Add support for parsing audio streams (though should be easy to extend to video)

via libavcodec.
Parsing can be done at the demuxer stage (currently disabled) or at the decoder
(ad_ffmpeg, enabled).
Should allow using the libavcodec AAC, DTS, ... decoders independent of container
format.


git-svn-id: svn://svn.mplayerhq.hu/mplayer/trunk@30130 b3059339-0415-0410-9bf9-f77b7e298cf2

7 files changed, 156 insertions, 3 deletions

diff --git a/libmpcodecs/ad_ffmpeg.c b/libmpcodecs/ad_ffmpeg.c
index fa005b4a15..068a32a307 100644
--- a/libmpcodecs/ad_ffmpeg.c
+++ b/libmpcodecs/ad_ffmpeg.c
@@ -167,7 +167,17 @@ static int decode_audio(sh_audio_t *sh_audio,unsigned char *buf,int minlen,int m
 	int len2=maxlen;
 	double pts;
 	int x=ds_get_packet_pts(sh_audio->ds,&start, &pts);
-	if(x<=0) break; // error
+	if(x<=0) {
+	    start = NULL;
+	    x = 0;
+	    ds_parse(sh_audio->ds, &start, &x, MP_NOPTS_VALUE, 0);
+	    if (x <= 0)
+	        break; // error
+	} else {
+	    int in_size = x;
+	    int consumed = ds_parse(sh_audio->ds, &start, &x, pts, 0);
+	    sh_audio->ds->buffer_pos -= in_size - consumed;
+	}
 	av_init_packet(&pkt);
 	pkt.data = start;
 	pkt.size = x;
@@ -178,7 +188,8 @@ static int decode_audio(sh_audio_t *sh_audio,unsigned char *buf,int minlen,int m
 	y=avcodec_decode_audio3(sh_audio->context,(int16_t*)buf,&len2,&pkt);
 //printf("return:%d samples_out:%d bitstream_in:%d sample_sum:%d\n", y, len2, x, len); fflush(stdout);
 	if(y<0){ mp_msg(MSGT_DECAUDIO,MSGL_V,"lavc_audio: error\n");break; }
-	if(y<x) sh_audio->ds->buffer_pos+=y-x;  // put back data (HACK!)
+	if(!sh_audio->needs_parsing && y<x)
+	    sh_audio->ds->buffer_pos+=y-x;  // put back data (HACK!)
 	if(len2>0){
 	  if (((AVCodecContext *)sh_audio->context)->channels >= 5) {
             int samplesize = av_get_bits_per_sample_format(((AVCodecContext *)
diff --git a/libmpdemux/aviheader.c b/libmpdemux/aviheader.c
index 2717478221..8435fa3c4f 100644
--- a/libmpdemux/aviheader.c
+++ b/libmpdemux/aviheader.c
@@ -219,6 +219,7 @@ while(1){
         mp_msg(MSGT_DEMUX, MSGL_INFO, MSGTR_AudioID, "aviheader", stream_id);
         memcpy(&sh_audio->audio,&h,sizeof(h));
         sh_audio->stream_delay = (float)sh_audio->audio.dwStart * sh_audio->audio.dwScale/sh_audio->audio.dwRate;
+        sh_audio->needs_parsing = 1;
       }
       last_fccType=h.fccType;
       last_fccHandler=h.fccHandler;
diff --git a/libmpdemux/demux_mpg.c b/libmpdemux/demux_mpg.c
index 09bcd95178..72ab2a9b27 100644
--- a/libmpdemux/demux_mpg.c
+++ b/libmpdemux/demux_mpg.c
@@ -270,6 +270,7 @@ static void new_audio_stream(demuxer_t *demux, int aid){
     sh_audio_t* sh_a;
     new_sh_audio(demux,aid);
     sh_a = (sh_audio_t*)demux->a_streams[aid];
+    sh_a->needs_parsing = 1;
     switch(aid & 0xE0){  // 1110 0000 b  (high 3 bit: type  low 5: id)
       case 0x00: sh_a->format=0x50;break; // mpeg
       case 0xA0: sh_a->format=0x10001;break;  // dvd pcm
diff --git a/libmpdemux/demux_ts.c b/libmpdemux/demux_ts.c
index c7966da9ba..af9e2109f4 100644
--- a/libmpdemux/demux_ts.c
+++ b/libmpdemux/demux_ts.c
@@ -305,6 +305,7 @@ static void ts_add_stream(demuxer_t * demuxer, ES_stream_t *es)
 		if(sh)
 		{
 			const char *lang = pid_lang_from_pmt(priv, es->pid);
+			sh->needs_parsing = 1;
 			sh->format = IS_AUDIO(es->type) ? es->type : es->subtype;
 			sh->ds = demuxer->audio;
 
diff --git a/libmpdemux/demuxer.c b/libmpdemux/demuxer.c
index 038b2999cf..216323a2a1 100644
--- a/libmpdemux/demuxer.c
+++ b/libmpdemux/demuxer.c
@@ -53,6 +53,13 @@
 #endif
 #endif
 
+// This is quite experimental, in particular it will mess up the pts values
+// in the queue - on the other hand it might fix some issues like generating
+// broken files with mencoder and stream copy.
+// Better leave it disabled for now, if we find no use for it this code should
+// just be removed again.
+#define PARSE_ON_ADD 0
+
 void resync_video_stream(sh_video_t *sh_video);
 void resync_audio_stream(sh_audio_t *sh_audio);
 
@@ -278,6 +285,10 @@ void free_sh_sub(sh_sub_t *sh)
         ass_free_track(sh->ass_track);
 #endif
     free(sh->lang);
+#ifdef CONFIG_LIBAVCODEC
+    av_parser_close(sh->parser);
+    av_freep(&sh->avctx);
+#endif
     free(sh);
 }
 
@@ -315,6 +326,10 @@ void free_sh_audio(demuxer_t *demuxer, int id)
     free(sh->wf);
     free(sh->codecdata);
     free(sh->lang);
+#ifdef CONFIG_LIBAVCODEC
+    av_parser_close(sh->parser);
+    av_freep(&sh->avctx);
+#endif
     free(sh);
 }
 
@@ -343,6 +358,10 @@ void free_sh_video(sh_video_t *sh)
 {
     mp_msg(MSGT_DEMUXER, MSGL_DBG2, "DEMUXER: freeing sh_video at %p\n", sh);
     free(sh->bih);
+#ifdef CONFIG_LIBAVCODEC
+    av_parser_close(sh->parser);
+    av_freep(&sh->avctx);
+#endif
     free(sh);
 }
 
@@ -396,7 +415,7 @@ void free_demuxer(demuxer_t *demuxer)
 }
 
 
-void ds_add_packet(demux_stream_t *ds, demux_packet_t *dp)
+static void ds_add_packet_internal(demux_stream_t *ds, demux_packet_t *dp)
 {
     // append packet to DS stream:
     ++ds->packs;
@@ -416,6 +435,109 @@ void ds_add_packet(demux_stream_t *ds, demux_packet_t *dp)
            ds->demuxer->video->packs);
 }
 
+#ifdef CONFIG_LIBAVCODEC
+static void allocate_parser(AVCodecContext **avctx, AVCodecParserContext **parser, unsigned format)
+{
+    enum CodecID codec_id = CODEC_ID_NONE;
+    extern int avcodec_initialized;
+    if (!avcodec_initialized) {
+        avcodec_init();
+        avcodec_register_all();
+        avcodec_initialized = 1;
+    }
+    switch (format) {
+    case 0x2000:
+    case 0x332D6361:
+    case 0x332D4341:
+    case MKTAG('d', 'n', 'e', 't'):
+    case MKTAG('s', 'a', 'c', '3'):
+        codec_id = CODEC_ID_AC3;
+        break;
+    case MKTAG('E', 'A', 'C', '3'):
+        codec_id = CODEC_ID_EAC3;
+        break;
+    case 0x2001:
+    case 0x86:
+        codec_id = CODEC_ID_DTS;
+        break;
+    case 0x55:
+    case 0x5500736d:
+    case MKTAG('.', 'm', 'p', '3'):
+    case MKTAG('M', 'P', 'E', ' '):
+    case MKTAG('L', 'A', 'M', 'E'):
+        codec_id = CODEC_ID_MP3;
+        break;
+    case 0x50:
+    case MKTAG('.', 'm', 'p', '2'):
+    case MKTAG('.', 'm', 'p', '1'):
+        codec_id = CODEC_ID_MP2;
+        break;
+    }
+    if (codec_id != CODEC_ID_NONE) {
+        *avctx = avcodec_alloc_context();
+        if (!*avctx)
+            return;
+        *parser = av_parser_init(codec_id);
+        if (!*parser)
+            av_freep(avctx);
+    }
+}
+
+static void get_parser(sh_common_t *sh, AVCodecContext **avctx, AVCodecParserContext **parser)
+{
+    *avctx  = NULL;
+    *parser = NULL;
+
+    if (!sh || !sh->needs_parsing)
+        return;
+
+    *avctx  = sh->avctx;
+    *parser = sh->parser;
+    if (*parser)
+        return;
+
+    allocate_parser(avctx, parser, sh->format);
+    sh->avctx  = *avctx;
+    sh->parser = *parser;
+}
+
+int ds_parse(demux_stream_t *ds, uint8_t **buffer, int *len, double pts, off_t pos)
+{
+    AVCodecContext *avctx;
+    AVCodecParserContext *parser;
+    get_parser(ds->sh, &avctx, &parser);
+    if (!parser)
+        return *len;
+    return av_parser_parse2(parser, avctx, buffer, len, *buffer, *len, pts, pts, pos);
+}
+#endif
+
+void ds_add_packet(demux_stream_t *ds, demux_packet_t *dp)
+{
+#if PARSE_ON_ADD && defined(CONFIG_LIBAVCODEC)
+    int len = dp->len;
+    int pos = 0;
+    while (len > 0) {
+        uint8_t *parsed_start = dp->buffer + pos;
+        int parsed_len = len;
+        int consumed = ds_parse(ds->sh, &parsed_start, &parsed_len, dp->pts, dp->pos);
+        pos += consumed;
+        len -= consumed;
+        if (parsed_start == dp->buffer && parsed_len == dp->len) {
+            ds_add_packet_internal(ds, dp);
+        } else if (parsed_len) {
+            demux_packet_t *dp2 = new_demux_packet(parsed_len);
+            dp2->pos = dp->pos;
+            dp2->pts = dp->pts; // should be parser->pts but that works badly
+            memcpy(dp2->buffer, parsed_start, parsed_len);
+            ds_add_packet_internal(ds, dp2);
+        }
+    }
+#else
+    ds_add_packet_internal(ds, dp);
+#endif
+}
+
 void ds_read_packet(demux_stream_t *ds, stream_t *stream, int len,
                     double pts, off_t pos, int flags)
 {
@@ -514,6 +636,18 @@ int ds_fill_buffer(demux_stream_t *ds)
             break;
         }
         if (!demux_fill_buffer(demux, ds)) {
+#if PARSE_ON_ADD && defined(CONFIG_LIBAVCODEC)
+            uint8_t *parsed_start = NULL;
+            int parsed_len = 0;
+            ds_parse(ds->sh, &parsed_start, &parsed_len, MP_NOPTS_VALUE, 0);
+            if (parsed_len) {
+                demux_packet_t *dp2 = new_demux_packet(parsed_len);
+                dp2->pts = MP_NOPTS_VALUE;
+                memcpy(dp2->buffer, parsed_start, parsed_len);
+                ds_add_packet_internal(ds, dp2);
+                continue;
+            }
+#endif
             mp_dbg(MSGT_DEMUXER, MSGL_DBG2,
                    "ds_fill_buffer()->demux_fill_buffer() failed\n");
             break; // EOF
diff --git a/libmpdemux/demuxer.h b/libmpdemux/demuxer.h
index a2e9185b1e..6ed858024c 100644
--- a/libmpdemux/demuxer.h
+++ b/libmpdemux/demuxer.h
@@ -382,6 +382,7 @@ int ds_get_packet(demux_stream_t *ds,unsigned char **start);
 int ds_get_packet_pts(demux_stream_t *ds, unsigned char **start, double *pts);
 int ds_get_packet_sub(demux_stream_t *ds,unsigned char **start);
 double ds_get_next_pts(demux_stream_t *ds);
+int ds_parse(demux_stream_t *sh, uint8_t **buffer, int *len, double pts, off_t pos);
 
 // This is defined here because demux_stream_t ins't defined in stream.h
 stream_t* new_ds_stream(demux_stream_t *ds);
diff --git a/libmpdemux/stheader.h b/libmpdemux/stheader.h
index 576ed68c08..99ae63ad0d 100644
--- a/libmpdemux/stheader.h
+++ b/libmpdemux/stheader.h
@@ -31,6 +31,10 @@
   unsigned int format; \
   int initialized; \
   float stream_delay; /* number of seconds stream should be delayed (according to dwStart or similar) */ \
+  /* things needed for parsing */ \
+  int needs_parsing; \
+  struct AVCodecContext *avctx; \
+  struct AVCodecParserContext *parser; \
   /* audio: last known pts value in output from decoder \
    * video: predicted/interpolated PTS of the current frame */ \
   double pts; \