2015-12-07 17:45:52 +02:00
|
|
|
/*
|
|
|
|
* Copyright (c) 2015 Paul B. Mahol
|
|
|
|
*
|
|
|
|
* 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
|
|
|
|
*/
|
|
|
|
|
2023-12-22 23:25:25 +02:00
|
|
|
#include "libavutil/avassert.h"
|
2015-12-07 17:45:52 +02:00
|
|
|
#include "libavutil/eval.h"
|
|
|
|
#include "libavutil/imgutils.h"
|
2024-03-25 02:30:37 +02:00
|
|
|
#include "libavutil/mem.h"
|
2015-12-07 17:45:52 +02:00
|
|
|
#include "libavutil/opt.h"
|
|
|
|
|
|
|
|
#include "avfilter.h"
|
|
|
|
#include "formats.h"
|
|
|
|
#include "internal.h"
|
|
|
|
#include "video.h"
|
|
|
|
|
|
|
|
typedef struct SwapRectContext {
|
|
|
|
const AVClass *class;
|
|
|
|
char *w, *h;
|
|
|
|
char *x1, *y1;
|
|
|
|
char *x2, *y2;
|
|
|
|
|
|
|
|
int nb_planes;
|
|
|
|
int pixsteps[4];
|
|
|
|
|
|
|
|
const AVPixFmtDescriptor *desc;
|
|
|
|
uint8_t *temp;
|
|
|
|
} SwapRectContext;
|
|
|
|
|
|
|
|
#define OFFSET(x) offsetof(SwapRectContext, x)
|
2021-02-09 12:38:37 +02:00
|
|
|
#define FLAGS AV_OPT_FLAG_VIDEO_PARAM|AV_OPT_FLAG_FILTERING_PARAM|AV_OPT_FLAG_RUNTIME_PARAM
|
2015-12-07 17:45:52 +02:00
|
|
|
static const AVOption swaprect_options[] = {
|
|
|
|
{ "w", "set rect width", OFFSET(w), AV_OPT_TYPE_STRING, {.str="w/2"}, 0, 0, .flags = FLAGS },
|
|
|
|
{ "h", "set rect height", OFFSET(h), AV_OPT_TYPE_STRING, {.str="h/2"}, 0, 0, .flags = FLAGS },
|
|
|
|
{ "x1", "set 1st rect x top left coordinate", OFFSET(x1), AV_OPT_TYPE_STRING, {.str="w/2"}, 0, 0, .flags = FLAGS },
|
|
|
|
{ "y1", "set 1st rect y top left coordinate", OFFSET(y1), AV_OPT_TYPE_STRING, {.str="h/2"}, 0, 0, .flags = FLAGS },
|
|
|
|
{ "x2", "set 2nd rect x top left coordinate", OFFSET(x2), AV_OPT_TYPE_STRING, {.str="0"}, 0, 0, .flags = FLAGS },
|
|
|
|
{ "y2", "set 2nd rect y top left coordinate", OFFSET(y2), AV_OPT_TYPE_STRING, {.str="0"}, 0, 0, .flags = FLAGS },
|
|
|
|
{ NULL },
|
|
|
|
};
|
|
|
|
|
|
|
|
AVFILTER_DEFINE_CLASS(swaprect);
|
|
|
|
|
|
|
|
static int query_formats(AVFilterContext *ctx)
|
|
|
|
{
|
2021-09-26 00:09:16 +02:00
|
|
|
int reject_flags = AV_PIX_FMT_FLAG_PAL |
|
|
|
|
AV_PIX_FMT_FLAG_HWACCEL |
|
|
|
|
AV_PIX_FMT_FLAG_BITSTREAM;
|
2015-12-07 17:45:52 +02:00
|
|
|
|
2021-09-26 00:09:16 +02:00
|
|
|
return ff_set_common_formats(ctx, ff_formats_pixdesc_filter(0, reject_flags));
|
2015-12-07 17:45:52 +02:00
|
|
|
}
|
|
|
|
|
lavu/frame: deprecate AVFrame.pkt_{pos,size}
These fields are supposed to store information about the packet the
frame was decoded from, specifically the byte offset it was stored at
and its size.
However,
- the fields are highly ad-hoc - there is no strong reason why
specifically those (and not any other) packet properties should have a
dedicated field in AVFrame; unlike e.g. the timestamps, there is no
fundamental link between coded packet offset/size and decoded frames
- they only make sense for frames produced by decoding demuxed packets,
and even then it is not always the case that the encoded data was
stored in the file as a contiguous sequence of bytes (in order for pos
to be well-defined)
- pkt_pos was added without much explanation, apparently to allow
passthrough of this information through lavfi in order to handle byte
seeking in ffplay. That is now implemented using arbitrary user data
passthrough in AVFrame.opaque_ref.
- several filters use pkt_pos as a variable available to user-supplied
expressions, but there seems to be no established motivation for using them.
- pkt_size was added for use in ffprobe, but that too is now handled
without using this field. Additonally, the values of this field
produced by libavcodec are flawed, as described in the previous
ffprobe conversion commit.
In summary - these fields are ill-defined and insufficiently motivated,
so deprecate them.
2023-03-10 11:48:34 +02:00
|
|
|
static const char *const var_names[] = { "w", "h", "a", "n", "t",
|
|
|
|
#if FF_API_FRAME_PKT
|
|
|
|
"pos",
|
|
|
|
#endif
|
|
|
|
"sar", "dar", NULL };
|
|
|
|
enum { VAR_W, VAR_H, VAR_A, VAR_N, VAR_T,
|
|
|
|
#if FF_API_FRAME_PKT
|
|
|
|
VAR_POS,
|
|
|
|
#endif
|
|
|
|
VAR_SAR, VAR_DAR, VAR_VARS_NB };
|
2015-12-07 17:45:52 +02:00
|
|
|
|
|
|
|
static int filter_frame(AVFilterLink *inlink, AVFrame *in)
|
|
|
|
{
|
|
|
|
AVFilterContext *ctx = inlink->dst;
|
|
|
|
AVFilterLink *outlink = ctx->outputs[0];
|
|
|
|
SwapRectContext *s = ctx->priv;
|
|
|
|
double var_values[VAR_VARS_NB];
|
|
|
|
int x1[4], y1[4];
|
|
|
|
int x2[4], y2[4];
|
|
|
|
int aw[4], ah[4];
|
|
|
|
int lw[4], lh[4];
|
|
|
|
int pw[4], ph[4];
|
|
|
|
double dw, dh;
|
|
|
|
double dx1, dy1;
|
|
|
|
double dx2, dy2;
|
|
|
|
int y, p, w, h, ret;
|
|
|
|
|
|
|
|
var_values[VAR_W] = inlink->w;
|
|
|
|
var_values[VAR_H] = inlink->h;
|
|
|
|
var_values[VAR_A] = (float) inlink->w / inlink->h;
|
|
|
|
var_values[VAR_SAR] = inlink->sample_aspect_ratio.num ? av_q2d(inlink->sample_aspect_ratio) : 1;
|
|
|
|
var_values[VAR_DAR] = var_values[VAR_A] * var_values[VAR_SAR];
|
2016-08-30 15:28:41 +02:00
|
|
|
var_values[VAR_N] = inlink->frame_count_out;
|
2015-12-07 17:45:52 +02:00
|
|
|
var_values[VAR_T] = in->pts == AV_NOPTS_VALUE ? NAN : in->pts * av_q2d(inlink->time_base);
|
lavu/frame: deprecate AVFrame.pkt_{pos,size}
These fields are supposed to store information about the packet the
frame was decoded from, specifically the byte offset it was stored at
and its size.
However,
- the fields are highly ad-hoc - there is no strong reason why
specifically those (and not any other) packet properties should have a
dedicated field in AVFrame; unlike e.g. the timestamps, there is no
fundamental link between coded packet offset/size and decoded frames
- they only make sense for frames produced by decoding demuxed packets,
and even then it is not always the case that the encoded data was
stored in the file as a contiguous sequence of bytes (in order for pos
to be well-defined)
- pkt_pos was added without much explanation, apparently to allow
passthrough of this information through lavfi in order to handle byte
seeking in ffplay. That is now implemented using arbitrary user data
passthrough in AVFrame.opaque_ref.
- several filters use pkt_pos as a variable available to user-supplied
expressions, but there seems to be no established motivation for using them.
- pkt_size was added for use in ffprobe, but that too is now handled
without using this field. Additonally, the values of this field
produced by libavcodec are flawed, as described in the previous
ffprobe conversion commit.
In summary - these fields are ill-defined and insufficiently motivated,
so deprecate them.
2023-03-10 11:48:34 +02:00
|
|
|
#if FF_API_FRAME_PKT
|
|
|
|
FF_DISABLE_DEPRECATION_WARNINGS
|
2020-01-15 10:39:07 +02:00
|
|
|
var_values[VAR_POS] = in->pkt_pos == -1 ? NAN : in->pkt_pos;
|
lavu/frame: deprecate AVFrame.pkt_{pos,size}
These fields are supposed to store information about the packet the
frame was decoded from, specifically the byte offset it was stored at
and its size.
However,
- the fields are highly ad-hoc - there is no strong reason why
specifically those (and not any other) packet properties should have a
dedicated field in AVFrame; unlike e.g. the timestamps, there is no
fundamental link between coded packet offset/size and decoded frames
- they only make sense for frames produced by decoding demuxed packets,
and even then it is not always the case that the encoded data was
stored in the file as a contiguous sequence of bytes (in order for pos
to be well-defined)
- pkt_pos was added without much explanation, apparently to allow
passthrough of this information through lavfi in order to handle byte
seeking in ffplay. That is now implemented using arbitrary user data
passthrough in AVFrame.opaque_ref.
- several filters use pkt_pos as a variable available to user-supplied
expressions, but there seems to be no established motivation for using them.
- pkt_size was added for use in ffprobe, but that too is now handled
without using this field. Additonally, the values of this field
produced by libavcodec are flawed, as described in the previous
ffprobe conversion commit.
In summary - these fields are ill-defined and insufficiently motivated,
so deprecate them.
2023-03-10 11:48:34 +02:00
|
|
|
FF_ENABLE_DEPRECATION_WARNINGS
|
|
|
|
#endif
|
2015-12-07 17:45:52 +02:00
|
|
|
|
|
|
|
ret = av_expr_parse_and_eval(&dw, s->w,
|
|
|
|
var_names, &var_values[0],
|
|
|
|
NULL, NULL, NULL, NULL,
|
|
|
|
0, 0, ctx);
|
|
|
|
if (ret < 0)
|
|
|
|
return ret;
|
|
|
|
|
|
|
|
ret = av_expr_parse_and_eval(&dh, s->h,
|
|
|
|
var_names, &var_values[0],
|
|
|
|
NULL, NULL, NULL, NULL,
|
|
|
|
0, 0, ctx);
|
|
|
|
if (ret < 0)
|
|
|
|
return ret;
|
|
|
|
|
|
|
|
ret = av_expr_parse_and_eval(&dx1, s->x1,
|
|
|
|
var_names, &var_values[0],
|
|
|
|
NULL, NULL, NULL, NULL,
|
|
|
|
0, 0, ctx);
|
|
|
|
if (ret < 0)
|
|
|
|
return ret;
|
|
|
|
|
|
|
|
ret = av_expr_parse_and_eval(&dy1, s->y1,
|
|
|
|
var_names, &var_values[0],
|
|
|
|
NULL, NULL, NULL, NULL,
|
|
|
|
0, 0, ctx);
|
|
|
|
if (ret < 0)
|
|
|
|
return ret;
|
|
|
|
|
|
|
|
ret = av_expr_parse_and_eval(&dx2, s->x2,
|
|
|
|
var_names, &var_values[0],
|
|
|
|
NULL, NULL, NULL, NULL,
|
|
|
|
0, 0, ctx);
|
|
|
|
if (ret < 0)
|
|
|
|
return ret;
|
|
|
|
|
|
|
|
ret = av_expr_parse_and_eval(&dy2, s->y2,
|
|
|
|
var_names, &var_values[0],
|
|
|
|
NULL, NULL, NULL, NULL,
|
|
|
|
0, 0, ctx);
|
|
|
|
if (ret < 0)
|
|
|
|
return ret;
|
|
|
|
|
|
|
|
w = dw; h = dh; x1[0] = dx1; y1[0] = dy1; x2[0] = dx2; y2[0] = dy2;
|
|
|
|
|
|
|
|
x1[0] = av_clip(x1[0], 0, inlink->w - 1);
|
2023-12-22 23:26:22 +02:00
|
|
|
y1[0] = av_clip(y1[0], 0, inlink->h - 1);
|
2015-12-07 17:45:52 +02:00
|
|
|
|
|
|
|
x2[0] = av_clip(x2[0], 0, inlink->w - 1);
|
2023-12-22 23:26:22 +02:00
|
|
|
y2[0] = av_clip(y2[0], 0, inlink->h - 1);
|
2015-12-07 17:45:52 +02:00
|
|
|
|
2018-05-08 12:44:57 +02:00
|
|
|
ah[1] = ah[2] = AV_CEIL_RSHIFT(h, s->desc->log2_chroma_h);
|
2015-12-07 17:45:52 +02:00
|
|
|
ah[0] = ah[3] = h;
|
2018-05-08 12:44:57 +02:00
|
|
|
aw[1] = aw[2] = AV_CEIL_RSHIFT(w, s->desc->log2_chroma_w);
|
2015-12-07 17:45:52 +02:00
|
|
|
aw[0] = aw[3] = w;
|
|
|
|
|
|
|
|
w = FFMIN3(w, inlink->w - x1[0], inlink->w - x2[0]);
|
|
|
|
h = FFMIN3(h, inlink->h - y1[0], inlink->h - y2[0]);
|
|
|
|
|
2018-05-08 12:44:57 +02:00
|
|
|
ph[1] = ph[2] = AV_CEIL_RSHIFT(h, s->desc->log2_chroma_h);
|
2015-12-07 17:45:52 +02:00
|
|
|
ph[0] = ph[3] = h;
|
2018-05-08 12:44:57 +02:00
|
|
|
pw[1] = pw[2] = AV_CEIL_RSHIFT(w, s->desc->log2_chroma_w);
|
2015-12-07 17:45:52 +02:00
|
|
|
pw[0] = pw[3] = w;
|
|
|
|
|
2018-05-08 12:44:57 +02:00
|
|
|
lh[1] = lh[2] = AV_CEIL_RSHIFT(inlink->h, s->desc->log2_chroma_h);
|
2015-12-07 17:45:52 +02:00
|
|
|
lh[0] = lh[3] = inlink->h;
|
2018-05-08 12:44:57 +02:00
|
|
|
lw[1] = lw[2] = AV_CEIL_RSHIFT(inlink->w, s->desc->log2_chroma_w);
|
2015-12-07 17:45:52 +02:00
|
|
|
lw[0] = lw[3] = inlink->w;
|
|
|
|
|
2023-12-22 23:27:08 +02:00
|
|
|
x1[1] = x1[2] = (x1[0] >> s->desc->log2_chroma_w);
|
2015-12-07 17:45:52 +02:00
|
|
|
x1[0] = x1[3] = x1[0];
|
2023-12-22 23:27:08 +02:00
|
|
|
y1[1] = y1[2] = (y1[0] >> s->desc->log2_chroma_h);
|
2015-12-07 17:45:52 +02:00
|
|
|
y1[0] = y1[3] = y1[0];
|
|
|
|
|
2023-12-22 23:27:08 +02:00
|
|
|
x2[1] = x2[2] = (x2[0] >> s->desc->log2_chroma_w);
|
2015-12-07 17:45:52 +02:00
|
|
|
x2[0] = x2[3] = x2[0];
|
2023-12-22 23:27:08 +02:00
|
|
|
y2[1] = y2[2] = (y2[0] >> s->desc->log2_chroma_h);
|
2015-12-07 17:45:52 +02:00
|
|
|
y2[0] = y2[3] = y2[0];
|
|
|
|
|
2023-12-22 23:25:25 +02:00
|
|
|
|
|
|
|
av_assert0(FFMAX(x1[1], x2[1]) + pw[1] <= lw[1]);
|
|
|
|
av_assert0(FFMAX(y1[1], y2[1]) + ph[1] <= lh[1]);
|
|
|
|
|
2015-12-07 17:45:52 +02:00
|
|
|
for (p = 0; p < s->nb_planes; p++) {
|
|
|
|
if (ph[p] == ah[p] && pw[p] == aw[p]) {
|
|
|
|
uint8_t *src = in->data[p] + y1[p] * in->linesize[p] + x1[p] * s->pixsteps[p];
|
|
|
|
uint8_t *dst = in->data[p] + y2[p] * in->linesize[p] + x2[p] * s->pixsteps[p];
|
|
|
|
|
|
|
|
for (y = 0; y < ph[p]; y++) {
|
|
|
|
memcpy(s->temp, src, pw[p] * s->pixsteps[p]);
|
|
|
|
memmove(src, dst, pw[p] * s->pixsteps[p]);
|
|
|
|
memcpy(dst, s->temp, pw[p] * s->pixsteps[p]);
|
|
|
|
src += in->linesize[p];
|
|
|
|
dst += in->linesize[p];
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return ff_filter_frame(outlink, in);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int config_input(AVFilterLink *inlink)
|
|
|
|
{
|
|
|
|
AVFilterContext *ctx = inlink->dst;
|
|
|
|
SwapRectContext *s = ctx->priv;
|
|
|
|
|
|
|
|
if (!s->w || !s->h ||
|
|
|
|
!s->x1 || !s->y1 ||
|
|
|
|
!s->x2 || !s->y2)
|
|
|
|
return AVERROR(EINVAL);
|
|
|
|
|
|
|
|
s->desc = av_pix_fmt_desc_get(inlink->format);
|
|
|
|
av_image_fill_max_pixsteps(s->pixsteps, NULL, s->desc);
|
|
|
|
s->nb_planes = av_pix_fmt_count_planes(inlink->format);
|
|
|
|
|
|
|
|
s->temp = av_malloc_array(inlink->w, s->pixsteps[0]);
|
|
|
|
if (!s->temp)
|
|
|
|
return AVERROR(ENOMEM);
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static av_cold void uninit(AVFilterContext *ctx)
|
|
|
|
{
|
|
|
|
SwapRectContext *s = ctx->priv;
|
|
|
|
av_freep(&s->temp);
|
|
|
|
}
|
|
|
|
|
|
|
|
static const AVFilterPad inputs[] = {
|
|
|
|
{
|
|
|
|
.name = "default",
|
|
|
|
.type = AVMEDIA_TYPE_VIDEO,
|
2021-08-16 00:07:41 +02:00
|
|
|
.flags = AVFILTERPAD_FLAG_NEEDS_WRITABLE,
|
2015-12-07 17:45:52 +02:00
|
|
|
.filter_frame = filter_frame,
|
|
|
|
.config_props = config_input,
|
|
|
|
},
|
|
|
|
};
|
|
|
|
|
2021-04-19 18:33:56 +02:00
|
|
|
const AVFilter ff_vf_swaprect = {
|
2015-12-07 17:45:52 +02:00
|
|
|
.name = "swaprect",
|
|
|
|
.description = NULL_IF_CONFIG_SMALL("Swap 2 rectangular objects in video."),
|
|
|
|
.priv_size = sizeof(SwapRectContext),
|
|
|
|
.priv_class = &swaprect_class,
|
|
|
|
.uninit = uninit,
|
2021-08-12 13:05:31 +02:00
|
|
|
FILTER_INPUTS(inputs),
|
2023-08-03 14:37:51 +02:00
|
|
|
FILTER_OUTPUTS(ff_video_default_filterpad),
|
2021-09-27 12:07:35 +02:00
|
|
|
FILTER_QUERY_FUNC(query_formats),
|
2016-02-08 16:43:09 +02:00
|
|
|
.flags = AVFILTER_FLAG_SUPPORT_TIMELINE_GENERIC,
|
2021-02-09 12:38:37 +02:00
|
|
|
.process_command = ff_filter_process_command,
|
2015-12-07 17:45:52 +02:00
|
|
|
};
|