[FFmpeg-devel] [PATCH] png parser
Michael Niedermayer
michaelni
Thu Jun 11 01:32:15 CEST 2009
cOn Wed, Jun 10, 2009 at 04:33:48PM +0200, Peter Holik wrote:
> >> 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()
well yes, but ive meant it should be used like also in the other parsers,
but lets rather look at actual problems of your approuch ... if you can
fix them and your code is simpler i wouldnt mind it but i doubt it
example:
2 packets
packet 1: frame 0, frame 1, fra
packet 2: me 2, frame 3 frame 4
your approuch will memcpy the whole 2nd packet, the normal ff_combine_frame
code will copy only "frame 2"
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
No human being will ever know the Truth, for even if they happen to say it
by chance, they would not even known they had done so. -- Xenophanes
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20090611/12c6c68a/attachment.pgp>
More information about the ffmpeg-devel
mailing list