1
0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-11-21 10:55:51 +02:00
FFmpeg/libavfilter/buffersrc.c
2012-06-26 19:08:13 +02:00

385 lines
12 KiB
C

/*
* Copyright (c) 2008 Vitor Sessak
*
* 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
* memory buffer source filter
*/
#include "audio.h"
#include "avfilter.h"
#include "buffersrc.h"
#include "formats.h"
#include "internal.h"
#include "video.h"
#include "libavutil/audioconvert.h"
#include "libavutil/fifo.h"
#include "libavutil/imgutils.h"
#include "libavutil/opt.h"
#include "libavutil/samplefmt.h"
typedef struct {
const AVClass *class;
AVFifoBuffer *fifo;
AVRational time_base; ///< time_base to set in the output link
/* video only */
int h, w;
enum PixelFormat pix_fmt;
AVRational pixel_aspect;
/* audio only */
int sample_rate;
enum AVSampleFormat sample_fmt;
char *sample_fmt_str;
uint64_t channel_layout;
char *channel_layout_str;
int eof;
} BufferSourceContext;
#define CHECK_VIDEO_PARAM_CHANGE(s, c, width, height, format)\
if (c->w != width || c->h != height || c->pix_fmt != format) {\
av_log(s, AV_LOG_ERROR, "Changing frame properties on the fly is not supported.\n");\
return AVERROR(EINVAL);\
}
#define CHECK_AUDIO_PARAM_CHANGE(s, c, srate, ch_layout, format)\
if (c->sample_fmt != format || c->sample_rate != srate ||\
c->channel_layout != ch_layout) {\
av_log(s, AV_LOG_ERROR, "Changing frame properties on the fly is not supported.\n");\
return AVERROR(EINVAL);\
}
int av_buffersrc_write_frame(AVFilterContext *buffer_filter, AVFrame *frame)
{
BufferSourceContext *c = buffer_filter->priv;
AVFilterBufferRef *buf;
int ret;
if (!frame) {
c->eof = 1;
return 0;
} else if (c->eof)
return AVERROR(EINVAL);
if (!av_fifo_space(c->fifo) &&
(ret = av_fifo_realloc2(c->fifo, av_fifo_size(c->fifo) +
sizeof(buf))) < 0)
return ret;
switch (buffer_filter->outputs[0]->type) {
case AVMEDIA_TYPE_VIDEO:
CHECK_VIDEO_PARAM_CHANGE(buffer_filter, c, frame->width, frame->height,
frame->format);
buf = ff_get_video_buffer(buffer_filter->outputs[0], AV_PERM_WRITE,
c->w, c->h);
av_image_copy(buf->data, buf->linesize, frame->data, frame->linesize,
c->pix_fmt, c->w, c->h);
break;
case AVMEDIA_TYPE_AUDIO:
CHECK_AUDIO_PARAM_CHANGE(buffer_filter, c, frame->sample_rate, frame->channel_layout,
frame->format);
buf = ff_get_audio_buffer(buffer_filter->outputs[0], AV_PERM_WRITE,
frame->nb_samples);
av_samples_copy(buf->extended_data, frame->extended_data,
0, 0, frame->nb_samples,
av_get_channel_layout_nb_channels(frame->channel_layout),
frame->format);
break;
default:
return AVERROR(EINVAL);
}
avfilter_copy_frame_props(buf, frame);
if ((ret = av_fifo_generic_write(c->fifo, &buf, sizeof(buf), NULL)) < 0) {
avfilter_unref_buffer(buf);
return ret;
}
return 0;
}
int av_buffersrc_buffer(AVFilterContext *s, AVFilterBufferRef *buf)
{
BufferSourceContext *c = s->priv;
int ret;
if (!buf) {
c->eof = 1;
return 0;
} else if (c->eof)
return AVERROR(EINVAL);
if (!av_fifo_space(c->fifo) &&
(ret = av_fifo_realloc2(c->fifo, av_fifo_size(c->fifo) +
sizeof(buf))) < 0)
return ret;
switch (s->outputs[0]->type) {
case AVMEDIA_TYPE_VIDEO:
CHECK_VIDEO_PARAM_CHANGE(s, c, buf->video->w, buf->video->h, buf->format);
break;
case AVMEDIA_TYPE_AUDIO:
CHECK_AUDIO_PARAM_CHANGE(s, c, buf->audio->sample_rate, buf->audio->channel_layout,
buf->format);
break;
default:
return AVERROR(EINVAL);
}
if ((ret = av_fifo_generic_write(c->fifo, &buf, sizeof(buf), NULL)) < 0)
return ret;
return 0;
}
static av_cold int init_video(AVFilterContext *ctx, const char *args)
{
BufferSourceContext *c = ctx->priv;
char pix_fmt_str[128];
int n = 0;
if (!args ||
(n = sscanf(args, "%d:%d:%127[^:]:%d:%d:%d:%d", &c->w, &c->h, pix_fmt_str,
&c->time_base.num, &c->time_base.den,
&c->pixel_aspect.num, &c->pixel_aspect.den)) != 7) {
av_log(ctx, AV_LOG_ERROR, "Expected 7 arguments, but %d found in '%s'\n", n, args);
return AVERROR(EINVAL);
}
if ((c->pix_fmt = av_get_pix_fmt(pix_fmt_str)) == PIX_FMT_NONE) {
char *tail;
c->pix_fmt = strtol(pix_fmt_str, &tail, 10);
if (*tail || c->pix_fmt < 0 || c->pix_fmt >= PIX_FMT_NB) {
av_log(ctx, AV_LOG_ERROR, "Invalid pixel format string '%s'\n", pix_fmt_str);
return AVERROR(EINVAL);
}
}
if (!(c->fifo = av_fifo_alloc(sizeof(AVFilterBufferRef*))))
return AVERROR(ENOMEM);
av_log(ctx, AV_LOG_VERBOSE, "w:%d h:%d pixfmt:%s\n", c->w, c->h, av_pix_fmt_descriptors[c->pix_fmt].name);
return 0;
}
#define OFFSET(x) offsetof(BufferSourceContext, x)
#define A AV_OPT_FLAG_AUDIO_PARAM
static const AVOption audio_options[] = {
{ "time_base", NULL, OFFSET(time_base), AV_OPT_TYPE_RATIONAL, { 0 }, 0, INT_MAX, A },
{ "sample_rate", NULL, OFFSET(sample_rate), AV_OPT_TYPE_INT, { 0 }, 0, INT_MAX, A },
{ "sample_fmt", NULL, OFFSET(sample_fmt_str), AV_OPT_TYPE_STRING, .flags = A },
{ "channel_layout", NULL, OFFSET(channel_layout_str), AV_OPT_TYPE_STRING, .flags = A },
{ NULL },
};
static const AVClass abuffer_class = {
.class_name = "abuffer source",
.item_name = av_default_item_name,
.option = audio_options,
.version = LIBAVUTIL_VERSION_INT,
};
static av_cold int init_audio(AVFilterContext *ctx, const char *args)
{
BufferSourceContext *s = ctx->priv;
int ret = 0;
s->class = &abuffer_class;
av_opt_set_defaults(s);
if ((ret = av_set_options_string(s, args, "=", ":")) < 0) {
av_log(ctx, AV_LOG_ERROR, "Error parsing options string: %s.\n", args);
goto fail;
}
s->sample_fmt = av_get_sample_fmt(s->sample_fmt_str);
if (s->sample_fmt == AV_SAMPLE_FMT_NONE) {
av_log(ctx, AV_LOG_ERROR, "Invalid sample format %s.\n",
s->sample_fmt_str);
ret = AVERROR(EINVAL);
goto fail;
}
s->channel_layout = av_get_channel_layout(s->channel_layout_str);
if (!s->channel_layout) {
av_log(ctx, AV_LOG_ERROR, "Invalid channel layout %s.\n",
s->channel_layout_str);
ret = AVERROR(EINVAL);
goto fail;
}
if (!(s->fifo = av_fifo_alloc(sizeof(AVFilterBufferRef*)))) {
ret = AVERROR(ENOMEM);
goto fail;
}
if (!s->time_base.num)
s->time_base = (AVRational){1, s->sample_rate};
av_log(ctx, AV_LOG_VERBOSE, "tb:%d/%d samplefmt:%s samplerate: %d "
"ch layout:%s\n", s->time_base.num, s->time_base.den, s->sample_fmt_str,
s->sample_rate, s->channel_layout_str);
fail:
av_opt_free(s);
return ret;
}
static av_cold void uninit(AVFilterContext *ctx)
{
BufferSourceContext *s = ctx->priv;
while (s->fifo && av_fifo_size(s->fifo)) {
AVFilterBufferRef *buf;
av_fifo_generic_read(s->fifo, &buf, sizeof(buf), NULL);
avfilter_unref_buffer(buf);
}
av_fifo_free(s->fifo);
s->fifo = NULL;
}
static int query_formats(AVFilterContext *ctx)
{
BufferSourceContext *c = ctx->priv;
AVFilterChannelLayouts *channel_layouts = NULL;
AVFilterFormats *formats = NULL;
AVFilterFormats *samplerates = NULL;
switch (ctx->outputs[0]->type) {
case AVMEDIA_TYPE_VIDEO:
ff_add_format(&formats, c->pix_fmt);
ff_set_common_formats(ctx, formats);
break;
case AVMEDIA_TYPE_AUDIO:
ff_add_format(&formats, c->sample_fmt);
ff_set_common_formats(ctx, formats);
ff_add_format(&samplerates, c->sample_rate);
ff_set_common_samplerates(ctx, samplerates);
ff_add_channel_layout(&channel_layouts, c->channel_layout);
ff_set_common_channel_layouts(ctx, channel_layouts);
break;
default:
return AVERROR(EINVAL);
}
return 0;
}
static int config_props(AVFilterLink *link)
{
BufferSourceContext *c = link->src->priv;
switch (link->type) {
case AVMEDIA_TYPE_VIDEO:
link->w = c->w;
link->h = c->h;
link->sample_aspect_ratio = c->pixel_aspect;
break;
case AVMEDIA_TYPE_AUDIO:
link->channel_layout = c->channel_layout;
link->sample_rate = c->sample_rate;
break;
default:
return AVERROR(EINVAL);
}
link->time_base = c->time_base;
return 0;
}
static int request_frame(AVFilterLink *link)
{
BufferSourceContext *c = link->src->priv;
AVFilterBufferRef *buf;
if (!av_fifo_size(c->fifo)) {
if (c->eof)
return AVERROR_EOF;
return AVERROR(EAGAIN);
}
av_fifo_generic_read(c->fifo, &buf, sizeof(buf), NULL);
switch (link->type) {
case AVMEDIA_TYPE_VIDEO:
ff_start_frame(link, avfilter_ref_buffer(buf, ~0));
ff_draw_slice(link, 0, link->h, 1);
ff_end_frame(link);
break;
case AVMEDIA_TYPE_AUDIO:
ff_filter_samples(link, avfilter_ref_buffer(buf, ~0));
break;
default:
return AVERROR(EINVAL);
}
avfilter_unref_buffer(buf);
return 0;
}
static int poll_frame(AVFilterLink *link)
{
BufferSourceContext *c = link->src->priv;
int size = av_fifo_size(c->fifo);
if (!size && c->eof)
return AVERROR_EOF;
return size/sizeof(AVFilterBufferRef*);
}
AVFilter avfilter_vsrc_buffer = {
.name = "buffer",
.description = NULL_IF_CONFIG_SMALL("Buffer video frames, and make them accessible to the filterchain."),
.priv_size = sizeof(BufferSourceContext),
.query_formats = query_formats,
.init = init_video,
.uninit = uninit,
.inputs = (AVFilterPad[]) {{ .name = NULL }},
.outputs = (AVFilterPad[]) {{ .name = "default",
.type = AVMEDIA_TYPE_VIDEO,
.request_frame = request_frame,
.poll_frame = poll_frame,
.config_props = config_props, },
{ .name = NULL}},
};
AVFilter avfilter_asrc_abuffer = {
.name = "abuffer",
.description = NULL_IF_CONFIG_SMALL("Buffer audio frames, and make them accessible to the filterchain."),
.priv_size = sizeof(BufferSourceContext),
.query_formats = query_formats,
.init = init_audio,
.uninit = uninit,
.inputs = (AVFilterPad[]) {{ .name = NULL }},
.outputs = (AVFilterPad[]) {{ .name = "default",
.type = AVMEDIA_TYPE_AUDIO,
.request_frame = request_frame,
.poll_frame = poll_frame,
.config_props = config_props, },
{ .name = NULL}},
};