[FFmpeg-devel] [PATCH] png parser
Peter Holik
peter
Wed Jun 10 16:33:48 CEST 2009
>> On Wed, Jun 10, 2009 at 01:27:10PM +0200, Peter Holik wrote:
>>> > On Wed, Jun 10, 2009 at 11:52:34AM +0200, Peter Holik wrote:
>>> >> >>> On Thu, Jun 04, 2009 at 10:40:55AM +0200, Peter Holik wrote:
>>> >> >>>> > 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.
>>> >> >>>
>>> >> >>> ff_combine_frame() does not always memcpy, your code does always
>>> >> >>
>>> >> >> change my patch to only memcpy when input is not exactly a png frame.
>>> >> >>
>>> >> >> cu Peter
>>> >>
>>> >> ping
>>> >
>>> > does your code do something else than ff_combine_frame() ?
>>> > if no its duplicate code
>>> > if yes its either worse or ff_combine_frame() should be improved
>>> >
>>>
>>> 1. ff_combine_frame is overhead
>>
>> elaborate
>
> there is more done than i need (pc->overread, pc->state,...)
>
>>> 2. i use a variable from PasteContext also altered by ff_combine_frame
>>
>> which variable is how used by you so that it conflicts with ff_combine_frame?
>
> * i use pc->last_index to remember where to parse from, because i do not like
> to parse every time from the beginning.
>
> * to return a packet starting with a png signature i skip the garbage if any
> and need this information in my combine_frame version.
>
>
> well, i will try to use ff_combine_frame()
now with ff_combine_frame()
cu Peter
-------------- next part --------------
A non-text attachment was scrubbed...
Name: png-parser.patch
Type: text/x-diff
Size: 5887 bytes
Desc: not available
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20090610/b9af1b8f/attachment.patch>
More information about the ffmpeg-devel
mailing list