On Thu, Jul 2, 2015 at 11:17 PM, Pavel Koshevoy <pkoshe...@gmail.com> wrote:
> On Thu, Jul 2, 2015 at 12:18 PM, Kieran Kunhya <kier...@ob-encoder.com>
> wrote:
>
>> ---
>> libavcodec/h264.c | 11 +++++++++++
>> libavcodec/h264.h | 2 ++
>> libavcodec/h264_sei.c | 35 ++++++++++++++++++++++++++++++++++-
>> 3 files changed, 47 insertions(+), 1 deletion(-)
>>
>> diff --git a/libavcodec/h264.c b/libavcodec/h264.c
>> index 1cbd4cb..1569ec8 100644
>> --- a/libavcodec/h264.c
>> +++ b/libavcodec/h264.c
>> @@ -609,6 +609,8 @@ static int h264_init_context(AVCodecContext *avctx,
>> H264Context *h)
>> h->prev_frame_num = -1;
>> h->sei_fpa.frame_packing_arrangement_cancel_flag = -1;
>> h->has_afd = 0;
>> + h->a53_caption_size = 0;
>> + h->a53_caption = NULL;
>>
>> h->next_outputed_poc = INT_MIN;
>> for (i = 0; i < MAX_DELAYED_PIC_COUNT; i++)
>> @@ -879,6 +881,15 @@ static void decode_postinit(H264Context *h, int
>> setup_finished)
>> }
>> }
>>
>> + if (h->a53_caption) {
>> + AVFrameSideData *sd =
>> + av_frame_new_side_data(cur->f, AV_FRAME_DATA_A53_CC,
>> h->a53_caption_size);
>> + if (sd)
>> + memcpy(sd->data, h->a53_caption, h->a53_caption_size);
>> + av_freep(&h->a53_caption);
>> + h->a53_caption_size = 0;
>> + }
>> +
>> cur->mmco_reset = h->mmco_reset;
>> h->mmco_reset = 0;
>>
>> diff --git a/libavcodec/h264.h b/libavcodec/h264.h
>> index 7565e03..3aef49f 100644
>> --- a/libavcodec/h264.h
>> +++ b/libavcodec/h264.h
>> @@ -784,6 +784,8 @@ typedef struct H264Context {
>> uint8_t afd;
>> int has_afd;
>>
>> + int a53_caption_size;
>> + uint8_t *a53_caption;
>>
>> // Timestamp stuff
>> int sei_buffering_period_present; ///< Buffering period SEI flag
>> diff --git a/libavcodec/h264_sei.c b/libavcodec/h264_sei.c
>> index b6ec5c7..5c6fcc0 100644
>> --- a/libavcodec/h264_sei.c
>> +++ b/libavcodec/h264_sei.c
>> @@ -111,7 +111,7 @@ static int decode_picture_timing(H264Context *h)
>> static int decode_user_data_itu_t_t35(H264Context *h, int size)
>> {
>> uint32_t user_identifier;
>> - int dtg_active_format;
>> + int dtg_active_format, cc_count, user_data_type_code;
>>
>> if (size < 7)
>> return -1;
>> @@ -143,6 +143,39 @@ FF_ENABLE_DEPRECATION_WARNINGS
>> skip_bits(&h->gb, 6);
>> }
>> break;
>> + case 0x47413934: // "GA94" closed captions
>>
>
>
> Pardon my ignorance, but is this endian safe? 0x47413934 is "GA94" on
> big-endian architectures, won't it be 0x34394147 on little-endian?
>
The way the value is read will always fill the variable with this
exact value, no matter what the endianness of the underlying memory
actually is.
- Hendrik
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
