| @@ -0,0 +1,384 @@ |
| 1 | +/* |
| 2 | + * ARIB STD-B24 caption decoder using the libaribb24 library |
| 3 | + * Copyright (c) 2019 Jan Ekström |
| 4 | + * |
| 5 | + * This file is part of FFmpeg. |
| 6 | + * |
| 7 | + * FFmpeg is free software; you can redistribute it and/or |
| 8 | + * modify it under the terms of the GNU Lesser General Public |
| 9 | + * License as published by the Free Software Foundation; either |
| 10 | + * version 2.1 of the License, or (at your option) any later version. |
| 11 | + * |
| 12 | + * FFmpeg is distributed in the hope that it will be useful, |
| 13 | + * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 14 | + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
| 15 | + * Lesser General Public License for more details. |
| 16 | + * |
| 17 | + * You should have received a copy of the GNU Lesser General Public |
| 18 | + * License along with FFmpeg; if not, write to the Free Software |
| 19 | + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA |
| 20 | + */ |
| 21 | + |
| 22 | +#include "avcodec.h" |
| 23 | +#include "libavcodec/ass.h" |
| 24 | +#include "libavutil/log.h" |
| 25 | +#include "libavutil/opt.h" |
| 26 | + |
| 27 | +#include <aribb24/aribb24.h> |
| 28 | +#include <aribb24/parser.h> |
| 29 | +#include <aribb24/decoder.h> |
| 30 | + |
| 31 | +typedef struct Libaribb24Context { |
| 32 | + AVClass *class; |
| 33 | + |
| 34 | + arib_instance_t *lib_instance; |
| 35 | + arib_parser_t *parser; |
| 36 | + arib_decoder_t *decoder; |
| 37 | + |
| 38 | + int read_order; |
| 39 | + |
| 40 | + char *aribb24_base_path; |
| 41 | + unsigned int aribb24_skip_ruby; |
| 42 | +} Libaribb24Context; |
| 43 | + |
| 44 | +static unsigned int get_profile_font_size(int profile) |
| 45 | +{ |
| 46 | + switch (profile) { |
| 47 | + case FF_PROFILE_ARIB_PROFILE_A: |
| 48 | + return 36; |
| 49 | + case FF_PROFILE_ARIB_PROFILE_C: |
| 50 | + return 18; |
| 51 | + default: |
| 52 | + return 0; |
| 53 | + } |
| 54 | +} |
| 55 | + |
| 56 | +static void libaribb24_log(void *p, const char *msg) |
| 57 | +{ |
| 58 | + av_log((AVCodecContext *)p, AV_LOG_INFO, "%s\n", msg); |
| 59 | +} |
| 60 | + |
| 61 | +static int libaribb24_generate_ass_header(AVCodecContext *avctx) |
| 62 | +{ |
| 63 | + unsigned int plane_width = 0; |
| 64 | + unsigned int plane_height = 0; |
| 65 | + unsigned int font_size = 0; |
| 66 | + |
| 67 | + switch (avctx->profile) { |
| 68 | + case FF_PROFILE_ARIB_PROFILE_A: |
| 69 | + plane_width = 960; |
| 70 | + plane_height = 540; |
| 71 | + font_size = get_profile_font_size(avctx->profile); |
| 72 | + break; |
| 73 | + case FF_PROFILE_ARIB_PROFILE_C: |
| 74 | + plane_width = 320; |
| 75 | + plane_height = 180; |
| 76 | + font_size = get_profile_font_size(avctx->profile); |
| 77 | + break; |
| 78 | + default: |
| 79 | + av_log(avctx, AV_LOG_ERROR, "Unknown or unsupported profile set!\n"); |
| 80 | + return AVERROR(EINVAL); |
| 81 | + } |
| 82 | + |
| 83 | + avctx->subtitle_header = av_asprintf( |
| 84 | + "[Script Info]\r\n" |
| 85 | + "; Script generated by FFmpeg/Lavc%s\r\n" |
| 86 | + "ScriptType: v4.00+\r\n" |
| 87 | + "PlayResX: %d\r\n" |
| 88 | + "PlayResY: %d\r\n" |
| 89 | + "\r\n" |
| 90 | + "[V4+ Styles]\r\n" |
| 91 | + |
| 92 | + /* ASSv4 header */ |
| 93 | + "Format: Name, " |
| 94 | + "Fontname, Fontsize, " |
| 95 | + "PrimaryColour, SecondaryColour, OutlineColour, BackColour, " |
| 96 | + "Bold, Italic, Underline, StrikeOut, " |
| 97 | + "ScaleX, ScaleY, " |
| 98 | + "Spacing, Angle, " |
| 99 | + "BorderStyle, Outline, Shadow, " |
| 100 | + "Alignment, MarginL, MarginR, MarginV, " |
| 101 | + "Encoding\r\n" |
| 102 | + |
| 103 | + "Style: " |
| 104 | + "Default," /* Name */ |
| 105 | + "%s,%d," /* Font{name,size} */ |
| 106 | + "&H%x,&H%x,&H%x,&H%x," /* {Primary,Secondary,Outline,Back}Colour */ |
| 107 | + "%d,%d,%d,0," /* Bold, Italic, Underline, StrikeOut */ |
| 108 | + "100,100," /* Scale{X,Y} */ |
| 109 | + "0,0," /* Spacing, Angle */ |
| 110 | + "%d,1,0," /* BorderStyle, Outline, Shadow */ |
| 111 | + "%d,10,10,10," /* Alignment, Margin[LRV] */ |
| 112 | + "0\r\n" /* Encoding */ |
| 113 | + |
| 114 | + "\r\n" |
| 115 | + "[Events]\r\n" |
| 116 | + "Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text\r\n", |
| 117 | + !(avctx->flags & AV_CODEC_FLAG_BITEXACT) ? AV_STRINGIFY(LIBAVCODEC_VERSION) : "", |
| 118 | + plane_width, plane_height, |
| 119 | + ASS_DEFAULT_FONT, font_size, ASS_DEFAULT_COLOR, |
| 120 | + ASS_DEFAULT_COLOR, ASS_DEFAULT_BACK_COLOR, ASS_DEFAULT_BACK_COLOR, |
| 121 | + -ASS_DEFAULT_BOLD, -ASS_DEFAULT_ITALIC, -ASS_DEFAULT_UNDERLINE, |
| 122 | + ASS_DEFAULT_BORDERSTYLE, ASS_DEFAULT_ALIGNMENT); |
| 123 | + |
| 124 | + if (!avctx->subtitle_header) |
| 125 | + return AVERROR(ENOMEM); |
| 126 | + |
| 127 | + avctx->subtitle_header_size = strlen(avctx->subtitle_header); |
| 128 | + |
| 129 | + return 0; |
| 130 | +} |
| 131 | + |
| 132 | +static int libaribb24_init(AVCodecContext *avctx) |
| 133 | +{ |
| 134 | + Libaribb24Context *b24 = avctx->priv_data; |
| 135 | + void(* arib_dec_init)(arib_decoder_t* decoder) = NULL; |
| 136 | + int ret_code = AVERROR_EXTERNAL; |
| 137 | + |
| 138 | + if (!(b24->lib_instance = arib_instance_new(avctx))) { |
| 139 | + av_log(avctx, AV_LOG_ERROR, "Failed to initialize libaribb24!\n"); |
| 140 | + goto init_fail; |
| 141 | + } |
| 142 | + |
| 143 | + if (b24->aribb24_base_path) { |
| 144 | + av_log(avctx, AV_LOG_INFO, "Setting the libaribb24 base path to '%s'\n", |
| 145 | + b24->aribb24_base_path); |
| 146 | + arib_set_base_path(b24->lib_instance, b24->aribb24_base_path); |
| 147 | + } |
| 148 | + |
| 149 | + arib_register_messages_callback(b24->lib_instance, libaribb24_log); |
| 150 | + |
| 151 | + if (!(b24->parser = arib_get_parser(b24->lib_instance))) { |
| 152 | + av_log(avctx, AV_LOG_ERROR, "Failed to initialize libaribb24 PES parser!\n"); |
| 153 | + goto init_fail; |
| 154 | + } |
| 155 | + if (!(b24->decoder = arib_get_decoder(b24->lib_instance))) { |
| 156 | + av_log(avctx, AV_LOG_ERROR, "Failed to initialize libaribb24 decoder!\n"); |
| 157 | + goto init_fail; |
| 158 | + } |
| 159 | + |
| 160 | + switch (avctx->profile) { |
| 161 | + case FF_PROFILE_ARIB_PROFILE_A: |
| 162 | + arib_dec_init = arib_initialize_decoder_a_profile; |
| 163 | + break; |
| 164 | + case FF_PROFILE_ARIB_PROFILE_C: |
| 165 | + arib_dec_init = arib_initialize_decoder_c_profile; |
| 166 | + break; |
| 167 | + default: |
| 168 | + av_log(avctx, AV_LOG_ERROR, "Unknown or unsupported profile set!\n"); |
| 169 | + ret_code = AVERROR(EINVAL); |
| 170 | + goto init_fail; |
| 171 | + } |
| 172 | + |
| 173 | + arib_dec_init(b24->decoder); |
| 174 | + |
| 175 | + if (libaribb24_generate_ass_header(avctx) < 0) { |
| 176 | + ret_code = AVERROR(ENOMEM); |
| 177 | + goto init_fail; |
| 178 | + } |
| 179 | + |
| 180 | + return 0; |
| 181 | + |
| 182 | +init_fail: |
| 183 | + if (b24->decoder) |
| 184 | + arib_finalize_decoder(b24->decoder); |
| 185 | + |
| 186 | + if (b24->lib_instance) |
| 187 | + arib_instance_destroy(b24->lib_instance); |
| 188 | + |
| 189 | + return ret_code; |
| 190 | +} |
| 191 | + |
| 192 | +static int libaribb24_close(AVCodecContext *avctx) |
| 193 | +{ |
| 194 | + Libaribb24Context *b24 = avctx->priv_data; |
| 195 | + |
| 196 | + if (b24->decoder) |
| 197 | + arib_finalize_decoder(b24->decoder); |
| 198 | + |
| 199 | + if (b24->lib_instance) |
| 200 | + arib_instance_destroy(b24->lib_instance); |
| 201 | + |
| 202 | + return 0; |
| 203 | +} |
| 204 | + |
| 205 | +#define RGB_TO_BGR(c) ((c & 0xff) << 16 | (c & 0xff00) | ((c >> 16) & 0xff)) |
| 206 | + |
| 207 | +static void libaribb24_handle_regions(AVCodecContext *avctx, AVSubtitle *sub) |
| 208 | +{ |
| 209 | + Libaribb24Context *b24 = avctx->priv_data; |
| 210 | + const arib_buf_region_t *region = arib_decoder_get_regions(b24->decoder); |
| 211 | + unsigned int profile_font_size = get_profile_font_size(avctx->profile); |
| 212 | + AVBPrint buf = { 0 }; |
| 213 | + |
| 214 | + av_bprint_init(&buf, 0, AV_BPRINT_SIZE_UNLIMITED); |
| 215 | + |
| 216 | + while (region) { |
| 217 | + ptrdiff_t region_length = region->p_end - region->p_start; |
| 218 | + unsigned int ruby_region = |
| 219 | + region->i_fontheight == (profile_font_size / 2); |
| 220 | + |
| 221 | + // ASS requires us to make the colors BGR, so we convert here |
| 222 | + int foreground_bgr_color = RGB_TO_BGR(region->i_foreground_color); |
| 223 | + int background_bgr_color = RGB_TO_BGR(region->i_background_color); |
| 224 | + |
| 225 | + if (region_length < 0) { |
| 226 | + av_log(avctx, AV_LOG_ERROR, "Invalid negative region length!\n"); |
| 227 | + break; |
| 228 | + } |
| 229 | + |
| 230 | + if (region_length == 0 || (ruby_region && b24->aribb24_skip_ruby)) { |
| 231 | + goto next_region; |
| 232 | + } |
| 233 | + |
| 234 | + // color and alpha |
| 235 | + if (foreground_bgr_color != ASS_DEFAULT_COLOR) |
| 236 | + av_bprintf(&buf, "{\\1c&H%06x&}", foreground_bgr_color); |
| 237 | + |
| 238 | + if (region->i_foreground_alpha != 0) |
| 239 | + av_bprintf(&buf, "{\\1a&H%02x&}", region->i_foreground_alpha); |
| 240 | + |
| 241 | + if (background_bgr_color != ASS_DEFAULT_BACK_COLOR) |
| 242 | + av_bprintf(&buf, "{\\3c&H%06x&}", background_bgr_color); |
| 243 | + |
| 244 | + if (region->i_background_alpha != 0) |
| 245 | + av_bprintf(&buf, "{\\3a&H%02x&}", region->i_background_alpha); |
| 246 | + |
| 247 | + // font size |
| 248 | + if (region->i_fontwidth != profile_font_size || |
| 249 | + region->i_fontheight != profile_font_size) { |
| 250 | + av_bprintf(&buf, "{\\fscx%d\\fscy%d}", |
| 251 | + (int)round(((double)region->i_fontwidth / |
| 252 | + (double)profile_font_size) * 100), |
| 253 | + (int)round(((double)region->i_fontheight / |
| 254 | + (double)profile_font_size) * 100)); |
| 255 | + } |
| 256 | + |
| 257 | + // TODO: positioning |
| 258 | + |
| 259 | + av_bprint_append_data(&buf, region->p_start, region_length); |
| 260 | + |
| 261 | + av_bprintf(&buf, "{\\r}"); |
| 262 | + |
| 263 | +next_region: |
| 264 | + region = region->p_next; |
| 265 | + } |
| 266 | + |
| 267 | + av_log(avctx, AV_LOG_DEBUG, "Styled ASS line: %s\n", |
| 268 | + buf.str); |
| 269 | + ff_ass_add_rect(sub, buf.str, b24->read_order++, |
| 270 | + 0, NULL, NULL); |
| 271 | + |
| 272 | + av_bprint_finalize(&buf, NULL); |
| 273 | +} |
| 274 | + |
| 275 | +static int libaribb24_decode(AVCodecContext *avctx, void *data, int *got_sub_ptr, AVPacket *pkt) |
| 276 | +{ |
| 277 | + Libaribb24Context *b24 = avctx->priv_data; |
| 278 | + AVSubtitle *sub = data; |
| 279 | + size_t parsed_data_size = 0; |
| 280 | + size_t decoded_subtitle_size = 0; |
| 281 | + const unsigned char *parsed_data = NULL; |
| 282 | + char *decoded_subtitle = NULL; |
| 283 | + time_t subtitle_duration = 0; |
| 284 | + |
| 285 | + if (pkt->size <= 0) |
| 286 | + return pkt->size; |
| 287 | + |
| 288 | + arib_parse_pes(b24->parser, pkt->data, pkt->size); |
| 289 | + |
| 290 | + parsed_data = arib_parser_get_data(b24->parser, |
| 291 | + &parsed_data_size); |
| 292 | + if (!parsed_data || !parsed_data_size) { |
| 293 | + av_log(avctx, AV_LOG_DEBUG, "No decode'able data was received from " |
| 294 | + "packet (dts: %"PRId64", pts: %"PRId64").\n", |
| 295 | + pkt->dts, pkt->pts); |
| 296 | + return pkt->size; |
| 297 | + } |
| 298 | + |
| 299 | + decoded_subtitle_size = parsed_data_size * 4; |
| 300 | + if (!(decoded_subtitle = av_mallocz(decoded_subtitle_size + 1))) { |
| 301 | + av_log(avctx, AV_LOG_ERROR, |
| 302 | + "Failed to allocate buffer for decoded subtitle!\n"); |
| 303 | + return AVERROR(ENOMEM); |
| 304 | + } |
| 305 | + |
| 306 | + decoded_subtitle_size = arib_decode_buffer(b24->decoder, |
| 307 | + parsed_data, |
| 308 | + parsed_data_size, |
| 309 | + decoded_subtitle, |
| 310 | + decoded_subtitle_size); |
| 311 | + |
| 312 | + subtitle_duration = arib_decoder_get_time(b24->decoder); |
| 313 | + |
| 314 | + if (avctx->pkt_timebase.num && pkt->pts != AV_NOPTS_VALUE) |
| 315 | + sub->pts = av_rescale_q(pkt->pts, |
| 316 | + avctx->pkt_timebase, AV_TIME_BASE_Q); |
| 317 | + |
| 318 | + sub->end_display_time = subtitle_duration ? |
| 319 | + av_rescale_q(subtitle_duration, |
| 320 | + AV_TIME_BASE_Q, |
| 321 | + (AVRational){1, 1000}) : |
| 322 | + UINT32_MAX; |
| 323 | + |
| 324 | + av_log(avctx, AV_LOG_DEBUG, |
| 325 | + "Result: '%s' (size: %zu, pkt_pts: %"PRId64", sub_pts: %"PRId64" " |
| 326 | + "duration: %"PRIu32", pkt_timebase: %d/%d, time_base: %d/%d')\n", |
| 327 | + decoded_subtitle ? decoded_subtitle : "<no subtitle>", |
| 328 | + decoded_subtitle_size, |
| 329 | + pkt->pts, sub->pts, |
| 330 | + sub->end_display_time, |
| 331 | + avctx->pkt_timebase.num, avctx->pkt_timebase.den, |
| 332 | + avctx->time_base.num, avctx->time_base.den); |
| 333 | + |
| 334 | + if (decoded_subtitle) |
| 335 | + libaribb24_handle_regions(avctx, sub); |
| 336 | + |
| 337 | + *got_sub_ptr = sub->num_rects > 0; |
| 338 | + |
| 339 | + av_free(decoded_subtitle); |
| 340 | + |
| 341 | + // flush the region buffers, otherwise the linked list keeps getting |
| 342 | + // longer and longer... |
| 343 | + arib_finalize_decoder(b24->decoder); |
| 344 | + |
| 345 | + return pkt->size; |
| 346 | +} |
| 347 | + |
| 348 | +static void libaribb24_flush(AVCodecContext *avctx) |
| 349 | +{ |
| 350 | + Libaribb24Context *b24 = avctx->priv_data; |
| 351 | + if (!(avctx->flags2 & AV_CODEC_FLAG2_RO_FLUSH_NOOP)) |
| 352 | + b24->read_order = 0; |
| 353 | +} |
| 354 | + |
| 355 | +#define OFFSET(x) offsetof(Libaribb24Context, x) |
| 356 | +#define SD AV_OPT_FLAG_SUBTITLE_PARAM | AV_OPT_FLAG_DECODING_PARAM |
| 357 | +static const AVOption options[] = { |
| 358 | + { "aribb24-base-path", "set the base path for the libaribb24 library", |
| 359 | + OFFSET(aribb24_base_path), AV_OPT_TYPE_STRING, { 0 }, 0, 0, SD }, |
| 360 | + { "aribb24-skip-ruby-text", "skip ruby text blocks during decoding", |
| 361 | + OFFSET(aribb24_skip_ruby), AV_OPT_TYPE_BOOL, { 1 }, 0, 1, SD }, |
| 362 | + { NULL } |
| 363 | +}; |
| 364 | + |
| 365 | +static const AVClass aribb24_class = { |
| 366 | + .class_name = "libaribb24 decoder", |
| 367 | + .item_name = av_default_item_name, |
| 368 | + .option = options, |
| 369 | + .version = LIBAVUTIL_VERSION_INT, |
| 370 | +}; |
| 371 | + |
| 372 | +AVCodec ff_libaribb24_decoder = { |
| 373 | + .name = "libaribb24", |
| 374 | + .long_name = NULL_IF_CONFIG_SMALL("libaribb24 ARIB STD-B24 caption decoder"), |
| 375 | + .type = AVMEDIA_TYPE_SUBTITLE, |
| 376 | + .id = AV_CODEC_ID_ARIB_CAPTION, |
| 377 | + .priv_data_size = sizeof(Libaribb24Context), |
| 378 | + .init = libaribb24_init, |
| 379 | + .close = libaribb24_close, |
| 380 | + .decode = libaribb24_decode, |
| 381 | + .flush = libaribb24_flush, |
| 382 | + .priv_class= &aribb24_class, |
| 383 | + .wrapper_name = "libaribb24", |
| 384 | +}; |