroskaTrackAudio;
double time_scale;
uint64_t default_duration;
uint64_t flag_default;
+ uint64_t flag_forced;
MatroskaTrackVideo video;
MatroskaTrackAudio audio;
EbmlList encodings;
uint64_t length;
} MatroskaLevel;
+typedef struct {
+ uint64_t timecode;
+ EbmlList blocks;
+} MatroskaCluster;
+
typedef struct {
AVFormatContext *ctx;
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;
uint64_t skip_to_timecode;
+
+ /* File has a CUES element, but we defer parsing until it is needed. */
+ int cues_parsing_deferred;
+
+ int current_cluster_num_blocks;
+ int64_t current_cluster_pos;
+ MatroskaCluster current_cluster;
+
+ /* File has SSA subtitles which prevent incremental cluster parsing. */
+ int contains_ssa;
} MatroskaDemuxContext;
typedef struct {
EbmlBin bin;
} MatroskaBlock;
-typedef struct {
- uint64_t timecode;
- EbmlList blocks;
-} MatroskaCluster;
-
static EbmlSyntax ebml_header[] = {
{ EBML_ID_EBMLREADVERSION, EBML_UINT, 0, offsetof(Ebml,version), {.u=EBML_VERSION} },
{ EBML_ID_EBMLMAXSIZELENGTH, EBML_UINT, 0, offsetof(Ebml,max_size), {.u=8} },
{ 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 EbmlSyntax matroska_cluster_incremental_parsing[] = {
+ { MATROSKA_ID_CLUSTERTIMECODE,EBML_UINT,0, offsetof(MatroskaCluster,timecode) },
+ { MATROSKA_ID_BLOCKGROUP, EBML_NEST, sizeof(MatroskaBlock), offsetof(MatroskaCluster,blocks), {.n=matroska_blockgroup} },
+ { MATROSKA_ID_SIMPLEBLOCK, EBML_PASS, sizeof(MatroskaBlock), offsetof(MatroskaCluster,blocks), {.n=matroska_blockgroup} },
+ { MATROSKA_ID_CLUSTERPOSITION,EBML_NONE },
+ { MATROSKA_ID_CLUSTERPREVSIZE,EBML_NONE },
+ { MATROSKA_ID_INFO, EBML_NONE },
+ { MATROSKA_ID_CUES, EBML_NONE },
+ { MATROSKA_ID_TAGS, EBML_NONE },
+ { MATROSKA_ID_SEEKHEAD, EBML_NONE },
+ { MATROSKA_ID_CLUSTER, EBML_STOP },
+ { 0 }
+};
+
+static EbmlSyntax matroska_cluster_incremental[] = {
+ { MATROSKA_ID_CLUSTERTIMECODE,EBML_UINT,0, offsetof(MatroskaCluster,timecode) },
+ { MATROSKA_ID_BLOCKGROUP, EBML_STOP },
+ { MATROSKA_ID_SIMPLEBLOCK, EBML_STOP },
+ { MATROSKA_ID_CLUSTERPOSITION,EBML_NONE },
+ { MATROSKA_ID_CLUSTERPREVSIZE,EBML_NONE },
+ { 0 }
+};
+
+static EbmlSyntax matroska_clusters_incremental[] = {
+ { MATROSKA_ID_CLUSTER, EBML_NEST, 0, 0, {.n=matroska_cluster_incremental} },
+ { MATROSKA_ID_INFO, EBML_NONE },
+ { MATROSKA_ID_CUES, EBML_NONE },
+ { MATROSKA_ID_TAGS, EBML_NONE },
+ { MATROSKA_ID_SEEKHEAD, EBML_NONE },
+ { 0 }
+};
+
+static const char *const matroska_doctypes[] = { "matroska", "webm" };
+
/*
* Return: Whether we reached the end of a level in the hierarchy or not.
*/
static int ebml_level_end(MatroskaDemuxContext *matroska)
{
- ByteIOContext *pb = matroska->ctx->pb;
- int64_t pos = url_ftell(pb);
+ AVIOContext *pb = matroska->ctx->pb;
+ int64_t pos = avio_tell(pb);
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;
}
* number.
* Returns: number of bytes read, < 0 on error
*/
-static int ebml_read_num(MatroskaDemuxContext *matroska, ByteIOContext *pb,
+static int ebml_read_num(MatroskaDemuxContext *matroska, AVIOContext *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
+ /* The first byte tells us the length in bytes - avio_r8() can normally
* return 0, but since that's not a valid first ebmlID byte, we can
* use it safely here to catch EOS. */
- if (!(total = get_byte(pb))) {
+ if (!(total = avio_r8(pb))) {
/* we might encounter EOS here */
- if (!url_feof(pb)) {
- int64_t pos = url_ftell(pb);
+ if (!pb->eof_reached) {
+ int64_t pos = avio_tell(pb);
av_log(matroska->ctx, AV_LOG_ERROR,
"Read error at pos. %"PRIu64" (0x%"PRIx64")\n",
pos, pos);
}
/* 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;
+ int64_t pos = avio_tell(pb) - 1;
av_log(matroska->ctx, AV_LOG_ERROR,
"Invalid EBML number size tag 0x%02x at pos %"PRIu64" (0x%"PRIx64")\n",
(uint8_t) total, pos, pos);
}
/* read out length */
- total &= ~len_mask;
+ total ^= 1 << ff_log2_tab[total];
while (n++ < read)
- total = (total << 8) | get_byte(pb);
+ total = (total << 8) | avio_r8(pb);
*number = total;
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, AVIOContext *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.
*/
-static int ebml_read_uint(ByteIOContext *pb, int size, uint64_t *num)
+static int ebml_read_uint(AVIOContext *pb, int size, uint64_t *num)
{
int n = 0;
- if (size < 1 || size > 8)
+ if (size > 8)
return AVERROR_INVALIDDATA;
/* big-endian ordering; build up number */
*num = 0;
while (n++ < size)
- *num = (*num << 8) | get_byte(pb);
+ *num = (*num << 8) | avio_r8(pb);
return 0;
}
* Read the next element as a float.
* 0 is success, < 0 is failure.
*/
-static int ebml_read_float(ByteIOContext *pb, int size, double *num)
+static int ebml_read_float(AVIOContext *pb, int size, double *num)
{
- if (size == 4) {
- *num= av_int2flt(get_be32(pb));
- } else if(size==8){
- *num= av_int2dbl(get_be64(pb));
+ if (size == 0) {
+ *num = 0;
+ } else if (size == 4) {
+ *num = av_int2float(avio_rb32(pb));
+ } else if (size == 8){
+ *num = av_int2double(avio_rb64(pb));
} else
return AVERROR_INVALIDDATA;
* Read the next element as an ASCII string.
* 0 is success, < 0 is failure.
*/
-static int ebml_read_ascii(ByteIOContext *pb, int size, char **str)
+static int ebml_read_ascii(AVIOContext *pb, int size, char **str)
{
- av_free(*str);
+ char *res;
+
/* EBML strings are usually not 0-terminated, so we allocate one
* byte more, read the string and NULL-terminate it ourselves. */
- if (!(*str = av_malloc(size + 1)))
+ if (!(res = av_malloc(size + 1)))
return AVERROR(ENOMEM);
- if (get_buffer(pb, (uint8_t *) *str, size) != size) {
- av_free(*str);
+ if (avio_read(pb, (uint8_t *) res, size) != size) {
+ av_free(res);
return AVERROR(EIO);
}
- (*str)[size] = '\0';
+ (res)[size] = '\0';
+ av_free(*str);
+ *str = res;
return 0;
}
* Read the next element as binary data.
* 0 is success, < 0 is failure.
*/
-static int ebml_read_binary(ByteIOContext *pb, int length, EbmlBin *bin)
+static int ebml_read_binary(AVIOContext *pb, int length, EbmlBin *bin)
{
av_free(bin->data);
if (!(bin->data = av_malloc(length)))
return AVERROR(ENOMEM);
bin->size = length;
- bin->pos = url_ftell(pb);
- if (get_buffer(pb, bin->data, length) != length)
+ bin->pos = avio_tell(pb);
+ if (avio_read(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;
+ AVIOContext *pb = matroska->ctx->pb;
MatroskaLevel *level;
if (matroska->num_levels >= EBML_MAX_DEPTH) {
}
level = &matroska->levels[matroska->num_levels++];
- level->start = url_ftell(pb);
+ level->start = avio_tell(pb);
level->length = length;
return 0;
static int matroska_ebmlnum_uint(MatroskaDemuxContext *matroska,
uint8_t *data, uint32_t size, uint64_t *num)
{
- ByteIOContext pb;
- init_put_byte(&pb, data, size, 0, NULL, NULL, NULL, NULL);
- return ebml_read_num(matroska, &pb, 8, num);
+ AVIOContext pb;
+ ffio_init_context(&pb, data, size, 0, NULL, NULL, NULL, NULL);
+ 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)
{
- ByteIOContext *pb = matroska->ctx->pb;
+ 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
+ };
+ AVIOContext *pb = matroska->ctx->pb;
uint32_t id = syntax->id;
uint64_t length;
int res;
+ void *newelem;
data = (char *)data + syntax->data_offset;
if (syntax->list_elem_size) {
EbmlList *list = data;
- list->elem = av_realloc(list->elem, (list->nb_elem+1)*syntax->list_elem_size);
+ newelem = av_realloc(list->elem, (list->nb_elem+1)*syntax->list_elem_size);
+ if (!newelem)
+ return AVERROR(ENOMEM);
+ list->elem = newelem;
data = (char*)list->elem + list->nb_elem*syntax->list_elem_size;
memset(data, 0, syntax->list_elem_size);
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;
case EBML_NEST: if ((res=ebml_read_master(matroska, length)) < 0)
return res;
if (id == MATROSKA_ID_SEGMENT)
- matroska->segment_start = url_ftell(matroska->ctx->pb);
+ matroska->segment_start = avio_tell(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;
- default: return url_fseek(pb,length,SEEK_CUR)<0 ? AVERROR(EIO) : 0;
+ case EBML_STOP: return 1;
+ default: return avio_skip(pb,length)<0 ? AVERROR(EIO) : 0;
}
if (res == AVERROR_INVALIDDATA)
av_log(matroska->ctx, AV_LOG_ERROR, "Invalid element\n");
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]);
+ if (total < probelen)
+ continue;
+ 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,
uint8_t* data = *buf;
int isize = *buf_size;
uint8_t* pkt_data = NULL;
+ uint8_t av_unused *newpktdata;
int pkt_size = isize;
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;
zstream.avail_in = isize;
do {
pkt_size *= 3;
- pkt_data = av_realloc(pkt_data, pkt_size);
+ newpktdata = av_realloc(pkt_data, pkt_size);
+ if (!newpktdata) {
+ inflateEnd(&zstream);
+ goto failed;
+ }
+ pkt_data = newpktdata;
zstream.avail_out = pkt_size - zstream.total_out;
zstream.next_out = pkt_data + zstream.total_out;
result = inflate(&zstream, Z_NO_FLUSH);
bzstream.avail_in = isize;
do {
pkt_size *= 3;
- pkt_data = av_realloc(pkt_data, pkt_size);
+ newpktdata = av_realloc(pkt_data, pkt_size);
+ if (!newpktdata) {
+ BZ2_bzDecompressEnd(&bzstream);
+ goto failed;
+ }
+ pkt_data = newpktdata;
bzstream.avail_out = pkt_size - bzstream.total_out_lo32;
bzstream.next_out = pkt_data + bzstream.total_out_lo32;
result = BZ2_bzDecompress(&bzstream);
}
}
-static void matroska_merge_packets(AVPacket *out, AVPacket *in)
+static int matroska_merge_packets(AVPacket *out, AVPacket *in)
{
- out->data = av_realloc(out->data, out->size+in->size);
+ void *newdata = av_realloc(out->data, out->size+in->size);
+ if (!newdata)
+ return AVERROR(ENOMEM);
+ out->data = newdata;
memcpy(out->data+out->size, in->data, in->size);
out->size += in->size;
av_destruct_packet(in);
av_free(in);
+ return 0;
}
static void matroska_convert_tag(AVFormatContext *s, EbmlList *list,
- AVMetadata **metadata, char *prefix)
+ AVDictionary **metadata, char *prefix)
{
MatroskaTag *tags = list->elem;
char key[1024];
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_set2(metadata, key, tags[i].string, 0);
+ av_dict_set(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_set2(metadata, key, tags[i].string, 0);
+ av_dict_set(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)
if (tags[i].target.attachuid) {
MatroskaAttachement *attachment = matroska->attachments.elem;
for (j=0; j<matroska->attachments.nb_elem; j++)
- if (attachment[j].uid == tags[i].target.attachuid)
+ if (attachment[j].uid == tags[i].target.attachuid
+ && attachment[j].stream)
matroska_convert_tag(s, &tags[i].tag,
&attachment[j].stream->metadata, NULL);
} else if (tags[i].target.chapteruid) {
MatroskaChapter *chapter = matroska->chapters.elem;
for (j=0; j<matroska->chapters.nb_elem; j++)
- if (chapter[j].uid == tags[i].target.chapteruid)
+ if (chapter[j].uid == tags[i].target.chapteruid
+ && chapter[j].chapter)
matroska_convert_tag(s, &tags[i].tag,
&chapter[j].chapter->metadata, NULL);
} else if (tags[i].target.trackuid) {
MatroskaTrack *track = matroska->tracks.elem;
for (j=0; j<matroska->tracks.nb_elem; j++)
- if (track[j].uid == tags[i].target.trackuid)
+ if (track[j].uid == tags[i].target.trackuid && track[j].stream)
matroska_convert_tag(s, &tags[i].tag,
&track[j].stream->metadata, NULL);
} else {
}
}
-static void matroska_execute_seekhead(MatroskaDemuxContext *matroska)
+static int matroska_parse_seekhead_entry(MatroskaDemuxContext *matroska, int idx)
{
EbmlList *seekhead_list = &matroska->seekhead;
MatroskaSeekhead *seekhead = seekhead_list->elem;
uint32_t level_up = matroska->level_up;
- int64_t before_pos = url_ftell(matroska->ctx->pb);
+ int64_t before_pos = avio_tell(matroska->ctx->pb);
+ uint32_t saved_id = matroska->current_id;
MatroskaLevel level;
- int i;
+ int64_t offset;
+ int ret = 0;
- for (i=0; i<seekhead_list->nb_elem; i++) {
- int64_t offset = seekhead[i].pos + matroska->segment_start;
-
- if (seekhead[i].pos <= before_pos
- || seekhead[i].id == MATROSKA_ID_SEEKHEAD
- || seekhead[i].id == MATROSKA_ID_CLUSTER)
- continue;
-
- /* seek */
- if (url_fseek(matroska->ctx->pb, offset, SEEK_SET) != offset)
- continue;
+ if (idx >= seekhead_list->nb_elem
+ || seekhead[idx].id == MATROSKA_ID_SEEKHEAD
+ || seekhead[idx].id == MATROSKA_ID_CLUSTER)
+ return 0;
+ /* seek */
+ offset = seekhead[idx].pos + matroska->segment_start;
+ if (avio_seek(matroska->ctx->pb, offset, SEEK_SET) == offset) {
/* We don't want to lose our seekhead level, so we add
* a dummy. This is a crude hack. */
if (matroska->num_levels == EBML_MAX_DEPTH) {
av_log(matroska->ctx, AV_LOG_INFO,
"Max EBML element depth (%d) reached, "
"cannot parse further.\n", EBML_MAX_DEPTH);
- break;
+ ret = AVERROR_INVALIDDATA;
+ } else {
+ level.start = 0;
+ level.length = (uint64_t)-1;
+ matroska->levels[matroska->num_levels] = level;
+ matroska->num_levels++;
+ matroska->current_id = 0;
+
+ ret = ebml_parse(matroska, matroska_segment, matroska);
+
+ /* remove dummy level */
+ while (matroska->num_levels) {
+ uint64_t length = matroska->levels[--matroska->num_levels].length;
+ if (length == (uint64_t)-1)
+ break;
+ }
}
+ }
+ /* seek back */
+ avio_seek(matroska->ctx->pb, before_pos, SEEK_SET);
+ matroska->level_up = level_up;
+ matroska->current_id = saved_id;
+
+ return ret;
+}
+
+static void matroska_execute_seekhead(MatroskaDemuxContext *matroska)
+{
+ EbmlList *seekhead_list = &matroska->seekhead;
+ int64_t before_pos = avio_tell(matroska->ctx->pb);
+ int i;
- level.start = 0;
- level.length = (uint64_t)-1;
- matroska->levels[matroska->num_levels] = level;
- matroska->num_levels++;
+ // we should not do any seeking in the streaming case
+ if (!matroska->ctx->pb->seekable ||
+ (matroska->ctx->flags & AVFMT_FLAG_IGNIDX))
+ return;
- ebml_parse(matroska, matroska_segment, matroska);
+ for (i = 0; i < seekhead_list->nb_elem; i++) {
+ MatroskaSeekhead *seekhead = seekhead_list->elem;
+ if (seekhead[i].pos <= before_pos)
+ continue;
- /* remove dummy level */
- while (matroska->num_levels) {
- uint64_t length = matroska->levels[--matroska->num_levels].length;
- if (length == (uint64_t)-1)
- break;
+ // defer cues parsing until we actually need cue data.
+ if (seekhead[i].id == MATROSKA_ID_CUES) {
+ matroska->cues_parsing_deferred = 1;
+ continue;
}
+
+ if (matroska_parse_seekhead_entry(matroska, i) < 0)
+ break;
}
+}
- /* seek back */
- url_fseek(matroska->ctx->pb, before_pos, SEEK_SET);
- matroska->level_up = level_up;
+static void matroska_parse_cues(MatroskaDemuxContext *matroska) {
+ EbmlList *seekhead_list = &matroska->seekhead;
+ MatroskaSeekhead *seekhead = seekhead_list->elem;
+ EbmlList *index_list;
+ MatroskaIndex *index;
+ int index_scale = 1;
+ int i, j;
+
+ for (i = 0; i < seekhead_list->nb_elem; i++)
+ if (seekhead[i].id == MATROSKA_ID_CUES)
+ break;
+ assert(i <= seekhead_list->nb_elem);
+
+ matroska_parse_seekhead_entry(matroska, i);
+
+ index_list = &matroska->index;
+ index = index_list->elem;
+ if (index_list->nb_elem
+ && index[0].time > 1E14/matroska->time_scale) {
+ av_log(matroska->ctx, AV_LOG_WARNING, "Working around broken index.\n");
+ index_scale = matroska->time_scale;
+ }
+ for (i = 0; i < index_list->nb_elem; i++) {
+ EbmlList *pos_list = &index[i].pos;
+ MatroskaIndexPos *pos = pos_list->elem;
+ for (j = 0; j < pos_list->nb_elem; j++) {
+ MatroskaTrack *track = matroska_find_track_by_num(matroska, pos[j].track);
+ if (track && track->stream)
+ av_add_index_entry(track->stream,
+ pos[j].pos + matroska->segment_start,
+ index[i].time/index_scale, 0, 0,
+ AVINDEX_KEYFRAME);
+ }
+ }
}
static int matroska_aac_profile(char *codec_id)
{
int sri;
- for (sri=0; sri<FF_ARRAY_ELEMS(ff_mpeg4audio_sample_rates); sri++)
- if (ff_mpeg4audio_sample_rates[sri] == samplerate)
+ for (sri=0; sri<FF_ARRAY_ELEMS(avpriv_mpeg4audio_sample_rates); sri++)
+ if (avpriv_mpeg4audio_sample_rates[sri] == samplerate)
break;
return sri;
}
-static int matroska_read_header(AVFormatContext *s, AVFormatParameters *ap)
+static int matroska_read_header(AVFormatContext *s)
{
MatroskaDemuxContext *matroska = s->priv_data;
EbmlList *attachements_list = &matroska->attachments;
EbmlList *chapters_list = &matroska->chapters;
MatroskaChapter *chapters;
MatroskaTrack *tracks;
- EbmlList *index_list;
- MatroskaIndex *index;
- int index_scale = 1;
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_set2(&s->metadata, "title", matroska->title, 0);
+ av_dict_set(&s->metadata, "title", matroska->title, 0);
tracks = matroska->tracks.elem;
for (i=0; i < matroska->tracks.nb_elem; i++) {
uint8_t *extradata = NULL;
int extradata_size = 0;
int extradata_offset = 0;
- ByteIOContext b;
+ AVIOContext b;
/* Apply some sanity checks. */
if (track->type != MATROSKA_TRACK_TYPE_VIDEO &&
continue;
if (track->type == MATROSKA_TRACK_TYPE_VIDEO) {
- if (!track->default_duration)
+ if (!track->default_duration && track->video.frame_rate > 0)
track->default_duration = 1000000000/track->video.frame_rate;
if (!track->video.display_width)
track->video.display_width = track->video.pixel_width;
}
if (encodings_list->nb_elem > 1) {
av_log(matroska->ctx, AV_LOG_ERROR,
- "Multiple combined encodings no supported");
+ "Multiple combined encodings not supported");
} else if (encodings_list->nb_elem == 1) {
if (encodings[0].type ||
(encodings[0].compression.algo != MATROSKA_TRACK_ENCODING_COMP_HEADERSTRIP &&
}
}
- st = track->stream = av_new_stream(s, 0);
+ st = track->stream = avformat_new_stream(s, NULL);
if (st == NULL)
return AVERROR(ENOMEM);
} else if (!strcmp(track->codec_id, "A_MS/ACM")
&& track->codec_priv.size >= 14
&& track->codec_priv.data != NULL) {
- init_put_byte(&b, track->codec_priv.data, track->codec_priv.size,
- URL_RDONLY, NULL, NULL, NULL, NULL);
- ff_get_wav_header(&b, st->codec, track->codec_priv.size);
+ int ret;
+ ffio_init_context(&b, track->codec_priv.data, track->codec_priv.size,
+ AVIO_FLAG_READ, NULL, NULL, NULL, NULL);
+ ret = ff_get_wav_header(&b, st->codec, track->codec_priv.size);
+ if (ret < 0)
+ return ret;
codec_id = st->codec->codec_id;
extradata_offset = FFMIN(track->codec_priv.size, 18);
} else if (!strcmp(track->codec_id, "V_QUICKTIME")
&& (track->codec_priv.size >= 86)
&& (track->codec_priv.data != NULL)) {
track->video.fourcc = AV_RL32(track->codec_priv.data);
- codec_id=ff_codec_get_id(codec_movvideo_tags, track->video.fourcc);
+ codec_id=ff_codec_get_id(ff_codec_movvideo_tags, track->video.fourcc);
} else if (codec_id == CODEC_ID_PCM_S16BE) {
switch (track->audio.bitdepth) {
case 8: codec_id = CODEC_ID_PCM_U8; break;
} else if (codec_id == CODEC_ID_AAC && !track->codec_priv.size) {
int profile = matroska_aac_profile(track->codec_id);
int sri = matroska_aac_sri(track->audio.samplerate);
- extradata = av_malloc(5);
+ extradata = av_mallocz(5 + FF_INPUT_BUFFER_PADDING_SIZE);
if (extradata == NULL)
return AVERROR(ENOMEM);
extradata[0] = (profile << 3) | ((sri&0x0E) >> 1);
extradata = av_mallocz(extradata_size);
if (extradata == NULL)
return AVERROR(ENOMEM);
- init_put_byte(&b, extradata, extradata_size, 1,
+ ffio_init_context(&b, extradata, extradata_size, 1,
NULL, NULL, NULL, NULL);
- put_buffer(&b, "TTA1", 4);
- put_le16(&b, 1);
- put_le16(&b, track->audio.channels);
- put_le16(&b, track->audio.bitdepth);
- put_le32(&b, track->audio.out_samplerate);
- put_le32(&b, matroska->ctx->duration * track->audio.out_samplerate);
+ avio_write(&b, "TTA1", 4);
+ avio_wl16(&b, 1);
+ avio_wl16(&b, track->audio.channels);
+ avio_wl16(&b, track->audio.bitdepth);
+ avio_wl32(&b, track->audio.out_samplerate);
+ avio_wl32(&b, matroska->ctx->duration * track->audio.out_samplerate);
} else if (codec_id == CODEC_ID_RV10 || codec_id == CODEC_ID_RV20 ||
codec_id == CODEC_ID_RV30 || codec_id == CODEC_ID_RV40) {
extradata_offset = 26;
} else if (codec_id == CODEC_ID_RA_288 || codec_id == CODEC_ID_COOK ||
codec_id == CODEC_ID_ATRAC3 || codec_id == CODEC_ID_SIPR) {
int flavor;
- init_put_byte(&b, track->codec_priv.data,track->codec_priv.size,
+ ffio_init_context(&b, track->codec_priv.data,track->codec_priv.size,
0, NULL, NULL, NULL, NULL);
- url_fskip(&b, 22);
- flavor = get_be16(&b);
- track->audio.coded_framesize = get_be32(&b);
- url_fskip(&b, 12);
- track->audio.sub_packet_h = get_be16(&b);
- track->audio.frame_size = get_be16(&b);
- track->audio.sub_packet_size = get_be16(&b);
+ avio_skip(&b, 22);
+ flavor = avio_rb16(&b);
+ track->audio.coded_framesize = avio_rb32(&b);
+ avio_skip(&b, 12);
+ track->audio.sub_packet_h = avio_rb16(&b);
+ track->audio.frame_size = avio_rb16(&b);
+ track->audio.sub_packet_size = avio_rb16(&b);
track->audio.buf = av_malloc(track->audio.frame_size * track->audio.sub_packet_h);
if (codec_id == CODEC_ID_RA_288) {
st->codec->block_align = track->audio.coded_framesize;
if (track->time_scale < 0.01)
track->time_scale = 1.0;
- av_set_pts_info(st, 64, matroska->time_scale*track->time_scale, 1000*1000*1000); /* 64 bit pts in ns */
+ avpriv_set_pts_info(st, 64, matroska->time_scale*track->time_scale, 1000*1000*1000); /* 64 bit pts in ns */
st->codec->codec_id = codec_id;
st->start_time = 0;
if (strcmp(track->language, "und"))
- av_metadata_set2(&st->metadata, "language", track->language, 0);
- av_metadata_set2(&st->metadata, "title", track->name, 0);
+ av_dict_set(&st->metadata, "language", track->language, 0);
+ av_dict_set(&st->metadata, "title", track->name, 0);
if (track->flag_default)
st->disposition |= AV_DISPOSITION_DEFAULT;
-
- if (track->default_duration)
- av_reduce(&st->codec->time_base.num, &st->codec->time_base.den,
- track->default_duration, 1000000000, 30000);
+ if (track->flag_forced)
+ st->disposition |= AV_DISPOSITION_FORCED;
if (!st->codec->extradata) {
if(extradata){
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);
+ if (track->default_duration) {
+ av_reduce(&st->r_frame_rate.num, &st->r_frame_rate.den,
+ 1000000000, track->default_duration, 30000);
+ st->avg_frame_rate = st->r_frame_rate;
+ }
} else if (track->type == MATROSKA_TRACK_TYPE_AUDIO) {
st->codec->codec_type = AVMEDIA_TYPE_AUDIO;
st->codec->sample_rate = track->audio.out_samplerate;
st->need_parsing = AVSTREAM_PARSE_HEADERS;
} else if (track->type == MATROSKA_TRACK_TYPE_SUBTITLE) {
st->codec->codec_type = AVMEDIA_TYPE_SUBTITLE;
+ if (st->codec->codec_id == CODEC_ID_SSA)
+ matroska->contains_ssa = 1;
}
}
attachements[j].bin.data && attachements[j].bin.size > 0)) {
av_log(matroska->ctx, AV_LOG_ERROR, "incomplete attachment\n");
} else {
- AVStream *st = av_new_stream(s, 0);
+ AVStream *st = avformat_new_stream(s, NULL);
if (st == NULL)
break;
- av_metadata_set2(&st->metadata, "filename",attachements[j].filename, 0);
+ av_dict_set(&st->metadata, "filename",attachements[j].filename, 0);
+ av_dict_set(&st->metadata, "mimetype", attachements[j].mime, 0);
st->codec->codec_id = CODEC_ID_NONE;
st->codec->codec_type = AVMEDIA_TYPE_ATTACHMENT;
st->codec->extradata = av_malloc(attachements[j].bin.size);
if (chapters[i].start != AV_NOPTS_VALUE && chapters[i].uid
&& (max_start==0 || chapters[i].start > max_start)) {
chapters[i].chapter =
- ff_new_chapter(s, chapters[i].uid, (AVRational){1, 1000000000},
+ avpriv_new_chapter(s, chapters[i].uid, (AVRational){1, 1000000000},
chapters[i].start, chapters[i].end,
chapters[i].title);
- av_metadata_set2(&chapters[i].chapter->metadata,
+ av_dict_set(&chapters[i].chapter->metadata,
"title", chapters[i].title, 0);
max_start = chapters[i].start;
}
- index_list = &matroska->index;
- index = index_list->elem;
- if (index_list->nb_elem
- && index[0].time > 100000000000000/matroska->time_scale) {
- av_log(matroska->ctx, AV_LOG_WARNING, "Working around broken index.\n");
- index_scale = matroska->time_scale;
- }
- for (i=0; i<index_list->nb_elem; i++) {
- EbmlList *pos_list = &index[i].pos;
- MatroskaIndexPos *pos = pos_list->elem;
- for (j=0; j<pos_list->nb_elem; j++) {
- MatroskaTrack *track = matroska_find_track_by_num(matroska,
- pos[j].track);
- if (track && track->stream)
- av_add_index_entry(track->stream,
- pos[j].pos + matroska->segment_start,
- index[i].time/index_scale, 0, 0,
- AVINDEX_KEYFRAME);
- }
- }
-
matroska_convert_tags(s);
return 0;
memcpy(pkt, matroska->packets[0], sizeof(AVPacket));
av_free(matroska->packets[0]);
if (matroska->num_packets > 1) {
+ void *newpackets;
memmove(&matroska->packets[0], &matroska->packets[1],
(matroska->num_packets - 1) * sizeof(AVPacket *));
- matroska->packets =
- av_realloc(matroska->packets, (matroska->num_packets - 1) *
- sizeof(AVPacket *));
+ newpackets = av_realloc(matroska->packets,
+ (matroska->num_packets - 1) * sizeof(AVPacket *));
+ if (newpackets)
+ matroska->packets = newpackets;
} else {
av_freep(&matroska->packets);
+ matroska->prev_pkt = NULL;
}
matroska->num_packets--;
return 0;
size -= n;
track = matroska_find_track_by_num(matroska, num);
- if (size <= 3 || !track || !track->stream) {
+ if (!track || !track->stream) {
av_log(matroska->ctx, AV_LOG_INFO,
"Invalid stream %"PRIu64" or size %u\n", num, size);
- return res;
- }
+ return AVERROR_INVALIDDATA;
+ } else if (size <= 3)
+ return 0;
st = track->stream;
if (st->discard >= AVDISCARD_ALL)
return res;
lace_size[n] = lace_size[n - 1] + snum;
total += lace_size[n];
}
- lace_size[n] = size - total;
+ lace_size[laces - 1] = size - total;
break;
}
}
int x;
if (!track->audio.pkt_cnt) {
- if (st->codec->codec_id == CODEC_ID_RA_288)
+ if (track->audio.sub_packet_cnt == 0)
+ track->audio.buf_timecode = timecode;
+ if (st->codec->codec_id == CODEC_ID_RA_288) {
+ if (size < cfs * h / 2) {
+ av_log(matroska->ctx, AV_LOG_ERROR,
+ "Corrupt int4 RM-style audio packet size\n");
+ res = AVERROR_INVALIDDATA;
+ goto end;
+ }
for (x=0; x<h/2; x++)
memcpy(track->audio.buf+x*2*w+y*cfs,
data+x*cfs, cfs);
- else if (st->codec->codec_id == CODEC_ID_SIPR)
+ } else if (st->codec->codec_id == CODEC_ID_SIPR) {
+ if (size < w) {
+ av_log(matroska->ctx, AV_LOG_ERROR,
+ "Corrupt sipr RM-style audio packet size\n");
+ res = AVERROR_INVALIDDATA;
+ goto end;
+ }
memcpy(track->audio.buf + y*w, data, w);
- else
+ } else {
+ if (size < sps * w / sps) {
+ av_log(matroska->ctx, AV_LOG_ERROR,
+ "Corrupt generic RM-style audio packet size\n");
+ res = AVERROR_INVALIDDATA;
+ goto end;
+ }
for (x=0; x<w/sps; x++)
memcpy(track->audio.buf+sps*(h*x+((h+1)/2)*(y&1)+(y>>1)), data+x*sps, sps);
+ }
if (++track->audio.sub_packet_cnt >= h) {
if (st->codec->codec_id == CODEC_ID_SIPR)
av_new_packet(pkt, a);
memcpy(pkt->data, track->audio.buf
+ a * (h*w / a - track->audio.pkt_cnt--), a);
+ pkt->pts = track->audio.buf_timecode;
+ track->audio.buf_timecode = AV_NOPTS_VALUE;
pkt->pos = pos;
pkt->stream_index = st->index;
dynarray_add(&matroska->packets,&matroska->num_packets,pkt);
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);
}
}
+end:
av_free(lace_size);
return res;
}
+static int matroska_parse_cluster_incremental(MatroskaDemuxContext *matroska)
+{
+ EbmlList *blocks_list;
+ MatroskaBlock *blocks;
+ int i, res;
+ res = ebml_parse(matroska,
+ matroska_cluster_incremental_parsing,
+ &matroska->current_cluster);
+ if (res == 1) {
+ /* New Cluster */
+ if (matroska->current_cluster_pos)
+ ebml_level_end(matroska);
+ ebml_free(matroska_cluster, &matroska->current_cluster);
+ memset(&matroska->current_cluster, 0, sizeof(MatroskaCluster));
+ matroska->current_cluster_num_blocks = 0;
+ matroska->current_cluster_pos = avio_tell(matroska->ctx->pb);
+ matroska->prev_pkt = NULL;
+ /* sizeof the ID which was already read */
+ if (matroska->current_id)
+ matroska->current_cluster_pos -= 4;
+ res = ebml_parse(matroska,
+ matroska_clusters_incremental,
+ &matroska->current_cluster);
+ /* Try parsing the block again. */
+ if (res == 1)
+ res = ebml_parse(matroska,
+ matroska_cluster_incremental_parsing,
+ &matroska->current_cluster);
+ }
+
+ if (!res &&
+ matroska->current_cluster_num_blocks <
+ matroska->current_cluster.blocks.nb_elem) {
+ blocks_list = &matroska->current_cluster.blocks;
+ blocks = blocks_list->elem;
+
+ matroska->current_cluster_num_blocks = blocks_list->nb_elem;
+ i = blocks_list->nb_elem - 1;
+ if (blocks[i].bin.size > 0 && blocks[i].bin.data) {
+ int is_keyframe = blocks[i].non_simple ? !blocks[i].reference : -1;
+ if (!blocks[i].non_simple)
+ blocks[i].duration = AV_NOPTS_VALUE;
+ res = matroska_parse_block(matroska,
+ blocks[i].bin.data, blocks[i].bin.size,
+ blocks[i].bin.pos,
+ matroska->current_cluster.timecode,
+ blocks[i].duration, is_keyframe,
+ matroska->current_cluster_pos);
+ }
+ }
+
+ if (res < 0) matroska->done = 1;
+ return res;
+}
+
static int matroska_parse_cluster(MatroskaDemuxContext *matroska)
{
MatroskaCluster cluster = { 0 };
EbmlList *blocks_list;
MatroskaBlock *blocks;
int i, res;
- int64_t pos = url_ftell(matroska->ctx->pb);
+ int64_t pos;
+ if (!matroska->contains_ssa)
+ return matroska_parse_cluster_incremental(matroska);
+ pos = avio_tell(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) {
+ for (i=0; i<blocks_list->nb_elem && !res; i++)
+ if (blocks[i].bin.size > 0 && blocks[i].bin.data) {
int is_keyframe = blocks[i].non_simple ? !blocks[i].reference : -1;
+ if (!blocks[i].non_simple)
+ blocks[i].duration = AV_NOPTS_VALUE;
res=matroska_parse_block(matroska,
blocks[i].bin.data, blocks[i].bin.size,
blocks[i].bin.pos, cluster.timecode,
static int matroska_read_packet(AVFormatContext *s, AVPacket *pkt)
{
MatroskaDemuxContext *matroska = s->priv_data;
+ int ret = 0;
- while (matroska_deliver_packet(matroska, pkt)) {
+ while (!ret && matroska_deliver_packet(matroska, pkt)) {
if (matroska->done)
return AVERROR_EOF;
- matroska_parse_cluster(matroska);
+ ret = matroska_parse_cluster(matroska);
}
- return 0;
+ return ret;
}
static int matroska_read_seek(AVFormatContext *s, int stream_index,
AVStream *st = s->streams[stream_index];
int i, index, index_sub, index_min;
+ /* Parse the CUES now since we need the index data to seek. */
+ if (matroska->cues_parsing_deferred) {
+ matroska_parse_cues(matroska);
+ matroska->cues_parsing_deferred = 0;
+ }
+
if (!st->nb_index_entries)
return 0;
timestamp = FFMAX(timestamp, st->index_entries[0].timestamp);
if ((index = av_index_search_timestamp(st, timestamp, flags)) < 0) {
- url_fseek(s->pb, st->index_entries[st->nb_index_entries-1].pos, SEEK_SET);
+ avio_seek(s->pb, st->index_entries[st->nb_index_entries-1].pos, SEEK_SET);
+ matroska->current_id = 0;
while ((index = av_index_search_timestamp(st, timestamp, flags)) < 0) {
matroska_clear_queue(matroska);
if (matroska_parse_cluster(matroska) < 0)
index_min = index;
for (i=0; i < matroska->tracks.nb_elem; i++) {
+ tracks[i].audio.pkt_cnt = 0;
+ tracks[i].audio.sub_packet_cnt = 0;
+ tracks[i].audio.buf_timecode = AV_NOPTS_VALUE;
tracks[i].end_timecode = 0;
if (tracks[i].type == MATROSKA_TRACK_TYPE_SUBTITLE
&& !tracks[i].stream->discard != AVDISCARD_ALL) {
}
}
- url_fseek(s->pb, st->index_entries[index_min].pos, SEEK_SET);
+ avio_seek(s->pb, st->index_entries[index_min].pos, SEEK_SET);
+ matroska->current_id = 0;
matroska->skip_to_keyframe = !(flags & AVSEEK_FLAG_ANY);
matroska->skip_to_timecode = st->index_entries[index].timestamp;
matroska->done = 0;
- av_update_cur_dts(s, st, st->index_entries[index].timestamp);
+ ff_update_cur_dts(s, st, st->index_entries[index].timestamp);
return 0;
}
for (n=0; n < matroska->tracks.nb_elem; n++)
if (tracks[n].type == MATROSKA_TRACK_TYPE_AUDIO)
av_free(tracks[n].audio.buf);
+ ebml_free(matroska_cluster, &matroska->current_cluster);
ebml_free(matroska_segment, matroska);
return 0;
}
-AVInputFormat matroska_demuxer = {
- "matroska",
- NULL_IF_CONFIG_SMALL("Matroska file format"),
- sizeof(MatroskaDemuxContext),
- matroska_probe,
- matroska_read_header,
- matroska_read_packet,
- matroska_read_close,
- matroska_read_seek,
- .metadata_conv = ff_mkv_metadata_conv,
+AVInputFormat ff_matroska_demuxer = {
+ .name = "matroska,webm",
+ .long_name = NULL_IF_CONFIG_SMALL("Matroska/WebM file format"),
+ .priv_data_size = sizeof(MatroskaDemuxContext),
+ .read_probe = matroska_probe,
+ .read_header = matroska_read_header,
+ .read_packet = matroska_read_packet,
+ .read_close = matroska_read_close,
+ .read_seek = matroska_read_seek,
};
projects / ffmpeg / blobdiff