King_DuckZ/libbpg
1
0
Fork 0
forked from mirror/libbpg
Code Issues Pull requests Releases Wiki Activity
libbpg/libbpg.c
King_DuckZ 357f186837 libbpg-0.9.5
2015-01-16 13:48:11 +01:00

1957 lines
54 KiB
C
Raw Blame History

/*
* libbpg
*
* Copyright (c) 2014 Fabrice Bellard
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to deal
* in the Software without restriction, including without limitation the rights
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
* copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
* THE SOFTWARE.
*/
#include <math.h>
#ifdef EMSCRIPTEN
#include <emscripten.h>
#endif
#include <libavutil/opt.h>
#include <libavcodec/avcodec.h>
#include <libavutil/common.h>
/* The following global defines are used:
- USE_VAR_BIT_DEPTH : support of bit depth > 8 bits
- USE_PRED : support of animations
*/
#ifndef EMSCRIPTEN
#define USE_RGB48 /* support all pixel formats */
//#define DEBUG
#endif
#if !defined(DEBUG)
#define NDEBUG
#endif
#include <assert.h>
#include "libbpg.h"
#define BPG_HEADER_MAGIC 0x425047fb
#define ITAPS2 4
#define ITAPS (2 * ITAPS2) /* number of taps of the interpolation filter */
#ifdef USE_VAR_BIT_DEPTH
typedef uint16_t PIXEL;
#else
typedef uint8_t PIXEL;
#endif
#define MAX_DATA_SIZE ((1 << 30) - 1)
typedef struct {
int c_shift;
int c_rnd;
int c_one;
int y_one, y_offset;
int c_r_cr, c_g_cb, c_g_cr, c_b_cb;
int c_center;
int bit_depth;
int limited_range;
} ColorConvertState;
typedef void ColorConvertFunc(ColorConvertState *s,
uint8_t *dst, const PIXEL *y_ptr,
const PIXEL *cb_ptr, const PIXEL *cr_ptr,
int n, int incr);
struct BPGDecoderContext {
AVCodecContext *dec_ctx;
AVCodecContext *alpha_dec_ctx;
AVFrame *frame;
AVFrame *alpha_frame;
int w, h;
BPGImageFormatEnum format;
uint8_t c_h_phase; /* only used for 422 and 420 */
uint8_t has_alpha; /* true if alpha or W plane */
uint8_t bit_depth;
uint8_t has_w_plane;
uint8_t limited_range;
uint8_t premultiplied_alpha;
uint8_t has_animation;
BPGColorSpaceEnum color_space;
uint8_t keep_extension_data; /* true if the extension data must be
kept during parsing */
uint8_t decode_animation; /* true if animation decoding is enabled */
BPGExtensionData *first_md;
/* animation */
uint16_t loop_count;
uint16_t frame_delay_num;
uint16_t frame_delay_den;
uint8_t *input_buf;
int input_buf_pos;
int input_buf_len;
/* the following is used for format conversion */
uint8_t output_inited;
BPGDecoderOutputFormat out_fmt;
uint8_t is_rgba;
uint8_t is_16bpp;
uint8_t is_cmyk;
int y; /* current line */
int w2, h2;
const uint8_t *y_buf, *cb_buf, *cr_buf, *a_buf;
int y_linesize, cb_linesize, cr_linesize, a_linesize;
PIXEL *cb_buf2, *cr_buf2, *cb_buf3[ITAPS], *cr_buf3[ITAPS];
int16_t *c_buf4;
ColorConvertState cvt;
ColorConvertFunc *cvt_func;
};
/* ffmpeg utilities */
#ifdef USE_AV_LOG
void av_log(void* avcl, int level, const char *fmt, ...)
{
#ifdef DEBUG
va_list ap;
va_start(ap, fmt);
vfprintf(stderr, fmt, ap);
va_end(ap);
#endif
}
void avpriv_report_missing_feature(void *avc, const char *msg, ...)
{
#ifdef DEBUG
va_list ap;
va_start(ap, msg);
vfprintf(stderr, msg, ap);
va_end(ap);
#endif
}
#endif /* USE_AV_LOG */
/* return < 0 if error, otherwise the consumed length */
static int get_ue32(uint32_t *pv, const uint8_t *buf, int len)
{
const uint8_t *p;
uint32_t v;
int a;
if (len <= 0)
return -1;
p = buf;
a = *p++;
len--;
if (a < 0x80) {
*pv = a;
return 1;
} else if (a == 0x80) {
/* we don't accept non canonical encodings */
return -1;
}
v = a & 0x7f;
for(;;) {
if (len <= 0)
return -1;
a = *p++;
len--;
v = (v << 7) | (a & 0x7f);
if (!(a & 0x80))
break;
}
*pv = v;
return p - buf;
}
static int get_ue(uint32_t *pv, const uint8_t *buf, int len)
{
int ret;
ret = get_ue32(pv, buf, len);
if (ret < 0)
return ret;
/* limit the maximum size to avoid overflows in buffer
computations */
if (*pv > MAX_DATA_SIZE)
return -1;
return ret;
}
static int build_msps(uint8_t **pbuf, int *pbuf_len,
const uint8_t *input_data, int input_data_len1,
int width, int height, int chroma_format_idc,
int bit_depth)
{
int input_data_len = input_data_len1;
int idx, msps_len, ret, buf_len, i;
uint32_t len;
uint8_t *buf, *msps_buf;
*pbuf = NULL;
/* build the modified SPS header to please libavcodec */
ret = get_ue(&len, input_data, input_data_len);
if (ret < 0)
return -1;
input_data += ret;
input_data_len -= ret;
if (len > input_data_len)
return -1;
msps_len = 1 + 4 + 4 + 1 + len;
msps_buf = av_malloc(msps_len);
idx = 0;
msps_buf[idx++] = chroma_format_idc;
msps_buf[idx++] = (width >> 24);
msps_buf[idx++] = (width >> 16);
msps_buf[idx++] = (width >> 8);
msps_buf[idx++] = (width >> 0);
msps_buf[idx++] = (height >> 24);
msps_buf[idx++] = (height >> 16);
msps_buf[idx++] = (height >> 8);
msps_buf[idx++] = (height >> 0);
msps_buf[idx++] = bit_depth - 8;
memcpy(msps_buf + idx, input_data, len);
idx += len;
assert(idx == msps_len);
input_data += len;
input_data_len -= len;
buf_len = 4 + 2 + msps_len * 2 + 4 + (input_data_len - len);
buf = av_malloc(buf_len);
idx = 0;
/* NAL header */
buf[idx++] = 0x00;
buf[idx++] = 0x00;
buf[idx++] = 0x00;
buf[idx++] = 0x01;
buf[idx++] = (48 << 1); /* application specific NAL unit type */
buf[idx++] = 1;
/* add the modified SPS with the correct escape codes */
i = 0;
while (i < msps_len) {
if ((i + 1) < msps_len && msps_buf[i] == 0 && msps_buf[i + 1] == 0) {
buf[idx++] = 0x00;
buf[idx++] = 0x00;
buf[idx++] = 0x03;
i += 2;
} else {
buf[idx++] = msps_buf[i++];
}
}
/* the last byte cannot be 0 */
if (idx == 0 || buf[idx - 1] == 0x00)
buf[idx++] = 0x80;
av_free(msps_buf);
*pbuf_len = idx;
*pbuf = buf;
return input_data_len1 - input_data_len;
}
/* return the position of the end of the NAL or -1 if error */
static int find_nal_end(const uint8_t *buf, int buf_len, int has_startcode)
{
int idx;
idx = 0;
if (has_startcode) {
if (buf_len >= 4 &&
buf[0] == 0 && buf[1] == 0 && buf[2] == 0 && buf[3] == 1) {
idx = 4;
} else if (buf_len >= 3 &&
buf[0] == 0 && buf[1] == 0 && buf[2] == 1) {
idx = 3;
} else {
return -1;
}
}
/* NAL header */
if (idx + 2 > buf_len)
return -1;
/* find the last byte */
for(;;) {
if (idx + 2 >= buf_len) {
idx = buf_len;
break;
}
if (buf[idx] == 0 && buf[idx + 1] == 0 && buf[idx + 2] == 1)
break;
if (idx + 3 < buf_len &&
buf[idx] == 0 && buf[idx + 1] == 0 && buf[idx + 2] == 0 && buf[idx + 3] == 1)
break;
idx++;
}
return idx;
}
typedef struct {
uint8_t *buf;
int size;
int len;
} DynBuf;
static void dyn_buf_init(DynBuf *s)
{
s->buf = NULL;
s->size = 0;
s->len = 0;
}
static int dyn_buf_resize(DynBuf *s, int size)
{
int new_size;
uint8_t *new_buf;
if (size <= s->size)
return 0;
new_size = (s->size * 3) / 2;
if (new_size < size)
new_size = size;
new_buf = av_realloc(s->buf, new_size);
if (!new_buf)
return -1;
s->buf = new_buf;
s->size = new_size;
return 0;
}
static int dyn_buf_push(DynBuf *s, const uint8_t *data, int len)
{
if (dyn_buf_resize(s, s->len + len) < 0)
return -1;
memcpy(s->buf + s->len, data, len);
s->len += len;
return 0;
}
extern AVCodec ff_hevc_decoder;
static int hevc_decode_init1(DynBuf *pbuf, AVFrame **pframe,
AVCodecContext **pc,
const uint8_t *buf, int buf_len,
int width, int height, int chroma_format_idc,
int bit_depth)
{
AVCodec *codec;
AVCodecContext *c;
AVFrame *frame;
uint8_t *nal_buf;
int nal_len, ret, ret1;
ret = build_msps(&nal_buf, &nal_len, buf, buf_len,
width, height, chroma_format_idc, bit_depth);
if (ret < 0)
return -1;
ret1 = dyn_buf_push(pbuf, nal_buf, nal_len);
av_free(nal_buf);
if (ret1 < 0)
return -1;
codec = &ff_hevc_decoder;
c = avcodec_alloc_context3(codec);
if (!c)
return -1;
frame = av_frame_alloc();
if (!frame)
return -1;
/* for testing: use the MD5 or CRC in SEI to check the decoded bit
stream. */
c->err_recognition |= AV_EF_CRCCHECK;
/* open it */
if (avcodec_open2(c, codec, NULL) < 0) {
av_frame_free(&frame);
return -1;
}
*pc = c;
*pframe = frame;
return ret;
}
static int hevc_write_frame(AVCodecContext *avctx,
AVFrame *frame,
uint8_t *buf, int buf_len)
{
AVPacket avpkt;
int len, got_frame;
av_init_packet(&avpkt);
avpkt.data = (uint8_t *)buf;
avpkt.size = buf_len;
/* avoid using uninitialized data */
memset(buf + buf_len, 0, FF_INPUT_BUFFER_PADDING_SIZE);
len = avcodec_decode_video2(avctx, frame, &got_frame, &avpkt);
if (len < 0 || !got_frame)
return -1;
else
return 0;
}
static int hevc_decode_frame_internal(BPGDecoderContext *s,
DynBuf *abuf, DynBuf *cbuf,
const uint8_t *buf, int buf_len1,
int first_nal)
{
int nal_len, start, nal_buf_len, ret, nuh_layer_id, buf_len, has_alpha;
int nut, frame_start_found[2];
DynBuf *pbuf;
uint8_t *nal_buf;
has_alpha = (s->alpha_dec_ctx != NULL);
buf_len = buf_len1;
frame_start_found[0] = 0;
frame_start_found[1] = 0;
while (buf_len > 0) {
if (buf_len < (first_nal ? 3 : 0) + 2)
goto fail;
if (first_nal)
start = 0;
else
start = 3 + (buf[2] == 0);
if (buf_len < start + 3)
goto fail;
nuh_layer_id = ((buf[start] & 1) << 5) | (buf[start + 1] >> 3);
nut = (buf[start] >> 1) & 0x3f;
#if 0
printf("nal: type=%d layer_id=%d fs=%d %d\n",
nut, nuh_layer_id, frame_start_found[0], frame_start_found[1]);
#endif
/* Note: we assume the alpha and color data are correctly
interleaved */
if ((nut >= 32 && nut <= 35) || nut == 39 || nut >= 41) {
if (frame_start_found[0] && frame_start_found[has_alpha])
break;
} else if ((nut <= 9 || (nut >= 16 && nut <= 21)) &&
start + 2 < buf_len && (buf[start + 2] & 0x80)) {
/* first slice segment */
if (frame_start_found[0] && frame_start_found[has_alpha])
break;
if (has_alpha && nuh_layer_id == 1)
frame_start_found[1] = 1;
else
frame_start_found[0] = 1;
}
nal_len = find_nal_end(buf, buf_len, !first_nal);
if (nal_len < 0)
goto fail;
nal_buf_len = nal_len - start + 3;
if (has_alpha && nuh_layer_id == 1)
pbuf = abuf;
else
pbuf = cbuf;
if (dyn_buf_resize(pbuf, pbuf->len + nal_buf_len) < 0)
goto fail;
nal_buf = pbuf->buf + pbuf->len;
nal_buf[0] = 0x00;
nal_buf[1] = 0x00;
nal_buf[2] = 0x01;
memcpy(nal_buf + 3, buf + start, nal_len - start);
if (has_alpha && nuh_layer_id == 1)
nal_buf[4] &= 0x7;
pbuf->len += nal_buf_len;
buf += nal_len;
buf_len -= nal_len;
first_nal = 0;
}
if (s->alpha_dec_ctx) {
if (dyn_buf_resize(abuf, abuf->len + FF_INPUT_BUFFER_PADDING_SIZE) < 0)
goto fail;
ret = hevc_write_frame(s->alpha_dec_ctx, s->alpha_frame, abuf->buf, abuf->len);
if (ret < 0)
goto fail;
}
if (dyn_buf_resize(cbuf, cbuf->len + FF_INPUT_BUFFER_PADDING_SIZE) < 0)
goto fail;
ret = hevc_write_frame(s->dec_ctx, s->frame, cbuf->buf, cbuf->len);
if (ret < 0)
goto fail;
ret = buf_len1 - buf_len;
done:
return ret;
fail:
ret = -1;
goto done;
}
/* decode the first frame */
static int hevc_decode_start(BPGDecoderContext *s,
const uint8_t *buf, int buf_len1,
int width, int height, int chroma_format_idc,
int bit_depth, int has_alpha)
{
int ret, buf_len;
DynBuf abuf_s, *abuf = &abuf_s;
DynBuf cbuf_s, *cbuf = &cbuf_s;
dyn_buf_init(abuf);
dyn_buf_init(cbuf);
buf_len = buf_len1;
if (has_alpha) {
ret = hevc_decode_init1(abuf, &s->alpha_frame, &s->alpha_dec_ctx,
buf, buf_len, width, height, 0, bit_depth);
if (ret < 0)
goto fail;
buf += ret;
buf_len -= ret;
}
ret = hevc_decode_init1(cbuf, &s->frame, &s->dec_ctx,
buf, buf_len, width, height, chroma_format_idc,
bit_depth);
if (ret < 0)
goto fail;
buf += ret;
buf_len -= ret;
ret = hevc_decode_frame_internal(s, abuf, cbuf, buf, buf_len, 1);
av_free(abuf->buf);
av_free(cbuf->buf);
if (ret < 0)
goto fail;
buf_len -= ret;
return buf_len1 - buf_len;
fail:
return -1;
}
#ifdef USE_PRED
static int hevc_decode_frame(BPGDecoderContext *s,
const uint8_t *buf, int buf_len)
{
int ret;
DynBuf abuf_s, *abuf = &abuf_s;
DynBuf cbuf_s, *cbuf = &cbuf_s;
dyn_buf_init(abuf);
dyn_buf_init(cbuf);
ret = hevc_decode_frame_internal(s, abuf, cbuf, buf, buf_len, 0);
av_free(abuf->buf);
av_free(cbuf->buf);
return ret;
}
#endif
static void hevc_decode_end(BPGDecoderContext *s)
{
if (s->alpha_dec_ctx) {
avcodec_close(s->alpha_dec_ctx);
av_free(s->alpha_dec_ctx);
s->alpha_dec_ctx = NULL;
}
if (s->dec_ctx) {
avcodec_close(s->dec_ctx);
av_free(s->dec_ctx);
s->dec_ctx = NULL;
}
}
uint8_t *bpg_decoder_get_data(BPGDecoderContext *img, int *pline_size, int plane)
{
int c_count;
if (img->format == BPG_FORMAT_GRAY)
c_count = 1;
else
c_count = 3;
if (plane < c_count) {
*pline_size = img->frame->linesize[plane];
return img->frame->data[plane];
} else if (img->has_alpha && plane == c_count) {
*pline_size = img->alpha_frame->linesize[0];
return img->alpha_frame->data[0];
} else {
*pline_size = 0;
return NULL;
}
}
int bpg_decoder_get_info(BPGDecoderContext *img, BPGImageInfo *p)
{
if (!img->frame)
return -1;
p->width = img->w;
p->height = img->h;
p->format = img->format;
p->has_alpha = img->has_alpha && !img->has_w_plane;
p->premultiplied_alpha = img->premultiplied_alpha;
p->has_w_plane = img->has_w_plane;
p->limited_range = img->limited_range;
p->color_space = img->color_space;
p->bit_depth = img->bit_depth;
p->has_animation = img->has_animation;
p->loop_count = img->loop_count;
return 0;
}
static inline int clamp_pix(int a, int pixel_max)
{
if (a < 0)
return 0;
else if (a > pixel_max)
return pixel_max;
else
return a;
}
static inline int clamp8(int a)
{
if (a < 0)
return 0;
else if (a > 255)
return 255;
else
return a;
}
/* 8 tap Lanczos interpolator (phase=0, symmetric) */
#define IP0C0 40
#define IP0C1 (-11)
#define IP0C2 4
#define IP0C3 (-1)
/* 7 tap Lanczos interpolator (phase=0.5) */
#define IP1C0 (-1)
#define IP1C1 4
#define IP1C2 (-10)
#define IP1C3 57
#define IP1C4 18
#define IP1C5 (-6)
#define IP1C6 2
/* interpolate by a factor of two assuming chroma is aligned with the
luma samples. */
static void interp2p0_simple(PIXEL *dst, const PIXEL *src, int n, int bit_depth)
{
int pixel_max;
pixel_max = (1 << bit_depth) - 1;
while (n >= 2) {
dst[0] = src[0];
dst[1] = clamp_pix(((src[-3] + src[4]) * IP0C3 +
(src[-2] + src[3]) * IP0C2 +
(src[-1] + src[2]) * IP0C1 +
(src[0] + src[1]) * IP0C0 + 32) >> 6, pixel_max);
dst += 2;
src++;
n -= 2;
}
if (n) {
dst[0] = src[0];
}
}
static void interp2p0_simple16(PIXEL *dst, const int16_t *src, int n, int bit_depth)
{
int shift1, offset1, shift0, offset0, pixel_max;
pixel_max = (1 << bit_depth) - 1;
shift0 = 14 - bit_depth;
offset0 = (1 << shift0) >> 1;
shift1 = 20 - bit_depth;
offset1 = 1 << (shift1 - 1);
while (n >= 2) {
dst[0] = clamp_pix((src[0] + offset0) >> shift0, pixel_max);
dst[1] = clamp_pix(((src[-3] + src[4]) * IP0C3 +
(src[-2] + src[3]) * IP0C2 +
(src[-1] + src[2]) * IP0C1 +
(src[0] + src[1]) * IP0C0 + offset1) >> shift1,
pixel_max);
dst += 2;
src++;
n -= 2;
}
if (n) {
dst[0] = clamp_pix((src[0] + offset0) >> shift0, pixel_max);
}
}
/* interpolate by a factor of two assuming chroma is between the luma
samples. */
static void interp2p1_simple(PIXEL *dst, const PIXEL *src, int n, int bit_depth)
{
int pixel_max, a0, a1, a2, a3, a4, a5, a6;
pixel_max = (1 << bit_depth) - 1;
a1 = src[-3];
a2 = src[-2];
a3 = src[-1];
a4 = src[0];
a5 = src[1];
a6 = src[2];
while (n >= 2) {
a0 = a1;
a1 = a2;
a2 = a3;
a3 = a4;
a4 = a5;
a5 = a6;
a6 = src[3];
dst[0] = clamp_pix((a0 * IP1C6 + a1 * IP1C5 + a2 * IP1C4 + a3 * IP1C3 +
a4 * IP1C2 + a5 * IP1C1 + a6 * IP1C0 + 32) >> 6,
pixel_max);
dst[1] = clamp_pix((a0 * IP1C0 + a1 * IP1C1 + a2 * IP1C2 + a3 * IP1C3 +
a4 * IP1C4 + a5 * IP1C5 + a6 * IP1C6 + 32) >> 6,
pixel_max);
dst += 2;
src++;
n -= 2;
}
if (n) {
a0 = a1;
a1 = a2;
a2 = a3;
a3 = a4;
a4 = a5;
a5 = a6;
a6 = src[3];
dst[0] = clamp_pix((a0 * IP1C6 + a1 * IP1C5 + a2 * IP1C4 + a3 * IP1C3 +
a4 * IP1C2 + a5 * IP1C1 + a6 * IP1C0 + 32) >> 6,
pixel_max);
}
}
static void interp2p1_simple16(PIXEL *dst, const int16_t *src, int n,
int bit_depth)
{
int shift, offset, pixel_max, a0, a1, a2, a3, a4, a5, a6;
pixel_max = (1 << bit_depth) - 1;
shift = 20 - bit_depth;
offset = 1 << (shift - 1);
a1 = src[-3];
a2 = src[-2];
a3 = src[-1];
a4 = src[0];
a5 = src[1];
a6 = src[2];
while (n >= 2) {
a0 = a1;
a1 = a2;
a2 = a3;
a3 = a4;
a4 = a5;
a5 = a6;
a6 = src[3];
dst[0] = clamp_pix((a0 * IP1C6 + a1 * IP1C5 + a2 * IP1C4 + a3 * IP1C3 +
a4 * IP1C2 + a5 * IP1C1 + a6 * IP1C0 + offset) >> shift,
pixel_max);
dst[1] = clamp_pix((a0 * IP1C0 + a1 * IP1C1 + a2 * IP1C2 + a3 * IP1C3 +
a4 * IP1C4 + a5 * IP1C5 + a6 * IP1C6 + offset) >> shift,
pixel_max);
dst += 2;
src++;
n -= 2;
}
if (n) {
a0 = a1;
a1 = a2;
a2 = a3;
a3 = a4;
a4 = a5;
a5 = a6;
a6 = src[3];
dst[0] = clamp_pix((a0 * IP1C6 + a1 * IP1C5 + a2 * IP1C4 + a3 * IP1C3 +
a4 * IP1C2 + a5 * IP1C1 + a6 * IP1C0 + offset) >> shift,
pixel_max);
}
}
/* tmp_buf is a temporary buffer of length (n2 + 2 * ITAPS2 - 1) */
static void interp2_h(PIXEL *dst, const PIXEL *src, int n, int bit_depth,
int phase, PIXEL *tmp_buf)
{
PIXEL *src1 = tmp_buf, v;
int i, n2;
/* add extra pixels and do the interpolation (XXX: could go faster) */
n2 = (n + 1) / 2;
memcpy(src1 + ITAPS2 - 1, src, n2 * sizeof(PIXEL));
v = src[0];
for(i = 0; i < ITAPS2 - 1; i++)
src1[i] = v;
v = src[n2 - 1];
for(i = 0; i < ITAPS2; i++)
src1[ITAPS2 - 1 + n2 + i] = v;
if (phase == 0)
interp2p0_simple(dst, src1 + ITAPS2 - 1, n, bit_depth);
else
interp2p1_simple(dst, src1 + ITAPS2 - 1, n, bit_depth);
}
/* y_pos is the position of the sample '0' in the 'src' circular
buffer. tmp_buf is a temporary buffer of length (n2 + 2 * ITAPS2 - 1) */
static void interp2_vh(PIXEL *dst, PIXEL **src, int n, int y_pos,
int16_t *tmp_buf, int bit_depth, int frac_pos,
int c_h_phase)
{
const PIXEL *src0, *src1, *src2, *src3, *src4, *src5, *src6;
int i, n2, shift, rnd;
int16_t v;
src0 = src[(y_pos - 3) & 7];
src1 = src[(y_pos - 2) & 7];
src2 = src[(y_pos - 1) & 7];
src3 = src[(y_pos + 0) & 7];
src4 = src[(y_pos + 1) & 7];
src5 = src[(y_pos + 2) & 7];
src6 = src[(y_pos + 3) & 7];
/* vertical interpolation first */
shift = bit_depth - 8;
rnd = (1 << shift) >> 1;
n2 = (n + 1) / 2;
if (frac_pos == 0) {
for(i = 0; i < n2; i++) {
tmp_buf[ITAPS2 - 1 + i] =
(src0[i] * IP1C6 + src1[i] * IP1C5 +
src2[i] * IP1C4 + src3[i] * IP1C3 +
src4[i] * IP1C2 + src5[i] * IP1C1 +
src6[i] * IP1C0 + rnd) >> shift;
}
} else {
for(i = 0; i < n2; i++) {
tmp_buf[ITAPS2 - 1 + i] =
(src0[i] * IP1C0 + src1[i] * IP1C1 +
src2[i] * IP1C2 + src3[i] * IP1C3 +
src4[i] * IP1C4 + src5[i] * IP1C5 +
src6[i] * IP1C6 + rnd) >> shift;
}
}
/* then horizontal interpolation */
v = tmp_buf[ITAPS2 - 1];
for(i = 0; i < ITAPS2 - 1; i++)
tmp_buf[i] = v;
v = tmp_buf[ITAPS2 - 1 + n2 - 1];
for(i = 0; i < ITAPS2; i++)
tmp_buf[ITAPS2 - 1 + n2 + i] = v;
if (c_h_phase == 0)
interp2p0_simple16(dst, tmp_buf + ITAPS2 - 1, n, bit_depth);
else
interp2p1_simple16(dst, tmp_buf + ITAPS2 - 1, n, bit_depth);
}
static void ycc_to_rgb24(ColorConvertState *s, uint8_t *dst, const PIXEL *y_ptr,
const PIXEL *cb_ptr, const PIXEL *cr_ptr,
int n, int incr)
{
uint8_t *q = dst;
int y_val, cb_val, cr_val, x;
int c_r_cr, c_g_cb, c_g_cr, c_b_cb, rnd, shift, center, c_one;
c_r_cr = s->c_r_cr;
c_g_cb = s->c_g_cb;
c_g_cr = s->c_g_cr;
c_b_cb = s->c_b_cb;
c_one = s->y_one;
rnd = s->y_offset;
shift = s->c_shift;
center = s->c_center;
for(x = 0; x < n; x++) {
y_val = y_ptr[x] * c_one;
cb_val = cb_ptr[x] - center;
cr_val = cr_ptr[x] - center;
q[0] = clamp8((y_val + c_r_cr * cr_val + rnd) >> shift);
q[1] = clamp8((y_val - c_g_cb * cb_val - c_g_cr * cr_val + rnd) >> shift);
q[2] = clamp8((y_val + c_b_cb * cb_val + rnd) >> shift);
q += incr;
}
}
static void ycgco_to_rgb24(ColorConvertState *s,
uint8_t *dst, const PIXEL *y_ptr,
const PIXEL *cb_ptr, const PIXEL *cr_ptr,
int n, int incr)
{
uint8_t *q = dst;
int y_val, cb_val, cr_val, x;
int rnd, shift, center, c_one;
c_one = s->y_one;
rnd = s->y_offset;
shift = s->c_shift;
center = s->c_center;
for(x = 0; x < n; x++) {
y_val = y_ptr[x];
cb_val = cb_ptr[x] - center;
cr_val = cr_ptr[x] - center;
q[0] = clamp8(((y_val - cb_val + cr_val) * c_one + rnd) >> shift);
q[1] = clamp8(((y_val + cb_val) * c_one + rnd) >> shift);
q[2] = clamp8(((y_val - cb_val - cr_val) * c_one + rnd) >> shift);
q += incr;
}
}
/* c = c * alpha */
static void alpha_combine8(ColorConvertState *s,
uint8_t *dst, const PIXEL *a_ptr, int n, int incr)
{
uint8_t *q = dst;
int x, a_val, shift, rnd;
shift = s->bit_depth;
rnd = 1 << (shift - 1);
for(x = 0; x < n; x++) {
a_val = a_ptr[x];
/* XXX: not accurate enough */
q[0] = (q[0] * a_val + rnd) >> shift;
q[1] = (q[1] * a_val + rnd) >> shift;
q[2] = (q[2] * a_val + rnd) >> shift;
q += incr;
}
}
static uint32_t divide8_table[256];
#define DIV8_BITS 16
static void alpha_divide8_init(void)
{
int i;
for(i = 1; i < 256; i++) {
/* Note: the 128 is added to have 100% correct results for all
the values */
divide8_table[i] = ((255 << DIV8_BITS) + (i / 2) + 128) / i;
}
}
static inline unsigned int comp_divide8(unsigned int val, unsigned int alpha,
unsigned int alpha_inv)
{
if (val >= alpha)
return 255;
return (val * alpha_inv + (1 << (DIV8_BITS - 1))) >> DIV8_BITS;
}
/* c = c / alpha */
static void alpha_divide8(uint8_t *dst, int n)
{
static int inited;
uint8_t *q = dst;
int x;
unsigned int a_val, a_inv;
if (!inited) {
inited = 1;
alpha_divide8_init();
}
for(x = 0; x < n; x++) {
a_val = q[3];
if (a_val == 0) {
q[0] = 255;
q[1] = 255;
q[2] = 255;
} else {
a_inv = divide8_table[a_val];
q[0] = comp_divide8(q[0], a_val, a_inv);
q[1] = comp_divide8(q[1], a_val, a_inv);
q[2] = comp_divide8(q[2], a_val, a_inv);
}
q += 4;
}
}
static void gray_to_rgb24(ColorConvertState *s,
uint8_t *dst, const PIXEL *y_ptr,
const PIXEL *cb_ptr, const PIXEL *cr_ptr,
int n, int incr)
{
uint8_t *q = dst;
int x, y_val, c, rnd, shift;
if (s->bit_depth == 8 && !s->limited_range) {
for(x = 0; x < n; x++) {
y_val = y_ptr[x];
q[0] = y_val;
q[1] = y_val;
q[2] = y_val;
q += incr;
}
} else {
c = s->y_one;
rnd = s->y_offset;
shift = s->c_shift;
for(x = 0; x < n; x++) {
y_val = clamp8((y_ptr[x] * c + rnd) >> shift);
q[0] = y_val;
q[1] = y_val;
q[2] = y_val;
q += incr;
}
}
}
static void rgb_to_rgb24(ColorConvertState *s, uint8_t *dst, const PIXEL *y_ptr,
const PIXEL *cb_ptr, const PIXEL *cr_ptr,
int n, int incr)
{
uint8_t *q = dst;
int x, c, rnd, shift;
if (s->bit_depth == 8 && !s->limited_range) {
for(x = 0; x < n; x++) {
q[0] = cr_ptr[x];
q[1] = y_ptr[x];
q[2] = cb_ptr[x];
q += incr;
}
} else {
c = s->y_one;
rnd = s->y_offset;
shift = s->c_shift;
for(x = 0; x < n; x++) {
q[0] = clamp8((cr_ptr[x] * c + rnd) >> shift);
q[1] = clamp8((y_ptr[x] * c + rnd) >> shift);
q[2] = clamp8((cb_ptr[x] * c + rnd) >> shift);
q += incr;
}
}
}
static void put_dummy_gray8(uint8_t *dst, int n, int incr)
{
int x;
for(x = 0; x < n; x++) {
dst[0] = 0xff;
dst += incr;
}
}
static void gray_to_gray8(ColorConvertState *s,
uint8_t *dst, const PIXEL *y_ptr,
int n, int incr)
{
uint8_t *q = dst;
int x, y_val, c, rnd, shift;
if (s->bit_depth == 8) {
for(x = 0; x < n; x++) {
y_val = y_ptr[x];
q[0] = y_val;
q += incr;
}
} else {
c = s->c_one;
rnd = s->c_rnd;
shift = s->c_shift;
for(x = 0; x < n; x++) {
y_val = (y_ptr[x] * c + rnd) >> shift;
q[0] = y_val;
q += incr;
}
}
}
static ColorConvertFunc *cs_to_rgb24[BPG_CS_COUNT] = {
ycc_to_rgb24,
rgb_to_rgb24,
ycgco_to_rgb24,
ycc_to_rgb24,
ycc_to_rgb24,
};
#ifdef USE_RGB48
/* 16 bit output */
static inline int clamp16(int a)
{
if (a < 0)
return 0;
else if (a > 65535)
return 65535;
else
return a;
}
static void ycc_to_rgb48(ColorConvertState *s, uint8_t *dst, const PIXEL *y_ptr,
const PIXEL *cb_ptr, const PIXEL *cr_ptr,
int n, int incr)
{
uint16_t *q = (uint16_t *)dst;
int y_val, cb_val, cr_val, x;
int c_r_cr, c_g_cb, c_g_cr, c_b_cb, rnd, shift, center, c_one;
c_r_cr = s->c_r_cr;
c_g_cb = s->c_g_cb;
c_g_cr = s->c_g_cr;
c_b_cb = s->c_b_cb;
c_one = s->y_one;
rnd = s->y_offset;
shift = s->c_shift;
center = s->c_center;
for(x = 0; x < n; x++) {
y_val = y_ptr[x] * c_one;
cb_val = cb_ptr[x] - center;
cr_val = cr_ptr[x] - center;
q[0] = clamp16((y_val + c_r_cr * cr_val + rnd) >> shift);
q[1] = clamp16((y_val - c_g_cb * cb_val - c_g_cr * cr_val + rnd) >> shift);
q[2] = clamp16((y_val + c_b_cb * cb_val + rnd) >> shift);
q += incr;
}
}
static void ycgco_to_rgb48(ColorConvertState *s,
uint8_t *dst, const PIXEL *y_ptr,
const PIXEL *cb_ptr, const PIXEL *cr_ptr,
int n, int incr)
{
uint16_t *q = (uint16_t *)dst;
int y_val, cb_val, cr_val, x;
int rnd, shift, center, c_one;
c_one = s->y_one;
rnd = s->y_offset;
shift = s->c_shift;
center = s->c_center;
for(x = 0; x < n; x++) {
y_val = y_ptr[x];
cb_val = cb_ptr[x] - center;
cr_val = cr_ptr[x] - center;
q[0] = clamp16(((y_val - cb_val + cr_val) * c_one + rnd) >> shift);
q[1] = clamp16(((y_val + cb_val) * c_one + rnd) >> shift);
q[2] = clamp16(((y_val - cb_val - cr_val) * c_one + rnd) >> shift);
q += incr;
}
}
static void gray_to_rgb48(ColorConvertState *s,
uint8_t *dst, const PIXEL *y_ptr,
const PIXEL *cb_ptr, const PIXEL *cr_ptr,
int n, int incr)
{
uint16_t *q = (uint16_t *)dst;
int x, y_val, c, rnd, shift;
c = s->y_one;
rnd = s->y_offset;
shift = s->c_shift;
for(x = 0; x < n; x++) {
y_val = clamp16((y_ptr[x] * c + rnd) >> shift);
q[0] = y_val;
q[1] = y_val;
q[2] = y_val;
q += incr;
}
}
static void gray_to_gray16(ColorConvertState *s,
uint16_t *dst, const PIXEL *y_ptr,
int n, int incr)
{
uint16_t *q = dst;
int x, y_val, c, rnd, shift;
c = s->c_one;
rnd = s->c_rnd;
shift = s->c_shift;
for(x = 0; x < n; x++) {
y_val = (y_ptr[x] * c + rnd) >> shift;
q[0] = y_val;
q += incr;
}
}
static void luma_to_gray16(ColorConvertState *s,
uint16_t *dst, const PIXEL *y_ptr,
int n, int incr)
{
uint16_t *q = dst;
int x, y_val, c, rnd, shift;
c = s->y_one;
rnd = s->y_offset;
shift = s->c_shift;
for(x = 0; x < n; x++) {
y_val = clamp16((y_ptr[x] * c + rnd) >> shift);
q[0] = y_val;
q += incr;
}
}
static void rgb_to_rgb48(ColorConvertState *s,
uint8_t *dst, const PIXEL *y_ptr,
const PIXEL *cb_ptr, const PIXEL *cr_ptr,
int n, int incr)
{
luma_to_gray16(s, (uint16_t *)dst + 1, y_ptr, n, incr);
luma_to_gray16(s, (uint16_t *)dst + 2, cb_ptr, n, incr);
luma_to_gray16(s, (uint16_t *)dst + 0, cr_ptr, n, incr);
}
static void put_dummy_gray16(uint16_t *dst, int n, int incr)
{
int x;
for(x = 0; x < n; x++) {
dst[0] = 0xffff;
dst += incr;
}
}
/* c = c * alpha */
static void alpha_combine16(ColorConvertState *s,
uint16_t *dst, const PIXEL *a_ptr, int n, int incr)
{
uint16_t *q = dst;
int x, a_val, shift, rnd;
shift = s->bit_depth;
rnd = 1 << (shift - 1);
for(x = 0; x < n; x++) {
a_val = a_ptr[x];
/* XXX: not accurate enough */
q[0] = (q[0] * a_val + rnd) >> shift;
q[1] = (q[1] * a_val + rnd) >> shift;
q[2] = (q[2] * a_val + rnd) >> shift;
q += incr;
}
}
#define DIV16_BITS 15
static unsigned int comp_divide16(unsigned int val, unsigned int alpha,
unsigned int alpha_inv)
{
if (val >= alpha)
return 65535;
return (val * alpha_inv + (1 << (DIV16_BITS - 1))) >> DIV16_BITS;
}
/* c = c / alpha */
static void alpha_divide16(uint16_t *dst, int n)
{
uint16_t *q = dst;
int x;
unsigned int a_val, a_inv;
for(x = 0; x < n; x++) {
a_val = q[3];
if (a_val == 0) {
q[0] = 65535;
q[1] = 65535;
q[2] = 65535;
} else {
a_inv = ((65535 << DIV16_BITS) + (a_val / 2)) / a_val;
q[0] = comp_divide16(q[0], a_val, a_inv);
q[1] = comp_divide16(q[1], a_val, a_inv);
q[2] = comp_divide16(q[2], a_val, a_inv);
}
q += 4;
}
}
static void gray_one_minus8(uint8_t *dst, int n, int incr)
{
int x;
for(x = 0; x < n; x++) {
dst[0] = 255 - dst[0];
dst += incr;
}
}
static void gray_one_minus16(uint16_t *dst, int n, int incr)
{
int x;
for(x = 0; x < n; x++) {
dst[0] = 65535 - dst[0];
dst += incr;
}
}
static ColorConvertFunc *cs_to_rgb48[BPG_CS_COUNT] = {
ycc_to_rgb48,
rgb_to_rgb48,
ycgco_to_rgb48,
ycc_to_rgb48,
ycc_to_rgb48,
};
#endif
static void convert_init(ColorConvertState *s,
int in_bit_depth, int out_bit_depth,
BPGColorSpaceEnum color_space,
int limited_range)
{
int c_shift, in_pixel_max, out_pixel_max;
double mult, k_r, k_b, mult_y, mult_c;
c_shift = 30 - out_bit_depth;
in_pixel_max = (1 << in_bit_depth) - 1;
out_pixel_max = (1 << out_bit_depth) - 1;
mult = (double)out_pixel_max * (1 << c_shift) / (double)in_pixel_max;
if (limited_range) {
mult_y = (double)out_pixel_max * (1 << c_shift) /
(double)(219 << (in_bit_depth - 8));
mult_c = (double)out_pixel_max * (1 << c_shift) /
(double)(224 << (in_bit_depth - 8));
} else {
mult_y = mult;
mult_c = mult;
}
switch(color_space) {
case BPG_CS_YCbCr:
k_r = 0.299;
k_b = 0.114;
goto convert_ycc;
case BPG_CS_YCbCr_BT709:
k_r = 0.2126;
k_b = 0.0722;
goto convert_ycc;
case BPG_CS_YCbCr_BT2020:
k_r = 0.2627;
k_b = 0.0593;
convert_ycc:
s->c_r_cr = lrint(2*(1-k_r) * mult_c);
s->c_g_cb = lrint(2*k_b*(1-k_b)/(1-k_b-k_r) * mult_c);
s->c_g_cr = lrint(2*k_r*(1-k_r)/(1-k_b-k_r) * mult_c);
s->c_b_cb = lrint(2*(1-k_b) * mult_c);
break;
default:
break;
}
s->c_one = lrint(mult);
s->c_shift = c_shift;
s->c_rnd = (1 << (c_shift - 1));
s->c_center = 1 << (in_bit_depth - 1);
if (limited_range) {
s->y_one = lrint(mult_y);
s->y_offset = -(16 << (in_bit_depth - 8)) * s->y_one + s->c_rnd;
} else {
s->y_one = s->c_one;
s->y_offset = s->c_rnd;
}
s->bit_depth = in_bit_depth;
s->limited_range = limited_range;
}
static int bpg_decoder_output_init(BPGDecoderContext *s,
BPGDecoderOutputFormat out_fmt)
{
int i;
#ifdef USE_RGB48
if ((unsigned)out_fmt > BPG_OUTPUT_FORMAT_CMYK64)
return -1;
#else
if ((unsigned)out_fmt > BPG_OUTPUT_FORMAT_RGBA32)
return -1;
#endif
s->is_rgba = (out_fmt == BPG_OUTPUT_FORMAT_RGBA32 ||
out_fmt == BPG_OUTPUT_FORMAT_RGBA64);
s->is_16bpp = (out_fmt == BPG_OUTPUT_FORMAT_RGB48 ||
out_fmt == BPG_OUTPUT_FORMAT_RGBA64 ||
out_fmt == BPG_OUTPUT_FORMAT_CMYK64);
s->is_cmyk = (out_fmt == BPG_OUTPUT_FORMAT_CMYK32 ||
out_fmt == BPG_OUTPUT_FORMAT_CMYK64);
if (s->format == BPG_FORMAT_420 || s->format == BPG_FORMAT_422) {
s->w2 = (s->w + 1) / 2;
s->h2 = (s->h + 1) / 2;
s->cb_buf2 = av_malloc(s->w * sizeof(PIXEL));
s->cr_buf2 = av_malloc(s->w * sizeof(PIXEL));
/* Note: too large if 422 and sizeof(PIXEL) = 1 */
s->c_buf4 = av_malloc((s->w2 + 2 * ITAPS2 - 1) * sizeof(int16_t));
if (s->format == BPG_FORMAT_420) {
for(i = 0; i < ITAPS; i++) {
s->cb_buf3[i] = av_malloc(s->w2 * sizeof(PIXEL));
s->cr_buf3[i] = av_malloc(s->w2 * sizeof(PIXEL));
}
}
}
convert_init(&s->cvt, s->bit_depth, s->is_16bpp ? 16 : 8,
s->color_space, s->limited_range);
if (s->format == BPG_FORMAT_GRAY) {
#ifdef USE_RGB48
if (s->is_16bpp) {
s->cvt_func = gray_to_rgb48;
} else
#endif
{
s->cvt_func = gray_to_rgb24;
}
} else {
#ifdef USE_RGB48
if (s->is_16bpp) {
s->cvt_func = cs_to_rgb48[s->color_space];
} else
#endif
{
s->cvt_func = cs_to_rgb24[s->color_space];
}
}
return 0;
}
static void bpg_decoder_output_end(BPGDecoderContext *s)
{
int i;
av_free(s->cb_buf2);
av_free(s->cr_buf2);
for(i = 0; i < ITAPS; i++) {
av_free(s->cb_buf3[i]);
av_free(s->cr_buf3[i]);
}
av_free(s->c_buf4);
}
int bpg_decoder_start(BPGDecoderContext *s, BPGDecoderOutputFormat out_fmt)
{
int ret, c_idx;
if (!s->frame)
return -1;
if (!s->output_inited) {
/* first frame is already decoded */
ret = bpg_decoder_output_init(s, out_fmt);
if (ret)
return ret;
s->output_inited = 1;
s->out_fmt = out_fmt;
} else {
#ifdef USE_PRED
if (s->has_animation && s->decode_animation) {
if (out_fmt != s->out_fmt)
return -1;
if (s->input_buf_pos >= s->input_buf_len) {
return -1;
} else {
ret = hevc_decode_frame(s, s->input_buf + s->input_buf_pos,
s->input_buf_len - s->input_buf_pos);
if (ret < 0)
return -1;
s->input_buf_pos += ret;
}
} else
#endif
{
return -1;
}
}
s->y_buf = bpg_decoder_get_data(s, &s->y_linesize, 0);
if (s->format != BPG_FORMAT_GRAY) {
s->cb_buf = bpg_decoder_get_data(s, &s->cb_linesize, 1);
s->cr_buf = bpg_decoder_get_data(s, &s->cr_linesize, 2);
c_idx = 3;
} else {
c_idx = 1;
}
if (s->has_alpha)
s->a_buf = bpg_decoder_get_data(s, &s->a_linesize, c_idx);
else
s->a_buf = NULL;
s->y = 0;
return 0;
}
void bpg_decoder_get_frame_duration(BPGDecoderContext *s, int *pnum, int *pden)
{
#ifdef USE_PRED
if (s->frame && s->has_animation) {
*pnum = s->frame_delay_num * (s->frame->pts);
*pden = s->frame_delay_den;
} else
#endif
{
*pnum = 0;
*pden = 1;
}
}
int bpg_decoder_get_line(BPGDecoderContext *s, void *rgb_line1)
{
uint8_t *rgb_line = rgb_line1;
int w, y, pos, y2, y1, incr, y_frac;
PIXEL *y_ptr, *cb_ptr, *cr_ptr, *a_ptr;
y = s->y;
if ((unsigned)y >= s->h)
return -1;
w = s->w;
y_ptr = (PIXEL *)(s->y_buf + y * s->y_linesize);
incr = 3 + (s->is_rgba || s->is_cmyk);
switch(s->format) {
case BPG_FORMAT_GRAY:
s->cvt_func(&s->cvt, rgb_line, y_ptr, NULL, NULL, w, incr);
break;
case BPG_FORMAT_420:
if (y == 0) {
int i;
/* init the vertical interpolation buffer */
for(i = 0; i < ITAPS; i++) {
y1 = i;
if (y1 > ITAPS2)
y1 -= ITAPS;
if (y1 < 0)
y1 = 0;
else if (y1 >= s->h2)
y1 = s->h2 - 1;
cb_ptr = (PIXEL *)(s->cb_buf + y1 * s->cb_linesize);
cr_ptr = (PIXEL *)(s->cr_buf + y1 * s->cr_linesize);
memcpy(s->cb_buf3[i], cb_ptr, s->w2 * sizeof(PIXEL));
memcpy(s->cr_buf3[i], cr_ptr, s->w2 * sizeof(PIXEL));
}
}
y2 = y >> 1;
pos = y2 % ITAPS;
y_frac = y & 1;
interp2_vh(s->cb_buf2, s->cb_buf3, w, pos, s->c_buf4,
s->bit_depth, y_frac, s->c_h_phase);
interp2_vh(s->cr_buf2, s->cr_buf3, w, pos, s->c_buf4,
s->bit_depth, y_frac, s->c_h_phase);
if (y_frac) {
/* add a new line in the circular buffer */
pos = (pos + ITAPS2 + 1) % ITAPS;
y1 = y2 + ITAPS2 + 1;
if (y1 >= s->h2)
y1 = s->h2 - 1;
cb_ptr = (PIXEL *)(s->cb_buf + y1 * s->cb_linesize);
cr_ptr = (PIXEL *)(s->cr_buf + y1 * s->cr_linesize);
memcpy(s->cb_buf3[pos], cb_ptr, s->w2 * sizeof(PIXEL));
memcpy(s->cr_buf3[pos], cr_ptr, s->w2 * sizeof(PIXEL));
}
s->cvt_func(&s->cvt, rgb_line, y_ptr, s->cb_buf2, s->cr_buf2, w, incr);
break;
case BPG_FORMAT_422:
cb_ptr = (PIXEL *)(s->cb_buf + y * s->cb_linesize);
cr_ptr = (PIXEL *)(s->cr_buf + y * s->cr_linesize);
interp2_h(s->cb_buf2, cb_ptr, w, s->bit_depth, s->c_h_phase,
(PIXEL *)s->c_buf4);
interp2_h(s->cr_buf2, cr_ptr, w, s->bit_depth, s->c_h_phase,
(PIXEL *)s->c_buf4);
s->cvt_func(&s->cvt, rgb_line, y_ptr, s->cb_buf2, s->cr_buf2, w, incr);
break;
case BPG_FORMAT_444:
cb_ptr = (PIXEL *)(s->cb_buf + y * s->cb_linesize);
cr_ptr = (PIXEL *)(s->cr_buf + y * s->cr_linesize);
s->cvt_func(&s->cvt, rgb_line, y_ptr, cb_ptr, cr_ptr, w, incr);
break;
default:
return -1;
}
/* alpha output or CMYK handling */
#ifdef USE_RGB48
if (s->is_cmyk) {
int i;
/* convert RGBW to CMYK */
if (s->is_16bpp) {
if (!s->has_w_plane)
put_dummy_gray16((uint16_t *)rgb_line + 3, w, 4);
for(i = 0; i < 4; i++)
gray_one_minus16((uint16_t *)rgb_line + i, w, 4);
} else {
if (!s->has_w_plane)
put_dummy_gray8(rgb_line + 3, w, 4);
for(i = 0; i < 4; i++)
gray_one_minus8(rgb_line + i, w, 4);
}
} else
#endif
if (s->has_w_plane) {
a_ptr = (PIXEL *)(s->a_buf + y * s->a_linesize);
#ifdef USE_RGB48
if (s->is_16bpp) {
alpha_combine16(&s->cvt, (uint16_t *)rgb_line, a_ptr, w, incr);
if (s->is_rgba)
put_dummy_gray16((uint16_t *)rgb_line + 3, w, 4);
} else
#endif
{
alpha_combine8(&s->cvt, rgb_line, a_ptr, w, incr);
if (s->is_rgba)
put_dummy_gray8(rgb_line + 3, w, 4);
}
} else if (s->is_rgba) {
#ifdef USE_RGB48
if (s->is_16bpp) {
if (s->has_alpha) {
a_ptr = (PIXEL *)(s->a_buf + y * s->a_linesize);
gray_to_gray16(&s->cvt,
(uint16_t *)rgb_line + 3, a_ptr, w, 4);
if (s->premultiplied_alpha)
alpha_divide16((uint16_t *)rgb_line, w);
} else {
put_dummy_gray16((uint16_t *)rgb_line + 3, w, 4);
}
} else
#endif
{
if (s->has_alpha) {
a_ptr = (PIXEL *)(s->a_buf + y * s->a_linesize);
gray_to_gray8(&s->cvt, rgb_line + 3, a_ptr, w, 4);
if (s->premultiplied_alpha)
alpha_divide8((uint8_t *)rgb_line, w);
} else {
put_dummy_gray8(rgb_line + 3, w, 4);
}
}
}
/* go to next line */
s->y++;
return 0;
}
BPGDecoderContext *bpg_decoder_open(void)
{
BPGDecoderContext *s;
s = av_mallocz(sizeof(BPGDecoderContext));
if (!s)
return NULL;
return s;
}
typedef struct {
uint32_t width, height;
BPGImageFormatEnum format;
uint8_t has_alpha;
uint8_t bit_depth;
uint8_t has_w_plane;
uint8_t premultiplied_alpha;
uint8_t limited_range;
uint8_t has_animation;
uint16_t loop_count;
uint16_t frame_delay_num;
uint16_t frame_delay_den;
BPGColorSpaceEnum color_space;
uint32_t hevc_data_len;
BPGExtensionData *first_md;
} BPGHeaderData;
static int bpg_decode_header(BPGHeaderData *h,
const uint8_t *buf, int buf_len,
int header_only, int load_extensions)
{
int idx, flags1, flags2, has_extension, ret, alpha1_flag, alpha2_flag;
uint32_t extension_data_len;
if (buf_len < 6)
return -1;
/* check magic */
if (buf[0] != ((BPG_HEADER_MAGIC >> 24) & 0xff) ||
buf[1] != ((BPG_HEADER_MAGIC >> 16) & 0xff) ||
buf[2] != ((BPG_HEADER_MAGIC >> 8) & 0xff) ||
buf[3] != ((BPG_HEADER_MAGIC >> 0) & 0xff))
return -1;
idx = 4;
flags1 = buf[idx++];
h->format = flags1 >> 5;
if (h->format > 5)
return -1;
alpha1_flag = (flags1 >> 4) & 1;
h->bit_depth = (flags1 & 0xf) + 8;
if (h->bit_depth > 14)
return -1;
flags2 = buf[idx++];
h->color_space = (flags2 >> 4) & 0xf;
has_extension = (flags2 >> 3) & 1;
alpha2_flag = (flags2 >> 2) & 1;
h->limited_range = (flags2 >> 1) & 1;
h->has_animation = flags2 & 1;
h->loop_count = 0;
h->frame_delay_num = 0;
h->frame_delay_den = 0;
h->has_alpha = 0;
h->has_w_plane = 0;
h->premultiplied_alpha = 0;
if (alpha1_flag) {
h->has_alpha = 1;
h->premultiplied_alpha = alpha2_flag;
} else if (alpha2_flag) {
h->has_alpha = 1;
h->has_w_plane = 1;
}
if (h->color_space >= BPG_CS_COUNT ||
(h->format == BPG_FORMAT_GRAY && h->color_space != 0) ||
(h->has_w_plane && h->format == BPG_FORMAT_GRAY))
return -1;
ret = get_ue(&h->width, buf + idx, buf_len - idx);
if (ret < 0)
return -1;
idx += ret;
ret = get_ue(&h->height, buf + idx, buf_len - idx);
if (ret < 0)
return -1;
idx += ret;
if (h->width == 0 || h->height == 0)
return -1;
if (header_only)
return idx;
ret = get_ue(&h->hevc_data_len, buf + idx, buf_len - idx);
if (ret < 0)
return -1;
idx += ret;
extension_data_len = 0;
if (has_extension) {
ret = get_ue(&extension_data_len, buf + idx, buf_len - idx);
if (ret < 0)
return -1;
idx += ret;
}
h->first_md = NULL;
if (has_extension) {
int ext_end;
ext_end = idx + extension_data_len;
if (ext_end > buf_len)
return -1;
if (load_extensions || h->has_animation) {
BPGExtensionData *md, **plast_md;
uint32_t tag, buf_len;
plast_md = &h->first_md;
while (idx < ext_end) {
ret = get_ue32(&tag, buf + idx, ext_end - idx);
if (ret < 0)
goto fail;
idx += ret;
ret = get_ue(&buf_len, buf + idx, ext_end - idx);
if (ret < 0)
goto fail;
idx += ret;
if (idx + buf_len > ext_end) {
fail:
bpg_decoder_free_extension_data(h->first_md);
return -1;
}
if (h->has_animation && tag == BPG_EXTENSION_TAG_ANIM_CONTROL) {
int idx1;
uint32_t loop_count, frame_delay_num, frame_delay_den;
idx1 = idx;
ret = get_ue(&loop_count, buf + idx1, ext_end - idx1);
if (ret < 0)
goto fail;
idx1 += ret;
ret = get_ue(&frame_delay_num, buf + idx1, ext_end - idx1);
if (ret < 0)
goto fail;
idx1 += ret;
ret = get_ue(&frame_delay_den, buf + idx1, ext_end - idx1);
if (ret < 0)
goto fail;
idx1 += ret;
if (frame_delay_num == 0 || frame_delay_den == 0 ||
(uint16_t)frame_delay_num != frame_delay_num ||
(uint16_t)frame_delay_den != frame_delay_den ||
(uint16_t)loop_count != loop_count)
goto fail;
h->loop_count = loop_count;
h->frame_delay_num = frame_delay_num;
h->frame_delay_den = frame_delay_den;
}
if (load_extensions) {
md = av_malloc(sizeof(BPGExtensionData));
md->tag = tag;
md->buf_len = buf_len;
*plast_md = md;
plast_md = &md->next;
md->buf = av_malloc(md->buf_len);
memcpy(md->buf, buf + idx, md->buf_len);
}
idx += buf_len;
}
} else
{
/* skip extension data */
idx += extension_data_len;
}
}
/* must have animation control extension for animations */
if (h->has_animation && h->frame_delay_num == 0)
goto fail;
if (h->hevc_data_len == 0)
h->hevc_data_len = buf_len - idx;
return idx;
}
int bpg_decoder_decode(BPGDecoderContext *img, const uint8_t *buf, int buf_len)
{
int idx, has_alpha, bit_depth, color_space, ret;
uint32_t width, height;
BPGHeaderData h_s, *h = &h_s;
idx = bpg_decode_header(h, buf, buf_len, 0, img->keep_extension_data);
if (idx < 0)
return idx;
width = h->width;
height = h->height;
has_alpha = h->has_alpha;
color_space = h->color_space;
bit_depth = h->bit_depth;
img->w = width;
img->h = height;
img->format = h->format;
if (h->format == BPG_FORMAT_422_VIDEO) {
img->format = BPG_FORMAT_422;
img->c_h_phase = 0;
} else if (h->format == BPG_FORMAT_420_VIDEO) {
img->format = BPG_FORMAT_420;
img->c_h_phase = 0;
} else {
img->format = h->format;
img->c_h_phase = 1;
}
img->has_alpha = has_alpha;
img->premultiplied_alpha = h->premultiplied_alpha;
img->has_w_plane = h->has_w_plane;
img->limited_range = h->limited_range;
img->color_space = color_space;
img->bit_depth = bit_depth;
img->has_animation = h->has_animation;
img->loop_count = h->loop_count;
img->frame_delay_num = h->frame_delay_num;
img->frame_delay_den = h->frame_delay_den;
img->first_md = h->first_md;
if (idx + h->hevc_data_len > buf_len)
goto fail;
/* decode the first frame */
ret = hevc_decode_start(img, buf + idx, buf_len - idx,
width, height, img->format, bit_depth, has_alpha);
if (ret < 0)
goto fail;
idx += ret;
#ifdef USE_PRED
/* XXX: add an option to avoid decoding animations ? */
img->decode_animation = 1;
if (img->has_animation && img->decode_animation) {
int len;
/* keep trailing bitstream to decode the next frames */
len = buf_len - idx;
img->input_buf = av_malloc(len);
if (!img->input_buf)
goto fail;
memcpy(img->input_buf, buf + idx, len);
img->input_buf_len = len;
img->input_buf_pos = 0;
} else
#endif
{
hevc_decode_end(img);
}
if (img->frame->width < img->w || img->frame->height < img->h)
goto fail;
img->y = -1;
return 0;
fail:
av_frame_free(&img->frame);
av_frame_free(&img->alpha_frame);
bpg_decoder_free_extension_data(img->first_md);
img->first_md = NULL;
return -1;
}
void bpg_decoder_close(BPGDecoderContext *s)
{
bpg_decoder_output_end(s);
av_free(s->input_buf);
hevc_decode_end(s);
av_frame_free(&s->frame);
av_frame_free(&s->alpha_frame);
bpg_decoder_free_extension_data(s->first_md);
av_free(s);
}
void bpg_decoder_free_extension_data(BPGExtensionData *first_md)
{
#ifndef EMSCRIPTEN
BPGExtensionData *md, *md_next;
for(md = first_md; md != NULL; md = md_next) {
md_next = md->next;
av_free(md->buf);
av_free(md);
}
#endif
}
#ifndef EMSCRIPTEN
void bpg_decoder_keep_extension_data(BPGDecoderContext *s, int enable)
{
s->keep_extension_data = enable;
}
BPGExtensionData *bpg_decoder_get_extension_data(BPGDecoderContext *s)
{
return s->first_md;
}
int bpg_decoder_get_info_from_buf(BPGImageInfo *p,
BPGExtensionData **pfirst_md,
const uint8_t *buf, int buf_len)
{
BPGHeaderData h_s, *h = &h_s;
int parse_extension;
parse_extension = (pfirst_md != NULL);
if (bpg_decode_header(h, buf, buf_len,
!parse_extension, parse_extension) < 0)
return -1;
p->width = h->width;
p->height = h->height;
p->format = h->format;
p->has_alpha = h->has_alpha && !h->has_w_plane;
p->premultiplied_alpha = h->premultiplied_alpha;
p->has_w_plane = h->has_w_plane;
p->limited_range = h->limited_range;
p->color_space = h->color_space;
p->bit_depth = h->bit_depth;
p->has_animation = h->has_animation;
p->loop_count = h->loop_count;
if (pfirst_md)
*pfirst_md = h->first_md;
return 0;
}
#endif
Reference in a new issue View git blame Copy permalink