1
0
mirror of https://github.com/mpv-player/mpv synced 2024-12-30 02:52:10 +00:00
mpv/common/common.c
wm4 76276c9210 video: rewrite filtering glue code
Get rid of the old vf.c code. Replace it with a generic filtering
framework, which can potentially handle more than just --vf. At least
reimplementing --af with this code is planned.

This changes some --vf semantics (including runtime behavior and the
"vf" command). The most important ones are listed in interface-changes.

vf_convert.c is renamed to f_swscale.c. It is now an internal filter
that can not be inserted by the user manually.

f_lavfi.c is a refactor of player/lavfi.c. The latter will be removed
once --lavfi-complex is reimplemented on top of f_lavfi.c. (which is
conceptually easy, but a big mess due to the data flow changes).

The existing filters are all changed heavily. The data flow of the new
filter framework is different. Especially EOF handling changes - EOF is
now a "frame" rather than a state, and must be passed through exactly
once.

Another major thing is that all filters must support dynamic format
changes. The filter reconfig() function goes away. (This sounds complex,
but since all filters need to handle EOF draining anyway, they can use
the same code, and it removes the mess with reconfig() having to predict
the output format, which completely breaks with libavfilter anyway.)

In addition, there is no automatic format negotiation or conversion.
libavfilter's primitive and insufficient API simply doesn't allow us to
do this in a reasonable way. Instead, filters can use f_autoconvert as
sub-filter, and tell it which formats they support. This filter will in
turn add actual conversion filters, such as f_swscale, to perform
necessary format changes.

vf_vapoursynth.c uses the same basic principle of operation as before,
but with worryingly different details in data flow. Still appears to
work.

The hardware deint filters (vf_vavpp.c, vf_d3d11vpp.c, vf_vdpaupp.c) are
heavily changed. Fortunately, they all used refqueue.c, which is for
sharing the data flow logic (especially for managing future/past
surfaces and such). It turns out it can be used to factor out most of
the data flow. Some of these filters accepted software input. Instead of
having ad-hoc upload code in each filter, surface upload is now
delegated to f_autoconvert, which can use f_hwupload to perform this.

Exporting VO capabilities is still a big mess (mp_stream_info stuff).

The D3D11 code drops the redundant image formats, and all code uses the
hw_subfmt (sw_format in FFmpeg) instead. Although that too seems to be a
big mess for now.

f_async_queue is unused.
2018-01-30 03:10:27 -08:00

316 lines
10 KiB
C

/*
* This file is part of mpv.
*
* mpv 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.
*
* mpv 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 mpv. If not, see <http://www.gnu.org/licenses/>.
*/
#include <stdarg.h>
#include <assert.h>
#include <libavutil/common.h>
#include <libavutil/error.h>
#include "mpv_talloc.h"
#include "misc/bstr.h"
#include "misc/ctype.h"
#include "common/common.h"
#include "osdep/strnlen.h"
#define appendf(ptr, ...) \
do {(*(ptr)) = talloc_asprintf_append_buffer(*(ptr), __VA_ARGS__);} while(0)
// Return a talloc'ed string formatted according to the format string in fmt.
// On error, return NULL.
// Valid formats:
// %H, %h: hour (%H is padded with 0 to two digits)
// %M: minutes from 00-59 (hours are subtracted)
// %m: total minutes (includes hours, unlike %M)
// %S: seconds from 00-59 (minutes and hours are subtracted)
// %s: total seconds (includes hours and minutes)
// %f: like %s, but as float
// %T: milliseconds (000-999)
char *mp_format_time_fmt(const char *fmt, double time)
{
if (time == MP_NOPTS_VALUE)
return talloc_strdup(NULL, "unknown");
char *sign = time < 0 ? "-" : "";
time = time < 0 ? -time : time;
long long int itime = time;
long long int h, m, tm, s;
int ms;
s = itime;
tm = s / 60;
h = s / 3600;
s -= h * 3600;
m = s / 60;
s -= m * 60;
ms = (time - itime) * 1000;
char *res = talloc_strdup(NULL, "");
while (*fmt) {
if (fmt[0] == '%') {
fmt++;
switch (fmt[0]) {
case 'h': appendf(&res, "%s%lld", sign, h); break;
case 'H': appendf(&res, "%s%02lld", sign, h); break;
case 'm': appendf(&res, "%s%lld", sign, tm); break;
case 'M': appendf(&res, "%02lld", m); break;
case 's': appendf(&res, "%s%lld", sign, itime); break;
case 'S': appendf(&res, "%02lld", s); break;
case 'T': appendf(&res, "%03d", ms); break;
case '%': appendf(&res, "%s", "%"); break;
default: goto error;
}
fmt++;
} else {
appendf(&res, "%c", *fmt);
fmt++;
}
}
return res;
error:
talloc_free(res);
return NULL;
}
char *mp_format_time(double time, bool fractions)
{
return mp_format_time_fmt(fractions ? "%H:%M:%S.%T" : "%H:%M:%S", time);
}
// Set rc to the union of rc and rc2
void mp_rect_union(struct mp_rect *rc, const struct mp_rect *rc2)
{
rc->x0 = FFMIN(rc->x0, rc2->x0);
rc->y0 = FFMIN(rc->y0, rc2->y0);
rc->x1 = FFMAX(rc->x1, rc2->x1);
rc->y1 = FFMAX(rc->y1, rc2->y1);
}
// Returns whether or not a point is contained by rc
bool mp_rect_contains(struct mp_rect *rc, int x, int y)
{
return rc->x0 <= x && x < rc->x1 && rc->y0 <= y && y < rc->y1;
}
// Set rc to the intersection of rc and src.
// Return false if the result is empty.
bool mp_rect_intersection(struct mp_rect *rc, const struct mp_rect *rc2)
{
rc->x0 = FFMAX(rc->x0, rc2->x0);
rc->y0 = FFMAX(rc->y0, rc2->y0);
rc->x1 = FFMIN(rc->x1, rc2->x1);
rc->y1 = FFMIN(rc->y1, rc2->y1);
return rc->x1 > rc->x0 && rc->y1 > rc->y0;
}
bool mp_rect_equals(struct mp_rect *rc1, struct mp_rect *rc2)
{
return rc1->x0 == rc2->x0 && rc1->y0 == rc2->y0 &&
rc1->x1 == rc2->x1 && rc1->y1 == rc2->y1;
}
// This works like snprintf(), except that it starts writing the first output
// character to str[strlen(str)]. This returns the number of characters the
// string would have *appended* assuming a large enough buffer, will make sure
// str is null-terminated, and will never write to str[size] or past.
// Example:
// int example(char *buf, size_t buf_size, double num, char *str) {
// int n = 0;
// n += mp_snprintf_cat(buf, size, "%f", num);
// n += mp_snprintf_cat(buf, size, "%s", str);
// return n; }
// Note how this can be chained with functions similar in style.
int mp_snprintf_cat(char *str, size_t size, const char *format, ...)
{
size_t len = strnlen(str, size);
assert(!size || len < size); // str with no 0-termination is not allowed
int r;
va_list ap;
va_start(ap, format);
r = vsnprintf(str + len, size - len, format, ap);
va_end(ap);
return r;
}
// Encode the unicode codepoint as UTF-8, and append to the end of the
// talloc'ed buffer. All guarantees bstr_xappend() give applies, such as
// implicit \0-termination for convenience.
void mp_append_utf8_bstr(void *talloc_ctx, struct bstr *buf, uint32_t codepoint)
{
char data[8];
uint8_t tmp;
char *output = data;
PUT_UTF8(codepoint, tmp, *output++ = tmp;);
bstr_xappend(talloc_ctx, buf, (bstr){data, output - data});
}
// Parse a C/JSON-style escape beginning at code, and append the result to *str
// using talloc. The input string (*code) must point to the first character
// after the initial '\', and after parsing *code is set to the first character
// after the current escape.
// On error, false is returned, and all input remains unchanged.
static bool mp_parse_escape(void *talloc_ctx, bstr *dst, bstr *code)
{
if (code->len < 1)
return false;
char replace = 0;
switch (code->start[0]) {
case '"': replace = '"'; break;
case '\\': replace = '\\'; break;
case '/': replace = '/'; break;
case 'b': replace = '\b'; break;
case 'f': replace = '\f'; break;
case 'n': replace = '\n'; break;
case 'r': replace = '\r'; break;
case 't': replace = '\t'; break;
case 'e': replace = '\x1b'; break;
case '\'': replace = '\''; break;
}
if (replace) {
bstr_xappend(talloc_ctx, dst, (bstr){&replace, 1});
*code = bstr_cut(*code, 1);
return true;
}
if (code->start[0] == 'x' && code->len >= 3) {
bstr num = bstr_splice(*code, 1, 3);
char c = bstrtoll(num, &num, 16);
if (num.len)
return false;
bstr_xappend(talloc_ctx, dst, (bstr){&c, 1});
*code = bstr_cut(*code, 3);
return true;
}
if (code->start[0] == 'u' && code->len >= 5) {
bstr num = bstr_splice(*code, 1, 5);
uint32_t c = bstrtoll(num, &num, 16);
if (num.len)
return false;
if (c >= 0xd800 && c <= 0xdbff) {
if (code->len < 5 + 6 // udddd + \udddd
|| code->start[5] != '\\' || code->start[6] != 'u')
return false;
*code = bstr_cut(*code, 5 + 1);
bstr num2 = bstr_splice(*code, 1, 5);
uint32_t c2 = bstrtoll(num2, &num2, 16);
if (num2.len || c2 < 0xdc00 || c2 > 0xdfff)
return false;
c = ((c - 0xd800) << 10) + 0x10000 + (c2 - 0xdc00);
}
mp_append_utf8_bstr(talloc_ctx, dst, c);
*code = bstr_cut(*code, 5);
return true;
}
return false;
}
// Like mp_append_escaped_string, but set *dst to sliced *src if no escape
// sequences have to be parsed (i.e. no memory allocation is required), and
// if dst->start was NULL on function entry.
bool mp_append_escaped_string_noalloc(void *talloc_ctx, bstr *dst, bstr *src)
{
bstr t = *src;
int cur = 0;
while (1) {
if (cur >= t.len || t.start[cur] == '"') {
*src = bstr_cut(t, cur);
t = bstr_splice(t, 0, cur);
if (dst->start == NULL) {
*dst = t;
} else {
bstr_xappend(talloc_ctx, dst, t);
}
return true;
} else if (t.start[cur] == '\\') {
bstr_xappend(talloc_ctx, dst, bstr_splice(t, 0, cur));
t = bstr_cut(t, cur + 1);
cur = 0;
if (!mp_parse_escape(talloc_ctx, dst, &t))
goto error;
} else {
cur++;
}
}
error:
return false;
}
// src is expected to point to a C-style string literal, *src pointing to the
// first char after the starting '"'. It will append the contents of the literal
// to *dst (using talloc_ctx) until the first '"' or the end of *str is found.
// See bstr_xappend() how data is appended to *dst.
// On success, *src will either start with '"', or be empty.
// On error, return false, and *dst will contain the string until the first
// error, *src is not changed.
// Note that dst->start will be implicitly \0-terminated on successful return,
// and if it was NULL or \0-terminated before calling the function.
// As mentioned above, the caller is responsible for skipping the '"' chars.
bool mp_append_escaped_string(void *talloc_ctx, bstr *dst, bstr *src)
{
if (mp_append_escaped_string_noalloc(talloc_ctx, dst, src)) {
// Guarantee copy (or allocation).
if (!dst->start || dst->start == src->start) {
bstr res = *dst;
*dst = (bstr){0};
bstr_xappend(talloc_ctx, dst, res);
}
return true;
}
return false;
}
// Behaves like strerror()/strerror_r(), but is thread- and GNU-safe.
char *mp_strerror_buf(char *buf, size_t buf_size, int errnum)
{
// This handles the nasty details of calling the right function for us.
av_strerror(AVERROR(errnum), buf, buf_size);
return buf;
}
char *mp_tag_str_buf(char *buf, size_t buf_size, uint32_t tag)
{
if (buf_size < 1)
return buf;
buf[0] = '\0';
for (int n = 0; n < 4; n++) {
uint8_t val = (tag >> (n * 8)) & 0xFF;
if (mp_isalnum(val) || val == '_' || val == ' ') {
mp_snprintf_cat(buf, buf_size, "%c", val);
} else {
mp_snprintf_cat(buf, buf_size, "[%d]", val);
}
}
return buf;
}
char *mp_tprintf_buf(char *buf, size_t buf_size, const char *format, ...)
{
va_list ap;
va_start(ap, format);
vsnprintf(buf, buf_size, format, ap);
va_end(ap);
return buf;
}
char **mp_dup_str_array(void *tctx, char **s)
{
char **r = NULL;
int num_r = 0;
for (int n = 0; s && s[n]; n++)
MP_TARRAY_APPEND(tctx, r, num_r, talloc_strdup(tctx, s[n]));
if (r)
MP_TARRAY_APPEND(tctx, r, num_r, NULL);
return r;
}