| 1 | /* |
| 2 | * SSA/ASS spliting functions |
| 3 | * Copyright (c) 2010 Aurelien Jacobs <aurel@gnuage.org> |
| 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 "ass_split.h" |
| 24 | |
| 25 | typedef enum { |
| 26 | ASS_STR, |
| 27 | ASS_INT, |
| 28 | ASS_FLT, |
| 29 | ASS_COLOR, |
| 30 | ASS_TIMESTAMP, |
| 31 | ASS_ALGN, |
| 32 | } ASSFieldType; |
| 33 | |
| 34 | typedef struct { |
| 35 | const char *name; |
| 36 | int type; |
| 37 | int offset; |
| 38 | } ASSFields; |
| 39 | |
| 40 | typedef struct { |
| 41 | const char *section; |
| 42 | const char *format_header; |
| 43 | const char *fields_header; |
| 44 | int size; |
| 45 | int offset; |
| 46 | int offset_count; |
| 47 | ASSFields fields[10]; |
| 48 | } ASSSection; |
| 49 | |
| 50 | static const ASSSection ass_sections[] = { |
| 51 | { .section = "Script Info", |
| 52 | .offset = offsetof(ASS, script_info), |
| 53 | .fields = {{"ScriptType", ASS_STR, offsetof(ASSScriptInfo, script_type)}, |
| 54 | {"Collisions", ASS_STR, offsetof(ASSScriptInfo, collisions) }, |
| 55 | {"PlayResX", ASS_INT, offsetof(ASSScriptInfo, play_res_x) }, |
| 56 | {"PlayResY", ASS_INT, offsetof(ASSScriptInfo, play_res_y) }, |
| 57 | {"Timer", ASS_FLT, offsetof(ASSScriptInfo, timer) }, |
| 58 | {0}, |
| 59 | } |
| 60 | }, |
| 61 | { .section = "V4+ Styles", |
| 62 | .format_header = "Format", |
| 63 | .fields_header = "Style", |
| 64 | .size = sizeof(ASSStyle), |
| 65 | .offset = offsetof(ASS, styles), |
| 66 | .offset_count = offsetof(ASS, styles_count), |
| 67 | .fields = {{"Name", ASS_STR, offsetof(ASSStyle, name) }, |
| 68 | {"Fontname", ASS_STR, offsetof(ASSStyle, font_name) }, |
| 69 | {"Fontsize", ASS_INT, offsetof(ASSStyle, font_size) }, |
| 70 | {"PrimaryColour",ASS_COLOR,offsetof(ASSStyle, primary_color)}, |
| 71 | {"BackColour", ASS_COLOR,offsetof(ASSStyle, back_color) }, |
| 72 | {"Bold", ASS_INT, offsetof(ASSStyle, bold) }, |
| 73 | {"Italic", ASS_INT, offsetof(ASSStyle, italic) }, |
| 74 | {"Underline", ASS_INT, offsetof(ASSStyle, underline) }, |
| 75 | {"Alignment", ASS_INT, offsetof(ASSStyle, alignment) }, |
| 76 | {0}, |
| 77 | } |
| 78 | }, |
| 79 | { .section = "V4 Styles", |
| 80 | .format_header = "Format", |
| 81 | .fields_header = "Style", |
| 82 | .size = sizeof(ASSStyle), |
| 83 | .offset = offsetof(ASS, styles), |
| 84 | .offset_count = offsetof(ASS, styles_count), |
| 85 | .fields = {{"Name", ASS_STR, offsetof(ASSStyle, name) }, |
| 86 | {"Fontname", ASS_STR, offsetof(ASSStyle, font_name) }, |
| 87 | {"Fontsize", ASS_INT, offsetof(ASSStyle, font_size) }, |
| 88 | {"PrimaryColour",ASS_COLOR,offsetof(ASSStyle, primary_color)}, |
| 89 | {"BackColour", ASS_COLOR,offsetof(ASSStyle, back_color) }, |
| 90 | {"Bold", ASS_INT, offsetof(ASSStyle, bold) }, |
| 91 | {"Italic", ASS_INT, offsetof(ASSStyle, italic) }, |
| 92 | {"Alignment", ASS_ALGN, offsetof(ASSStyle, alignment) }, |
| 93 | {0}, |
| 94 | } |
| 95 | }, |
| 96 | { .section = "Events", |
| 97 | .format_header = "Format", |
| 98 | .fields_header = "Dialogue", |
| 99 | .size = sizeof(ASSDialog), |
| 100 | .offset = offsetof(ASS, dialogs), |
| 101 | .offset_count = offsetof(ASS, dialogs_count), |
| 102 | .fields = {{"Layer", ASS_INT, offsetof(ASSDialog, layer) }, |
| 103 | {"Start", ASS_TIMESTAMP, offsetof(ASSDialog, start) }, |
| 104 | {"End", ASS_TIMESTAMP, offsetof(ASSDialog, end) }, |
| 105 | {"Style", ASS_STR, offsetof(ASSDialog, style) }, |
| 106 | {"Text", ASS_STR, offsetof(ASSDialog, text) }, |
| 107 | {0}, |
| 108 | } |
| 109 | }, |
| 110 | }; |
| 111 | |
| 112 | |
| 113 | typedef int (*ASSConvertFunc)(void *dest, const char *buf, int len); |
| 114 | |
| 115 | static int convert_str(void *dest, const char *buf, int len) |
| 116 | { |
| 117 | char *str = av_malloc(len + 1); |
| 118 | if (str) { |
| 119 | memcpy(str, buf, len); |
| 120 | str[len] = 0; |
| 121 | if (*(void **)dest) |
| 122 | av_free(*(void **)dest); |
| 123 | *(char **)dest = str; |
| 124 | } |
| 125 | return !str; |
| 126 | } |
| 127 | static int convert_int(void *dest, const char *buf, int len) |
| 128 | { |
| 129 | return sscanf(buf, "%d", (int *)dest) == 1; |
| 130 | } |
| 131 | static int convert_flt(void *dest, const char *buf, int len) |
| 132 | { |
| 133 | return sscanf(buf, "%f", (float *)dest) == 1; |
| 134 | } |
| 135 | static int convert_color(void *dest, const char *buf, int len) |
| 136 | { |
| 137 | return sscanf(buf, "&H%8x", (int *)dest) == 1 || |
| 138 | sscanf(buf, "%d", (int *)dest) == 1; |
| 139 | } |
| 140 | static int convert_timestamp(void *dest, const char *buf, int len) |
| 141 | { |
| 142 | int c, h, m, s, cs; |
| 143 | if ((c = sscanf(buf, "%d:%02d:%02d.%02d", &h, &m, &s, &cs)) == 4) |
| 144 | *(int *)dest = 360000*h + 6000*m + 100*s + cs; |
| 145 | return c == 4; |
| 146 | } |
| 147 | static int convert_alignment(void *dest, const char *buf, int len) |
| 148 | { |
| 149 | int a; |
| 150 | if (sscanf(buf, "%d", &a) == 1) { |
| 151 | /* convert V4 Style alignment to V4+ Style */ |
| 152 | *(int *)dest = a + ((a&4) >> 1) - 5*!!(a&8); |
| 153 | return 1; |
| 154 | } |
| 155 | return 0; |
| 156 | } |
| 157 | |
| 158 | static const ASSConvertFunc convert_func[] = { |
| 159 | [ASS_STR] = convert_str, |
| 160 | [ASS_INT] = convert_int, |
| 161 | [ASS_FLT] = convert_flt, |
| 162 | [ASS_COLOR] = convert_color, |
| 163 | [ASS_TIMESTAMP] = convert_timestamp, |
| 164 | [ASS_ALGN] = convert_alignment, |
| 165 | }; |
| 166 | |
| 167 | |
| 168 | struct ASSSplitContext { |
| 169 | ASS ass; |
| 170 | int current_section; |
| 171 | int field_number[FF_ARRAY_ELEMS(ass_sections)]; |
| 172 | int *field_order[FF_ARRAY_ELEMS(ass_sections)]; |
| 173 | }; |
| 174 | |
| 175 | |
| 176 | static uint8_t *realloc_section_array(ASSSplitContext *ctx) |
| 177 | { |
| 178 | const ASSSection *section = &ass_sections[ctx->current_section]; |
| 179 | int *count = (int *)((uint8_t *)&ctx->ass + section->offset_count); |
| 180 | void **section_ptr = (void **)((uint8_t *)&ctx->ass + section->offset); |
| 181 | uint8_t *tmp = av_realloc(*section_ptr, (*count+1)*section->size); |
| 182 | if (!tmp) |
| 183 | return NULL; |
| 184 | *section_ptr = tmp; |
| 185 | tmp += *count * section->size; |
| 186 | memset(tmp, 0, section->size); |
| 187 | (*count)++; |
| 188 | return tmp; |
| 189 | } |
| 190 | |
| 191 | static inline int is_eol(char buf) |
| 192 | { |
| 193 | return buf == '\r' || buf == '\n' || buf == 0; |
| 194 | } |
| 195 | |
| 196 | static inline const char *skip_space(const char *buf) |
| 197 | { |
| 198 | while (*buf == ' ') |
| 199 | buf++; |
| 200 | return buf; |
| 201 | } |
| 202 | |
| 203 | static const char *ass_split_section(ASSSplitContext *ctx, const char *buf) |
| 204 | { |
| 205 | const ASSSection *section = &ass_sections[ctx->current_section]; |
| 206 | int *number = &ctx->field_number[ctx->current_section]; |
| 207 | int *order = ctx->field_order[ctx->current_section]; |
| 208 | int *tmp, i, len; |
| 209 | |
| 210 | while (buf && *buf) { |
| 211 | if (buf[0] == '[') { |
| 212 | ctx->current_section = -1; |
| 213 | break; |
| 214 | } |
| 215 | if (buf[0] == ';' || (buf[0] == '!' && buf[1] == ':')) { |
| 216 | /* skip comments */ |
| 217 | } else if (section->format_header && !order) { |
| 218 | len = strlen(section->format_header); |
| 219 | if (strncmp(buf, section->format_header, len) || buf[len] != ':') |
| 220 | return NULL; |
| 221 | buf += len + 1; |
| 222 | while (!is_eol(*buf)) { |
| 223 | buf = skip_space(buf); |
| 224 | len = strcspn(buf, ", \r\n"); |
| 225 | if (!(tmp = av_realloc(order, (*number + 1) * sizeof(*order)))) |
| 226 | return NULL; |
| 227 | order = tmp; |
| 228 | order[*number] = -1; |
| 229 | for (i=0; section->fields[i].name; i++) |
| 230 | if (!strncmp(buf, section->fields[i].name, len)) { |
| 231 | order[*number] = i; |
| 232 | break; |
| 233 | } |
| 234 | (*number)++; |
| 235 | buf = skip_space(buf + len + (buf[len] == ',')); |
| 236 | } |
| 237 | ctx->field_order[ctx->current_section] = order; |
| 238 | } else if (section->fields_header) { |
| 239 | len = strlen(section->fields_header); |
| 240 | if (!strncmp(buf, section->fields_header, len) && buf[len] == ':') { |
| 241 | uint8_t *ptr, *struct_ptr = realloc_section_array(ctx); |
| 242 | if (!struct_ptr) return NULL; |
| 243 | buf += len + 1; |
| 244 | for (i=0; !is_eol(*buf) && i < *number; i++) { |
| 245 | int last = i == *number - 1; |
| 246 | buf = skip_space(buf); |
| 247 | len = strcspn(buf, last ? "\r\n" : ",\r\n"); |
| 248 | if (order[i] >= 0) { |
| 249 | ASSFieldType type = section->fields[order[i]].type; |
| 250 | ptr = struct_ptr + section->fields[order[i]].offset; |
| 251 | convert_func[type](ptr, buf, len); |
| 252 | } |
| 253 | buf += len; |
| 254 | if (!last && *buf) buf++; |
| 255 | buf = skip_space(buf); |
| 256 | } |
| 257 | } |
| 258 | } else { |
| 259 | len = strcspn(buf, ":\r\n"); |
| 260 | if (buf[len] == ':') { |
| 261 | for (i=0; section->fields[i].name; i++) |
| 262 | if (!strncmp(buf, section->fields[i].name, len)) { |
| 263 | ASSFieldType type = section->fields[i].type; |
| 264 | uint8_t *ptr = (uint8_t *)&ctx->ass + section->offset; |
| 265 | ptr += section->fields[i].offset; |
| 266 | buf = skip_space(buf + len + 1); |
| 267 | convert_func[type](ptr, buf, strcspn(buf, "\r\n")); |
| 268 | break; |
| 269 | } |
| 270 | } |
| 271 | } |
| 272 | buf += strcspn(buf, "\n"); |
| 273 | buf += !!*buf; |
| 274 | } |
| 275 | return buf; |
| 276 | } |
| 277 | |
| 278 | static int ass_split(ASSSplitContext *ctx, const char *buf) |
| 279 | { |
| 280 | char c, section[16]; |
| 281 | int i; |
| 282 | |
| 283 | if (ctx->current_section >= 0) |
| 284 | buf = ass_split_section(ctx, buf); |
| 285 | |
| 286 | while (buf && *buf) { |
| 287 | if (sscanf(buf, "[%15[0-9A-Za-z+ ]]%c", section, &c) == 2) { |
| 288 | buf += strcspn(buf, "\n"); |
| 289 | buf += !!*buf; |
| 290 | for (i=0; i<FF_ARRAY_ELEMS(ass_sections); i++) |
| 291 | if (!strcmp(section, ass_sections[i].section)) { |
| 292 | ctx->current_section = i; |
| 293 | buf = ass_split_section(ctx, buf); |
| 294 | } |
| 295 | } else { |
| 296 | buf += strcspn(buf, "\n"); |
| 297 | buf += !!*buf; |
| 298 | } |
| 299 | } |
| 300 | return buf ? 0 : AVERROR_INVALIDDATA; |
| 301 | } |
| 302 | |
| 303 | ASSSplitContext *ff_ass_split(const char *buf) |
| 304 | { |
| 305 | ASSSplitContext *ctx = av_mallocz(sizeof(*ctx)); |
| 306 | ctx->current_section = -1; |
| 307 | if (ass_split(ctx, buf) < 0) { |
| 308 | ff_ass_split_free(ctx); |
| 309 | return NULL; |
| 310 | } |
| 311 | return ctx; |
| 312 | } |
| 313 | |
| 314 | static void free_section(ASSSplitContext *ctx, const ASSSection *section) |
| 315 | { |
| 316 | uint8_t *ptr = (uint8_t *)&ctx->ass + section->offset; |
| 317 | int i, j, *count, c = 1; |
| 318 | |
| 319 | if (section->format_header) { |
| 320 | ptr = *(void **)ptr; |
| 321 | count = (int *)((uint8_t *)&ctx->ass + section->offset_count); |
| 322 | } else |
| 323 | count = &c; |
| 324 | |
| 325 | if (ptr) |
| 326 | for (i=0; i<*count; i++, ptr += section->size) |
| 327 | for (j=0; section->fields[j].name; j++) { |
| 328 | const ASSFields *field = §ion->fields[j]; |
| 329 | if (field->type == ASS_STR) |
| 330 | av_freep(ptr + field->offset); |
| 331 | } |
| 332 | *count = 0; |
| 333 | |
| 334 | if (section->format_header) |
| 335 | av_freep((uint8_t *)&ctx->ass + section->offset); |
| 336 | } |
| 337 | |
| 338 | ASSDialog *ff_ass_split_dialog(ASSSplitContext *ctx, const char *buf, |
| 339 | int cache, int *number) |
| 340 | { |
| 341 | ASSDialog *dialog = NULL; |
| 342 | int i, count; |
| 343 | if (!cache) |
| 344 | for (i=0; i<FF_ARRAY_ELEMS(ass_sections); i++) |
| 345 | if (!strcmp(ass_sections[i].section, "Events")) { |
| 346 | free_section(ctx, &ass_sections[i]); |
| 347 | break; |
| 348 | } |
| 349 | count = ctx->ass.dialogs_count; |
| 350 | if (ass_split(ctx, buf) == 0) |
| 351 | dialog = ctx->ass.dialogs + count; |
| 352 | if (number) |
| 353 | *number = ctx->ass.dialogs_count - count; |
| 354 | return dialog; |
| 355 | } |
| 356 | |
| 357 | void ff_ass_split_free(ASSSplitContext *ctx) |
| 358 | { |
| 359 | if (ctx) { |
| 360 | int i; |
| 361 | for (i=0; i<FF_ARRAY_ELEMS(ass_sections); i++) { |
| 362 | free_section(ctx, &ass_sections[i]); |
| 363 | av_freep(&(ctx->field_order[i])); |
| 364 | } |
| 365 | av_free(ctx); |
| 366 | } |
| 367 | } |
| 368 | |
| 369 | |
| 370 | int ff_ass_split_override_codes(const ASSCodesCallbacks *callbacks, void *priv, |
| 371 | const char *buf) |
| 372 | { |
| 373 | const char *text = NULL; |
| 374 | char new_line[2]; |
| 375 | int text_len = 0; |
| 376 | |
| 377 | while (buf && *buf) { |
| 378 | if (text && callbacks->text && |
| 379 | (sscanf(buf, "\\%1[nN]", new_line) == 1 || |
| 380 | !strncmp(buf, "{\\", 2))) { |
| 381 | callbacks->text(priv, text, text_len); |
| 382 | text = NULL; |
| 383 | } |
| 384 | if (sscanf(buf, "\\%1[nN]", new_line) == 1) { |
| 385 | if (callbacks->new_line) |
| 386 | callbacks->new_line(priv, new_line[0] == 'N'); |
| 387 | buf += 2; |
| 388 | } else if (!strncmp(buf, "{\\", 2)) { |
| 389 | buf++; |
| 390 | while (*buf == '\\') { |
| 391 | char style[2], c[2], sep[2], c_num[2] = "0", tmp[128] = {0}; |
| 392 | unsigned int color = 0xFFFFFFFF; |
| 393 | int len, size = -1, an = -1, alpha = -1; |
| 394 | int x1, y1, x2, y2, t1 = -1, t2 = -1; |
| 395 | if (sscanf(buf, "\\%1[bisu]%1[01\\}]%n", style, c, &len) > 1) { |
| 396 | int close = c[0] == '0' ? 1 : c[0] == '1' ? 0 : -1; |
| 397 | len += close != -1; |
| 398 | if (callbacks->style) |
| 399 | callbacks->style(priv, style[0], close); |
| 400 | } else if (sscanf(buf, "\\c%1[\\}]%n", sep, &len) > 0 || |
| 401 | sscanf(buf, "\\c&H%X&%1[\\}]%n", &color, sep, &len) > 1 || |
| 402 | sscanf(buf, "\\%1[1234]c%1[\\}]%n", c_num, sep, &len) > 1 || |
| 403 | sscanf(buf, "\\%1[1234]c&H%X&%1[\\}]%n", c_num, &color, sep, &len) > 2) { |
| 404 | if (callbacks->color) |
| 405 | callbacks->color(priv, color, c_num[0] - '0'); |
| 406 | } else if (sscanf(buf, "\\alpha%1[\\}]%n", sep, &len) > 0 || |
| 407 | sscanf(buf, "\\alpha&H%2X&%1[\\}]%n", &alpha, sep, &len) > 1 || |
| 408 | sscanf(buf, "\\%1[1234]a%1[\\}]%n", c_num, sep, &len) > 1 || |
| 409 | sscanf(buf, "\\%1[1234]a&H%2X&%1[\\}]%n", c_num, &alpha, sep, &len) > 2) { |
| 410 | if (callbacks->alpha) |
| 411 | callbacks->alpha(priv, alpha, c_num[0] - '0'); |
| 412 | } else if (sscanf(buf, "\\fn%1[\\}]%n", sep, &len) > 0 || |
| 413 | sscanf(buf, "\\fn%127[^\\}]%1[\\}]%n", tmp, sep, &len) > 1) { |
| 414 | if (callbacks->font_name) |
| 415 | callbacks->font_name(priv, tmp[0] ? tmp : NULL); |
| 416 | } else if (sscanf(buf, "\\fs%1[\\}]%n", sep, &len) > 0 || |
| 417 | sscanf(buf, "\\fs%u%1[\\}]%n", &size, sep, &len) > 1) { |
| 418 | if (callbacks->font_size) |
| 419 | callbacks->font_size(priv, size); |
| 420 | } else if (sscanf(buf, "\\a%1[\\}]%n", sep, &len) > 0 || |
| 421 | sscanf(buf, "\\a%2u%1[\\}]%n", &an, sep, &len) > 1 || |
| 422 | sscanf(buf, "\\an%1[\\}]%n", sep, &len) > 0 || |
| 423 | sscanf(buf, "\\an%1u%1[\\}]%n", &an, sep, &len) > 1) { |
| 424 | if (an != -1 && buf[2] != 'n') |
| 425 | an = (an&3) + (an&4 ? 6 : an&8 ? 3 : 0); |
| 426 | if (callbacks->alignment) |
| 427 | callbacks->alignment(priv, an); |
| 428 | } else if (sscanf(buf, "\\r%1[\\}]%n", sep, &len) > 0 || |
| 429 | sscanf(buf, "\\r%127[^\\}]%1[\\}]%n", tmp, sep, &len) > 1) { |
| 430 | if (callbacks->cancel_overrides) |
| 431 | callbacks->cancel_overrides(priv, tmp); |
| 432 | } else if (sscanf(buf, "\\move(%d,%d,%d,%d)%1[\\}]%n", &x1, &y1, &x2, &y2, sep, &len) > 4 || |
| 433 | sscanf(buf, "\\move(%d,%d,%d,%d,%d,%d)%1[\\}]%n", &x1, &y1, &x2, &y2, &t1, &t2, sep, &len) > 6) { |
| 434 | if (callbacks->move) |
| 435 | callbacks->move(priv, x1, y1, x2, y2, t1, t2); |
| 436 | } else if (sscanf(buf, "\\pos(%d,%d)%1[\\}]%n", &x1, &y1, sep, &len) > 2) { |
| 437 | if (callbacks->move) |
| 438 | callbacks->move(priv, x1, y1, x1, y1, -1, -1); |
| 439 | } else if (sscanf(buf, "\\org(%d,%d)%1[\\}]%n", &x1, &y1, sep, &len) > 2) { |
| 440 | if (callbacks->origin) |
| 441 | callbacks->origin(priv, x1, y1); |
| 442 | } else { |
| 443 | len = strcspn(buf+1, "\\}") + 2; /* skip unknown code */ |
| 444 | } |
| 445 | buf += len - 1; |
| 446 | } |
| 447 | if (*buf++ != '}') |
| 448 | return AVERROR_INVALIDDATA; |
| 449 | } else { |
| 450 | if (!text) { |
| 451 | text = buf; |
| 452 | text_len = 1; |
| 453 | } else |
| 454 | text_len++; |
| 455 | buf++; |
| 456 | } |
| 457 | } |
| 458 | if (text && callbacks->text) |
| 459 | callbacks->text(priv, text, text_len); |
| 460 | if (callbacks->end) |
| 461 | callbacks->end(priv); |
| 462 | return 0; |
| 463 | } |
| 464 | |
| 465 | ASSStyle *ff_ass_style_get(ASSSplitContext *ctx, const char *style) |
| 466 | { |
| 467 | ASS *ass = &ctx->ass; |
| 468 | int i; |
| 469 | |
| 470 | if (!style || !*style) |
| 471 | style = "Default"; |
| 472 | for (i=0; i<ass->styles_count; i++) |
| 473 | if (!strcmp(ass->styles[i].name, style)) |
| 474 | return ass->styles + i; |
| 475 | return NULL; |
| 476 | } |