Parse matroska clusters incrementally as opposed to all at once.

ffmpeg/patches/ugly/41_matroska_cluster_incremental.patch

+Index: libavformat/matroskadec.c
+===================================================================
+--- libavformat/matroskadec.c   (revision 105457)
++++ libavformat/matroskadec.c   (working copy)
+@@ -228,6 +228,11 @@
+ } MatroskaLevel;
+ 
+ typedef struct {
++    uint64_t timecode;
++    EbmlList blocks;
++} MatroskaCluster;
++
++typedef struct {
+     AVFormatContext *ctx;
+ 
+     /* EBML stuff */
+@@ -262,6 +267,10 @@
+ 
+     /* 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;
+ } MatroskaDemuxContext;
+ 
+ typedef struct {
+@@ -271,11 +280,6 @@
+     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} },
+@@ -546,6 +550,38 @@
+     { 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 *matroska_doctypes[] = { "matroska", "webm" };
+ 
+ /*
+@@ -1690,6 +1726,7 @@
+                            sizeof(AVPacket *));
+         } else {
+             av_freep(&matroska->packets);
++            matroska->prev_pkt = NULL;
+         }
+         matroska->num_packets--;
+         return 0;
+@@ -1993,6 +2030,59 @@
+     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 agiain. */
++        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;
++            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_read_packet(AVFormatContext *s, AVPacket *pkt)
+ {
+     MatroskaDemuxContext *matroska = s->priv_data;
+@@ -2001,7 +2091,7 @@
+     while (!ret && matroska_deliver_packet(matroska, pkt)) {
+         if (matroska->done)
+             return AVERROR_EOF;
+-        ret = matroska_parse_cluster(matroska);
++        ret = matroska_parse_cluster_incremental(matroska);
+     }
+ 
+     return ret;
+@@ -2075,6 +2165,7 @@
+     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;