Re: [FFmpeg-devel] Test case for Closed caption decoder.

Anshul Sun, 03 May 2015 06:17:33 -0700


On 05/02/2015 08:24 PM, Clément Bœsch wrote:

On Tue, Apr 28, 2015 at 07:50:15PM +0530, Anshul wrote:


On 04/28/2015 02:14 PM, Clément Bœsch wrote:

Then FATE test patch should be applied after the CC patch.

New patch attached. I have used other ass api.

+fate-sub-cc: CMD = fmtstdout ass -f lavfi -i 
"movie=$(TARGET_SAMPLES)/sub/Closedcaption_atsc_rollup.ts[out0+subcc]"

BTW, do we really need to go through libavfilter to extract the cc?

Yes, there is no other way then libavfilter that I know. Closed caption was
possible only after nicolas patch.
if you want more details, Why libavfilter was chosen, you might need to read
last year's mail chain.
there has been long debate on it. so I don't want to indulge myself again in
it.

+
  FATE_SUBTITLES_ASS-$(call DEMDEC, JACOSUB, JACOSUB) += fate-sub-jacosub
  fate-sub-jacosub: CMD = fmtstdout ass -i 
$(TARGET_SAMPLES)/sub/JACOsub_capability_tester.jss
diff --git a/tests/ref/fate/sub-cc b/tests/ref/fate/sub-cc
new file mode 100644
index 0000000..035a51a
--- /dev/null
+++ b/tests/ref/fate/sub-cc
@@ -0,0 +1,38 @@
+[Script Info]
+; Script generated by FFmpeg/Lavc
+ScriptType: v4.00+
+PlayResX: 384
+PlayResY: 288
+
+[V4+ Styles]
+Format: Name, Fontname, Fontsize, PrimaryColour, SecondaryColour, 
OutlineColour, BackColour, Bold, Italic, Underline, StrikeOut, ScaleX, ScaleY, 
Spacing, Angle, BorderStyle, Outline, Shadow, Alignment, MarginL, MarginR, 
MarginV, Encoding
+Style: 
Default,Arial,16,&Hffffff,&Hffffff,&H0,&H0,0,0,0,0,100,100,0,0,1,1,0,2,10,10,10,0
+
+[Events]
+Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
+Dialogue: 0,0:00:36.87,0:00:38.37,Default,,0,0,0,,( explosion )
+Dialogue: 0,0:00:38.37,0:00:40.51,Default,,0,0,0,,( explosion )\N( inaudible 
radio chatter )
+Dialogue: 0,0:00:40.51,0:00:41.88,Default,,0,0,0,,( inaudible radio chatter 
)\N>> Safety remains our number one
+Dialogue: 0,0:00:41.88,0:00:43.34,Default,,0,0,0,,>> Safety remains our number 
one\Npriority.
+Dialogue: 0,0:00:43.34,0:00:44.51,Default,,0,0,0,,priority.\N>> BP can talk 
about safety all
+Dialogue: 0,0:00:44.51,0:00:45.88,Default,,0,0,0,,>> BP can talk about safety 
all\Nthey want, but they're not going
+Dialogue: 0,0:00:45.88,0:00:47.95,Default,,0,0,0,,they want, but they're not 
going\Nto become a safer company.
+Dialogue: 0,0:00:47.95,0:00:49.52,Default,,0,0,0,,to become a safer 
company.\N>> They base everything on risk.
+Dialogue: 0,0:00:49.52,0:00:50.65,Default,,0,0,0,,>> They base everything on 
risk.\N"How many lives can we afford to
+Dialogue: 0,0:00:50.65,0:00:51.65,Default,,0,0,0,,"How many lives can we 
afford to\Nlose before we need to deal with
+Dialogue: 0,0:00:51.65,0:00:52.55,Default,,0,0,0,,lose before we need to deal 
with\Nthis?"
+Dialogue: 0,0:00:52.55,0:00:53.25,Default,,0,0,0,,this?"\N>> 9-1.
+Dialogue: 0,0:00:53.25,0:00:55.22,Default,,0,0,0,,>> 9-1.\N>> Yes, this plant 
just blew up!
+Dialogue: 0,0:00:55.22,0:00:57.19,Default,,0,0,0,,>> Yes, this plant just blew 
up!\N>> From Texas and Alaska to the
+Dialogue: 0,0:00:57.19,0:00:58.39,Default,,0,0,0,,>> From Texas and Alaska to 
the\NGulf of Mexico...
+Dialogue: 0,0:00:58.39,0:01:00.19,Default,,0,0,0,,Gulf of Mexico...\N>> BP 
apologized again...
+Dialogue: 0,0:01:00.19,0:01:02.33,Default,,0,0,0,,>> BP apologized 
again...\N>> ... Apology after apology.
+Dialogue: 0,0:01:02.33,0:01:03.10,Default,,0,0,0,,>> ... Apology after 
apology.\N>> They pledged repeatedly to
+Dialogue: 0,0:01:03.10,0:01:05.40,Default,,0,0,0,,>> They pledged repeatedly 
to\Nrun a safer operation, yet they
+Dialogue: 0,0:01:05.40,0:01:07.17,Default,,0,0,0,,run a safer operation, yet 
they\Ncontinued to cut costs.
+Dialogue: 0,0:01:07.17,0:01:08.67,Default,,0,0,0,,continued to cut costs.\N>> 
We have a facility here that
+Dialogue: 0,0:01:08.67,0:01:10.17,Default,,0,0,0,,>> We have a facility here 
that\Ncould produce a cloud of gas
+Dialogue: 0,0:01:10.17,0:01:11.57,Default,,0,0,0,,could produce a cloud of 
gas\Nthat would make this place look
+Dialogue: 0,0:01:11.57,0:01:13.04,Default,,0,0,0,,that would make this place 
look\Nlike Hiroshima.
+Dialogue: 0,0:01:13.04,0:01:15.21,Default,,0,0,0,,like Hiroshima.\N>> What 
went wrong at BP?
+Dialogue: 0,0:01:15.21,0:01:17.11,Default,,0,0,0,,>> What went wrong at 
BP?\N>> The culture of BP management

This output looks better.



_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

 From 13dedb5bd7df3efd88f6a3b0dbc521e869d8f6a2 Mon Sep 17 00:00:00 2001
From: Anshul Maheshwari <er.anshul.maheshw...@gmail.com>
Date: Tue, 28 Apr 2015 19:41:28 +0530
Subject: [PATCH] Correcting jumbled ass of captions

Signed-off-by: Anshul Maheshwari <er.anshul.maheshw...@gmail.com>
---
  libavcodec/ccaption_dec.c | 6 +++++-
  1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/libavcodec/ccaption_dec.c b/libavcodec/ccaption_dec.c
index da43ca6..540e17f 100644
--- a/libavcodec/ccaption_dec.c
+++ b/libavcodec/ccaption_dec.c
@@ -338,6 +338,10 @@ static int reap_screen(CCaptionSubContext *ctx, int64_t 
pts)
          }

}

+    if(screen->row_used) {
+        ctx->buffer.len -= 2;
+        ctx->buffer.str[ctx->buffer.len] = 0;
+    }

Please add an extra check about the ctx->buffer.len being >= 2

      ctx->startv_time = pts;
      ctx->end_time = pts;
      return ret;
@@ -550,7 +554,7 @@ static int decode(AVCodecContext *avctx, void *data, int 
*got_sub, AVPacket *avp
              int start_time = av_rescale_q(ctx->start_time, avctx->time_base, 
(AVRational){ 1, 100 });
              int end_time = av_rescale_q(ctx->end_time, avctx->time_base, 
(AVRational){ 1, 100 });
              av_dlog(ctx, "cdp writing data (%s)\n",ctx->buffer.str);
-            ret = ff_ass_add_rect(sub, ctx->buffer.str, start_time, end_time - 
start_time , 0);
+            ret = ff_ass_add_rect_bprint(sub, &ctx->buffer, start_time, 
end_time - start_time);
              if (ret < 0)
                  return ret;
              sub->pts = av_rescale_q(ctx->start_time, avctx->time_base, 
AV_TIME_BASE_Q);

LGTM otherwise

New patch attached.

-Anshul

>From 5e801ff90102f886507213bab7a7e1e827e5b2ab Mon Sep 17 00:00:00 2001
From: Anshul Maheshwari <er.anshul.maheshw...@gmail.com>
Date: Sun, 3 May 2015 18:37:40 +0530
Subject: [PATCH] correcting line breaks in cc

Signed-off-by: Anshul Maheshwari <er.anshul.maheshw...@gmail.com>
---
 libavcodec/ccaption_dec.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/libavcodec/ccaption_dec.c b/libavcodec/ccaption_dec.c
index da43ca6..264d21c 100644
--- a/libavcodec/ccaption_dec.c
+++ b/libavcodec/ccaption_dec.c
@@ -338,6 +338,10 @@ static int reap_screen(CCaptionSubContext *ctx, int64_t pts)
         }
 
     }
+    if(screen->row_used && ctx->buffer.len >= 2 ) {
+        ctx->buffer.len -= 2;
+        ctx->buffer.str[ctx->buffer.len] = 0;
+    }
     ctx->startv_time = pts;
     ctx->end_time = pts;
     return ret;
@@ -550,7 +554,7 @@ static int decode(AVCodecContext *avctx, void *data, int *got_sub, AVPacket *avp
             int start_time = av_rescale_q(ctx->start_time, avctx->time_base, (AVRational){ 1, 100 });
             int end_time = av_rescale_q(ctx->end_time, avctx->time_base, (AVRational){ 1, 100 });
             av_dlog(ctx, "cdp writing data (%s)\n",ctx->buffer.str);
-            ret = ff_ass_add_rect(sub, ctx->buffer.str, start_time, end_time - start_time , 0);
+            ret = ff_ass_add_rect_bprint(sub, &ctx->buffer, start_time, end_time - start_time);
             if (ret < 0)
                 return ret;
             sub->pts = av_rescale_q(ctx->start_time, avctx->time_base, AV_TIME_BASE_Q);
-- 
2.1.4

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel

Re: [FFmpeg-devel] Test case for Closed caption decoder.

Reply via email to