[FFmpeg-devel] [PATCH] png parser

Peter Holik peter
Thu Jun 4 10:40:55 CEST 2009


> On Mon, Jun 01, 2009 at 12:27:52PM +0200, Peter Holik wrote:
>> Hi!
>>
>> I've now reworked my patch to really parse a png for end of frame.
>>
>> Now you can filter out png's of any stream like
>>
>> cat garbage a.png garbage b.png garbage | ffmpeg -f image2pipe -vcodec png -i - c%d.png
>>
>> cu Peter
>>  Makefile     |    1
>>  allcodecs.c  |    1
>>  png_parser.c |  124 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>>  3 files changed, 126 insertions(+)
>> 63f7cd1bb826726f1d2a9e7254410cd7ec98a81c  png-parser.patch
>> diff --git a/libavcodec/Makefile b/libavcodec/Makefile
>> index 14ef9f4..189231b 100644
>> --- a/libavcodec/Makefile
>> +++ b/libavcodec/Makefile
>> @@ -411,6 +411,7 @@ OBJS-$(CONFIG_MLP_PARSER)              += mlp_parser.o mlp.o
>>  OBJS-$(CONFIG_MPEG4VIDEO_PARSER)       += mpeg4video_parser.o h263.o mpeg12data.o mpegvideo.o
>> error_resilience.o
>>  OBJS-$(CONFIG_MPEGAUDIO_PARSER)        += mpegaudio_parser.o mpegaudiodecheader.o
>> mpegaudiodata.o
>>  OBJS-$(CONFIG_MPEGVIDEO_PARSER)        += mpegvideo_parser.o mpeg12.o mpeg12data.o mpegvideo.o
>> error_resilience.o
>> +OBJS-$(CONFIG_PNG_PARSER)              += png_parser.o
>>  OBJS-$(CONFIG_PNM_PARSER)              += pnm_parser.o pnm.o
>>  OBJS-$(CONFIG_VC1_PARSER)              += vc1_parser.o
>>  OBJS-$(CONFIG_VP3_PARSER)              += vp3_parser.o
>> diff --git a/libavcodec/allcodecs.c b/libavcodec/allcodecs.c
>> index e8d5d27..362021f 100644
>> --- a/libavcodec/allcodecs.c
>> +++ b/libavcodec/allcodecs.c
>> @@ -334,6 +334,7 @@ void avcodec_register_all(void)
>>      REGISTER_PARSER  (MPEG4VIDEO, mpeg4video);
>>      REGISTER_PARSER  (MPEGAUDIO, mpegaudio);
>>      REGISTER_PARSER  (MPEGVIDEO, mpegvideo);
>> +    REGISTER_PARSER  (PNG, png);
>>      REGISTER_PARSER  (PNM, pnm);
>>      REGISTER_PARSER  (VC1, vc1);
>>      REGISTER_PARSER  (VP3, vp3);
>> diff --git a/libavcodec/png_parser.c b/libavcodec/png_parser.c
>> new file mode 100644
>> index 0000000..64f2356
>> --- /dev/null
>> +++ b/libavcodec/png_parser.c
>> @@ -0,0 +1,124 @@
>> +/*
>> + * PNG parser
>> + * Copyright (c) 2009 Peter Holik
>> + *
>> + * This file is part of FFmpeg.
>> + *
>> + * FFmpeg is free software; you can redistribute it and/or
>> + * modify it under the terms of the GNU Lesser General Public
>> + * License as published by the Free Software Foundation; either
>> + * version 2.1 of the License, or (at your option) any later version.
>> + *
>> + * FFmpeg is distributed in the hope that it will be useful,
>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
>> + * Lesser General Public License for more details.
>> + *
>> + * You should have received a copy of the GNU Lesser General Public
>> + * License along with FFmpeg; if not, write to the Free Software
>> + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
>> + */
>> +
>> +/**
>> + * @file libavcodec/png_parser.c
>> + * PNG parser
>> + */
>> +
>> +#include "parser.h"
>> +#include "bytestream.h"
>> +#include "png.h"
>> +
>> +static int png_parse(AVCodecParserContext *s, AVCodecContext *avctx,
>> +                     const uint8_t **poutbuf, int *poutbuf_size,
>> +                     const uint8_t *buf, int buf_size)
>> +{
>> +    ParseContext *pc = s->priv_data;
>> +    uint8_t *bytestream;
>> +    uint8_t *bytestream_start;
>> +    uint8_t *bytestream_end;
>> +
>> +    /* EOF considered as end of frame */
>> +    if (buf_size == 0)
>> +        goto flush;
>> +
>> +    *poutbuf      = NULL;
>> +    *poutbuf_size = 0;
>> +
>> +    /* first add new buffer to parsebuffer */
>> +    bytestream_start = av_fast_realloc(pc->buffer, &pc->buffer_size,
>> +                          pc->index + buf_size + FF_INPUT_BUFFER_PADDING_SIZE);
>> +    if(!bytestream_start)
>> +        goto fail;
>> +    pc->buffer = bytestream_start;
>> +    memcpy(&pc->buffer[pc->index], buf, buf_size);
>
> mandatory copying is unacceptable
> the case where each input is exactly a png frame for example should
> be free of any timeconsuming operation
>
> maybe ff_combine_frame() would be usefull

But ff_combine_frame also does memcpy, so there is no difference to my approach.

cu Peter





More information about the ffmpeg-devel mailing list