Added --analyzevideo

docs/CHANGES.TXT

@@ -6,7 +6,11 @@
 - New: Added build/installation script for .pkg.tar.xz (Arch Linux).
 - New: Added tarball generation script.
 - New: Added Autoconf build scripts for CCExtractor to generate makefiles.
-  
+- New: Added --analyzevideo. If present the video stream will be processed even if the
+  subtitles are in a different stream. This is useful when we want video information
+  (resolution, frame type, etc). -vides now implies this option too. 
+  [Note: Tentative - some possibly breaking changed were made for this, so if you
+  use it validate results]
 
 0.85b (2017-1-26)
 -----------------

src/lib_ccx/ccx_common_option.c

@@ -69,6 +69,7 @@ void init_options (struct ccx_s_options *options)
 	options->ocr_oem = 0; // By default, set Tesseract OEM mode OEM_TESSERACT_ONLY (0)
 	options->mkvlang = NULL; // By default, all the languages are extracted 
 	options->ignore_pts_jumps = 1;
+	options->analyze_video_stream = 0;
 
 	/*HardsubX related stuff*/
 	options->hardsubx_ocr_mode = 0;

src/lib_ccx/ccx_common_option.h

@@ -134,6 +134,7 @@ struct ccx_s_options // Options from user parameters
 	char *ocrlang;                    // The name of the .traineddata file to be loaded with tesseract
 	int ocr_oem;                      // The Tesseract OEM mode, could be 0 (default), 1 or 2
 	char *mkvlang;                    // The name of the language stream for MKV
+	int analyze_video_stream;         // If 1, the video stream will be processed even if we're using a different one for subtitles.
 
 	/*HardsubX related stuff*/
 	int hardsubx_ocr_mode;

src/lib_ccx/es_functions.c

@@ -242,7 +242,7 @@ static int es_video_sequence(struct lib_cc_decode *ctx, struct bitstream *esstre
 			ctx->no_bitstream_error = 0;
 			break;
 		}
-
+		// Sequence header
 		if (!ctx->in_pic_data && startcode == 0xB3)
 		{
 			if (!read_seq_info(ctx, esstream))
@@ -254,7 +254,7 @@ static int es_video_sequence(struct lib_cc_decode *ctx, struct bitstream *esstre
 			ctx->saw_seqgoppic = 1;
 			continue;
 		}
-
+		// Group of pictures
 		if (!ctx->in_pic_data && startcode == 0xB8)
 		{
 			if (!read_gop_info(ctx, esstream,sub))
@@ -266,7 +266,7 @@ static int es_video_sequence(struct lib_cc_decode *ctx, struct bitstream *esstre
 			ctx->saw_seqgoppic = 2;
 			continue;
 		}
-
+		// Picture
 		if (!ctx->in_pic_data && startcode == 0x00)
 		{
 			if (!read_pic_info(ctx, esstream, sub))
@@ -679,6 +679,7 @@ static int gop_header(struct lib_cc_decode *ctx, struct bitstream *esstream, str
 // will point to where we want to restart after getting more.
 static int read_pic_info(struct lib_cc_decode *ctx, struct bitstream *esstream, struct cc_subtitle *sub)
 {
+	char frame_type_to_char[] = { '?', 'I', 'P','B', 'D', '?', '?','?' };
 	debug("Read PIC Info\n");
 
 	// We only get here after seeing that start code
@@ -728,10 +729,17 @@ static int read_pic_info(struct lib_cc_decode *ctx, struct bitstream *esstream,
 	{
 		set_fts(ctx->timing); // Initialize fts
 	}
-
+	/*
-	dbg_print(CCX_DMT_VIDES, "  t:%d r:%d p:%d", ctx->top_field_first,
+	dbg_print(CCX_DMT_VIDES, "  frametype: %d (%c) t:%d r:%d p:%d", ctx->picture_coding_type, 
+		frame_type_to_char[ctx->picture_coding_type], ctx->top_field_first,
 			ctx->repeat_first_field, ctx->progressive_frame);
-	dbg_print(CCX_DMT_VIDES, "  FTS: %s\n", print_mstime_static(get_fts(ctx->timing, ctx->current_field)));
+	dbg_print(CCX_DMT_VIDES, "  FTS: %lld  (%s)\n", ctx->timing->current_pts, 
+		print_mstime_static(get_fts(ctx->timing, ctx->current_field)));
+		*/
+	if (ctx->picture_coding_type == 1) // Write I-Frame in ffprobe format for easy comparison
+	{
+		dbg_print(CCX_DMT_VIDES, "key_frame=1|pkt_pts=%lld|pict_type=I\n", ctx->timing->current_pts);
+	}
 
 	// Set min_pts/sync_pts according to the current time stamp.
 	// Use fts_at_gop_start as reference when a GOP header was seen

src/lib_ccx/general_loop.c

@@ -799,6 +799,7 @@ void segment_output_file(struct lib_ccx_ctx *ctx, struct lib_cc_decode *dec_ctx)
 		}
 	}
 }
+
 int general_loop(struct lib_ccx_ctx *ctx)
 {
 	struct lib_cc_decode *dec_ctx = NULL;
@@ -863,6 +864,7 @@ int general_loop(struct lib_ccx_ctx *ctx)
 		{
 			struct cap_info* cinfo = NULL;
 			struct encoder_ctx *enc_ctx = NULL;
+			// Find most promising stream: teletext, DVB, ISDB, ATSC, in that order
 			int pid = get_best_stream(ctx->demux_ctx);
 			if(pid < 0)
 			{
@@ -873,6 +875,39 @@ int general_loop(struct lib_ccx_ctx *ctx)
 				ignore_other_stream(ctx->demux_ctx, pid);
 				data_node = get_data_stream(datalist, pid);
 			}
+			if (ccx_options.analyze_video_stream)
+			{
+				int video_pid= get_video_stream(ctx->demux_ctx);
+				if (video_pid != pid && video_pid!=-1)
+				{					
+					struct cap_info* cinfo_video = get_cinfo(ctx->demux_ctx, pid); // Must be pid, not video_pid or DVB crashes (possibly buffer consumption?) - TODO
+					struct lib_cc_decode *dec_ctx_video = update_decoder_list_cinfo(ctx, cinfo_video);
+					struct cc_subtitle *dec_sub_video = &dec_ctx_video->dec_sub;
+					struct demuxer_data *data_node_video = get_data_stream(datalist, video_pid);					
+					if (data_node_video)
+					{
+						if (data_node_video->pts != CCX_NOPTS)
+						{
+							struct ccx_rational tb = { 1,MPEG_CLOCK_FREQ };
+							LLONG pts;
+							if (data_node_video->tb.num != 1 || data_node_video->tb.den != MPEG_CLOCK_FREQ)
+							{
+								pts = change_timebase(data_node_video->pts, data_node_video->tb, tb);
+							}
+							else
+								pts = data_node_video->pts;
+							set_current_pts(dec_ctx_video->timing, pts);
+							set_fts(dec_ctx_video->timing);
+						}
+						size_t got = process_m2v(dec_ctx_video, data_node_video->buffer, data_node_video->len, dec_sub_video);
+						if (got > 0)
+						{
+							memmove(data_node_video->buffer, data_node_video->buffer + got, (size_t)(data_node_video->len - got));
+							data_node_video->len -= got;
+						}
+					}
+				}
+			}
 
 			cinfo = get_cinfo(ctx->demux_ctx, pid);
 			enc_ctx = update_encoder_list_cinfo(ctx, cinfo);

src/lib_ccx/params.c

@@ -1877,6 +1877,12 @@ int parse_parameters (struct ccx_s_options *opt, int argc, char *argv[])
 		if (strcmp (argv[i],"-vides")==0)
 		{
 			opt->debug_mask |= CCX_DMT_VIDES;
+			opt->analyze_video_stream = 1;
+			continue;
+		}
+		if (strcmp(argv[i], "-anvid") == 0 || strcmp(argv[i], "--analyzevideo") == 0)
+		{
+			opt->analyze_video_stream = 1;
 			continue;
 		}
 		if (strcmp (argv[i],"-xds")==0)

src/lib_ccx/ts_functions.c

@@ -578,7 +578,8 @@ int copy_payload_to_capbuf(struct cap_info *cinfo, struct ts_payload *payload)
 {
 	int newcapbuflen;
 
-	if(cinfo->ignore == CCX_TRUE)
+	if(cinfo->ignore == CCX_TRUE && 
+		(cinfo->stream != CCX_STREAM_TYPE_VIDEO_MPEG2 || !ccx_options.analyze_video_stream))
 	{
 		return CCX_OK;
 	}
@@ -914,7 +915,8 @@ long ts_readstream(struct ccx_demuxer *ctx, struct demuxer_data **data)
 				look_for_caption_data (ctx, &payload);
 			continue;
 		}
-		else if (cinfo->ignore)
+		else if (cinfo->ignore == CCX_TRUE &&
+			(cinfo->stream != CCX_STREAM_TYPE_VIDEO_MPEG2 || !ccx_options.analyze_video_stream))
 		{
 			if(cinfo->codec_private_data)
 			{

src/lib_ccx/ts_info.c

@@ -97,6 +97,17 @@ void ignore_other_sib_stream(struct cap_info* head, int pid)
 	}
 }
 
+int get_video_stream(struct ccx_demuxer *ctx)
+{
+	struct cap_info* iter;
+	list_for_each_entry(iter, &ctx->cinfo_tree.all_stream, all_stream, struct cap_info)
+	{
+		if (iter->stream == CCX_STREAM_TYPE_VIDEO_MPEG2)
+			return iter->pid;
+	}
+	return -1;
+}
+
 int get_best_stream(struct ccx_demuxer *ctx)
 {
 	struct cap_info* iter;

src/lib_ccx/ts_tables.c

@@ -133,7 +133,7 @@ int parse_PMT (struct ccx_demuxer *ctx, unsigned char *buf, int len,  struct pro
 	}
 	else if (table_id == 0xC1)
 	{
-                //SCTE 57 2003
+        //SCTE 57 2003
 		dbg_print(CCX_DMT_PARSE, "PMT: PROGRAM NAME Table need implementation");
 		unsigned c0length = (buf[1] << 8 | buf[2]) & 0xFFF; // 12 bytes
 		dbg_print(CCX_DMT_PARSE, "Program name message length: %u", c0length);