mirror of
https://git.ffmpeg.org/ffmpeg.git
synced 2024-12-30 11:22:14 +00:00
57231e4d5b
Signed-off-by: Paul B Mahol <onemda@gmail.com> Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>
129 lines
4.3 KiB
C
129 lines
4.3 KiB
C
/*
|
|
* TAK parser
|
|
* Copyright (c) 2012 Michael Niedermayer
|
|
*
|
|
* This file is part of Libav.
|
|
*
|
|
* Libav 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.
|
|
*
|
|
* Libav 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 Libav; if not, write to the Free Software
|
|
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
|
|
*/
|
|
|
|
/**
|
|
* @file
|
|
* TAK parser
|
|
**/
|
|
|
|
#include "tak.h"
|
|
#include "parser.h"
|
|
|
|
typedef struct TAKParseContext {
|
|
ParseContext pc;
|
|
TAKStreamInfo ti;
|
|
int index;
|
|
} TAKParseContext;
|
|
|
|
static av_cold int tak_init(AVCodecParserContext *s)
|
|
{
|
|
ff_tak_init_crc();
|
|
return 0;
|
|
}
|
|
|
|
static int tak_parse(AVCodecParserContext *s, AVCodecContext *avctx,
|
|
const uint8_t **poutbuf, int *poutbuf_size,
|
|
const uint8_t *buf, int buf_size)
|
|
{
|
|
TAKParseContext *t = s->priv_data;
|
|
ParseContext *pc = &t->pc;
|
|
int next = END_NOT_FOUND;
|
|
GetBitContext gb;
|
|
int consumed = 0;
|
|
int needed = buf_size ? TAK_MAX_FRAME_HEADER_BYTES : 8;
|
|
|
|
if (s->flags & PARSER_FLAG_COMPLETE_FRAMES) {
|
|
TAKStreamInfo ti;
|
|
init_get_bits(&gb, buf, buf_size);
|
|
if (!ff_tak_decode_frame_header(avctx, &gb, &ti, 127))
|
|
s->duration = t->ti.last_frame_samples ? t->ti.last_frame_samples
|
|
: t->ti.frame_samples;
|
|
*poutbuf = buf;
|
|
*poutbuf_size = buf_size;
|
|
return buf_size;
|
|
}
|
|
|
|
while (buf_size || t->index + needed <= pc->index) {
|
|
if (buf_size && t->index + TAK_MAX_FRAME_HEADER_BYTES > pc->index) {
|
|
int tmp_buf_size = FFMIN(2 * TAK_MAX_FRAME_HEADER_BYTES,
|
|
buf_size);
|
|
const uint8_t *tmp_buf = buf;
|
|
|
|
ff_combine_frame(pc, END_NOT_FOUND, &tmp_buf, &tmp_buf_size);
|
|
consumed += tmp_buf_size;
|
|
buf += tmp_buf_size;
|
|
buf_size -= tmp_buf_size;
|
|
}
|
|
|
|
for (; t->index + needed <= pc->index; t->index++)
|
|
if (pc->buffer[t->index] == 0xFF &&
|
|
pc->buffer[t->index + 1] == 0xA0) {
|
|
TAKStreamInfo ti;
|
|
|
|
init_get_bits(&gb, pc->buffer + t->index,
|
|
8 * (pc->index - t->index));
|
|
if (!ff_tak_decode_frame_header(avctx, &gb,
|
|
pc->frame_start_found ? &ti
|
|
: &t->ti,
|
|
127) &&
|
|
!ff_tak_check_crc(pc->buffer + t->index,
|
|
get_bits_count(&gb) / 8)) {
|
|
if (!pc->frame_start_found) {
|
|
pc->frame_start_found = 1;
|
|
s->duration = t->ti.last_frame_samples ?
|
|
t->ti.last_frame_samples :
|
|
t->ti.frame_samples;
|
|
} else {
|
|
pc->frame_start_found = 0;
|
|
next = t->index - pc->index;
|
|
t->index = 0;
|
|
goto found;
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
found:
|
|
if (consumed && !buf_size && next == END_NOT_FOUND ||
|
|
ff_combine_frame(pc, next, &buf, &buf_size) < 0) {
|
|
*poutbuf = NULL;
|
|
*poutbuf_size = 0;
|
|
return buf_size + consumed;
|
|
}
|
|
|
|
if (next != END_NOT_FOUND) {
|
|
next += consumed;
|
|
pc->overread = FFMAX(0, -next);
|
|
}
|
|
|
|
*poutbuf = buf;
|
|
*poutbuf_size = buf_size;
|
|
return next;
|
|
}
|
|
|
|
AVCodecParser ff_tak_parser = {
|
|
.codec_ids = { AV_CODEC_ID_TAK },
|
|
.priv_data_size = sizeof(TAKParseContext),
|
|
.parser_init = tak_init,
|
|
.parser_parse = tak_parse,
|
|
.parser_close = ff_parse_close,
|
|
};
|