| 1 | /* Copyright (C) 2010-2020 The RetroArch team |
| 2 | * |
| 3 | * --------------------------------------------------------------------------------------- |
| 4 | * The following license statement only applies to this file (encoding_utf.c). |
| 5 | * --------------------------------------------------------------------------------------- |
| 6 | * |
| 7 | * Permission is hereby granted, free of charge, |
| 8 | * to any person obtaining a copy of this software and associated documentation files (the "Software"), |
| 9 | * to deal in the Software without restriction, including without limitation the rights to |
| 10 | * use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, |
| 11 | * and to permit persons to whom the Software is furnished to do so, subject to the following conditions: |
| 12 | * |
| 13 | * The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. |
| 14 | * |
| 15 | * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, |
| 16 | * INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, |
| 17 | * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. |
| 18 | * IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, |
| 19 | * WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, |
| 20 | * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. |
| 21 | */ |
| 22 | |
| 23 | #include <stdint.h> |
| 24 | #include <stdlib.h> |
| 25 | #include <stddef.h> |
| 26 | #include <string.h> |
| 27 | |
| 28 | #include <boolean.h> |
| 29 | #include <compat/strl.h> |
| 30 | #include <retro_inline.h> |
| 31 | |
| 32 | #include <encodings/utf.h> |
| 33 | |
| 34 | #if defined(_WIN32) && !defined(_XBOX) |
| 35 | #include <windows.h> |
| 36 | #elif defined(_XBOX) |
| 37 | #include <xtl.h> |
| 38 | #endif |
| 39 | |
| 40 | #define UTF8_WALKBYTE(string) (*((*(string))++)) |
| 41 | |
| 42 | static unsigned leading_ones(uint8_t c) |
| 43 | { |
| 44 | unsigned ones = 0; |
| 45 | while (c & 0x80) |
| 46 | { |
| 47 | ones++; |
| 48 | c <<= 1; |
| 49 | } |
| 50 | |
| 51 | return ones; |
| 52 | } |
| 53 | |
| 54 | /** |
| 55 | * utf8_conv_utf32: |
| 56 | * |
| 57 | * Simple implementation. Assumes the sequence is |
| 58 | * properly synchronized and terminated. |
| 59 | **/ |
| 60 | size_t utf8_conv_utf32(uint32_t *out, size_t out_chars, |
| 61 | const char *in, size_t in_size) |
| 62 | { |
| 63 | unsigned i; |
| 64 | size_t ret = 0; |
| 65 | while (in_size && out_chars) |
| 66 | { |
| 67 | unsigned extra, shift; |
| 68 | uint32_t c; |
| 69 | uint8_t first = *in++; |
| 70 | unsigned ones = leading_ones(first); |
| 71 | |
| 72 | if (ones > 6 || ones == 1) /* Invalid or desync. */ |
| 73 | break; |
| 74 | |
| 75 | extra = ones ? ones - 1 : ones; |
| 76 | if (1 + extra > in_size) /* Overflow. */ |
| 77 | break; |
| 78 | |
| 79 | shift = (extra - 1) * 6; |
| 80 | c = (first & ((1 << (7 - ones)) - 1)) << (6 * extra); |
| 81 | |
| 82 | for (i = 0; i < extra; i++, in++, shift -= 6) |
| 83 | c |= (*in & 0x3f) << shift; |
| 84 | |
| 85 | *out++ = c; |
| 86 | in_size -= 1 + extra; |
| 87 | out_chars--; |
| 88 | ret++; |
| 89 | } |
| 90 | |
| 91 | return ret; |
| 92 | } |
| 93 | |
| 94 | /** |
| 95 | * utf16_conv_utf8: |
| 96 | * |
| 97 | * Leaf function. |
| 98 | **/ |
| 99 | bool utf16_conv_utf8(uint8_t *out, size_t *out_chars, |
| 100 | const uint16_t *in, size_t in_size) |
| 101 | { |
| 102 | size_t out_pos = 0; |
| 103 | size_t in_pos = 0; |
| 104 | static const |
| 105 | uint8_t utf8_limits[5] = { 0xC0, 0xE0, 0xF0, 0xF8, 0xFC }; |
| 106 | |
| 107 | for (;;) |
| 108 | { |
| 109 | unsigned num_adds; |
| 110 | uint32_t value; |
| 111 | |
| 112 | if (in_pos == in_size) |
| 113 | { |
| 114 | *out_chars = out_pos; |
| 115 | return true; |
| 116 | } |
| 117 | value = in[in_pos++]; |
| 118 | if (value < 0x80) |
| 119 | { |
| 120 | if (out) |
| 121 | out[out_pos] = (char)value; |
| 122 | out_pos++; |
| 123 | continue; |
| 124 | } |
| 125 | |
| 126 | if (value >= 0xD800 && value < 0xE000) |
| 127 | { |
| 128 | uint32_t c2; |
| 129 | |
| 130 | if (value >= 0xDC00 || in_pos == in_size) |
| 131 | break; |
| 132 | c2 = in[in_pos++]; |
| 133 | if (c2 < 0xDC00 || c2 >= 0xE000) |
| 134 | break; |
| 135 | value = (((value - 0xD800) << 10) | (c2 - 0xDC00)) + 0x10000; |
| 136 | } |
| 137 | |
| 138 | for (num_adds = 1; num_adds < 5; num_adds++) |
| 139 | if (value < (((uint32_t)1) << (num_adds * 5 + 6))) |
| 140 | break; |
| 141 | if (out) |
| 142 | out[out_pos] = (char)(utf8_limits[num_adds - 1] |
| 143 | + (value >> (6 * num_adds))); |
| 144 | out_pos++; |
| 145 | do |
| 146 | { |
| 147 | num_adds--; |
| 148 | if (out) |
| 149 | out[out_pos] = (char)(0x80 |
| 150 | + ((value >> (6 * num_adds)) & 0x3F)); |
| 151 | out_pos++; |
| 152 | }while (num_adds != 0); |
| 153 | } |
| 154 | |
| 155 | *out_chars = out_pos; |
| 156 | return false; |
| 157 | } |
| 158 | |
| 159 | /** |
| 160 | * utf8cpy: |
| 161 | * |
| 162 | * Acts mostly like strlcpy. |
| 163 | * |
| 164 | * Copies the given number of UTF-8 characters, |
| 165 | * but at most @d_len bytes. |
| 166 | * |
| 167 | * Always NULL terminates. Does not copy half a character. |
| 168 | * @s is assumed valid UTF-8. |
| 169 | * Use only if @chars is considerably less than @d_len. |
| 170 | * |
| 171 | * @return Number of bytes. |
| 172 | **/ |
| 173 | size_t utf8cpy(char *d, size_t d_len, const char *s, size_t chars) |
| 174 | { |
| 175 | const uint8_t *sb = (const uint8_t*)s; |
| 176 | const uint8_t *sb_org = sb; |
| 177 | |
| 178 | if (!s) |
| 179 | return 0; |
| 180 | |
| 181 | while (*sb && chars-- > 0) |
| 182 | { |
| 183 | sb++; |
| 184 | while ((*sb & 0xC0) == 0x80) |
| 185 | sb++; |
| 186 | } |
| 187 | |
| 188 | if ((size_t)(sb - sb_org) > d_len-1 /* NUL */) |
| 189 | { |
| 190 | sb = sb_org + d_len-1; |
| 191 | while ((*sb & 0xC0) == 0x80) |
| 192 | sb--; |
| 193 | } |
| 194 | |
| 195 | memcpy(d, sb_org, sb-sb_org); |
| 196 | d[sb-sb_org] = '\0'; |
| 197 | |
| 198 | return sb-sb_org; |
| 199 | } |
| 200 | |
| 201 | /** |
| 202 | * utf8skip: |
| 203 | * |
| 204 | * Leaf function |
| 205 | **/ |
| 206 | const char *utf8skip(const char *str, size_t chars) |
| 207 | { |
| 208 | const uint8_t *strb = (const uint8_t*)str; |
| 209 | |
| 210 | if (!chars) |
| 211 | return str; |
| 212 | |
| 213 | do |
| 214 | { |
| 215 | strb++; |
| 216 | while ((*strb & 0xC0)==0x80) |
| 217 | strb++; |
| 218 | chars--; |
| 219 | }while (chars); |
| 220 | |
| 221 | return (const char*)strb; |
| 222 | } |
| 223 | |
| 224 | /** |
| 225 | * utf8len: |
| 226 | * |
| 227 | * Leaf function. |
| 228 | **/ |
| 229 | size_t utf8len(const char *string) |
| 230 | { |
| 231 | size_t ret = 0; |
| 232 | |
| 233 | if (!string) |
| 234 | return 0; |
| 235 | |
| 236 | while (*string) |
| 237 | { |
| 238 | if ((*string & 0xC0) != 0x80) |
| 239 | ret++; |
| 240 | string++; |
| 241 | } |
| 242 | return ret; |
| 243 | } |
| 244 | |
| 245 | /** |
| 246 | * utf8_walk: |
| 247 | * |
| 248 | * Does not validate the input. |
| 249 | * |
| 250 | * Leaf function. |
| 251 | * |
| 252 | * @return Returns garbage if it's not UTF-8. |
| 253 | **/ |
| 254 | uint32_t utf8_walk(const char **string) |
| 255 | { |
| 256 | uint8_t first = UTF8_WALKBYTE(string); |
| 257 | uint32_t ret = 0; |
| 258 | |
| 259 | if (first < 128) |
| 260 | return first; |
| 261 | |
| 262 | ret = (ret << 6) | (UTF8_WALKBYTE(string) & 0x3F); |
| 263 | if (first >= 0xE0) |
| 264 | { |
| 265 | ret = (ret << 6) | (UTF8_WALKBYTE(string) & 0x3F); |
| 266 | if (first >= 0xF0) |
| 267 | { |
| 268 | ret = (ret << 6) | (UTF8_WALKBYTE(string) & 0x3F); |
| 269 | return ret | (first & 7) << 18; |
| 270 | } |
| 271 | return ret | (first & 15) << 12; |
| 272 | } |
| 273 | |
| 274 | return ret | (first & 31) << 6; |
| 275 | } |
| 276 | |
| 277 | static bool utf16_to_char(uint8_t **utf_data, |
| 278 | size_t *dest_len, const uint16_t *in) |
| 279 | { |
| 280 | unsigned len = 0; |
| 281 | while (in[len] != '\0') |
| 282 | len++; |
| 283 | utf16_conv_utf8(NULL, dest_len, in, len); |
| 284 | *dest_len += 1; |
| 285 | if ((*utf_data = (uint8_t*)malloc(*dest_len)) != 0) |
| 286 | return utf16_conv_utf8(*utf_data, dest_len, in, len); |
| 287 | return false; |
| 288 | } |
| 289 | |
| 290 | /** |
| 291 | * utf16_to_char_string: |
| 292 | **/ |
| 293 | bool utf16_to_char_string(const uint16_t *in, char *s, size_t len) |
| 294 | { |
| 295 | size_t dest_len = 0; |
| 296 | uint8_t *utf16_data = NULL; |
| 297 | bool ret = utf16_to_char(&utf16_data, &dest_len, in); |
| 298 | |
| 299 | if (ret) |
| 300 | { |
| 301 | utf16_data[dest_len] = 0; |
| 302 | strlcpy(s, (const char*)utf16_data, len); |
| 303 | } |
| 304 | |
| 305 | free(utf16_data); |
| 306 | utf16_data = NULL; |
| 307 | |
| 308 | return ret; |
| 309 | } |
| 310 | |
| 311 | #if defined(_WIN32) && !defined(_XBOX) && !defined(UNICODE) |
| 312 | /** |
| 313 | * mb_to_mb_string_alloc: |
| 314 | * |
| 315 | * @return Returned pointer MUST be freed by the caller if non-NULL. |
| 316 | **/ |
| 317 | static char *mb_to_mb_string_alloc(const char *str, |
| 318 | enum CodePage cp_in, enum CodePage cp_out) |
| 319 | { |
| 320 | wchar_t *path_buf_wide = NULL; |
| 321 | int path_buf_wide_len = MultiByteToWideChar(cp_in, 0, str, -1, NULL, 0); |
| 322 | |
| 323 | /* Windows 95 will return 0 from these functions with |
| 324 | * a UTF8 codepage set without MSLU. |
| 325 | * |
| 326 | * From an unknown MSDN version (others omit this info): |
| 327 | * - CP_UTF8 Windows 98/Me, Windows NT 4.0 and later: |
| 328 | * Translate using UTF-8. When this is set, dwFlags must be zero. |
| 329 | * - Windows 95: Under the Microsoft Layer for Unicode, |
| 330 | * MultiByteToWideChar also supports CP_UTF7 and CP_UTF8. |
| 331 | */ |
| 332 | |
| 333 | if (!path_buf_wide_len) |
| 334 | return strdup(str); |
| 335 | |
| 336 | if ((path_buf_wide = (wchar_t*) |
| 337 | calloc(path_buf_wide_len + sizeof(wchar_t), sizeof(wchar_t)))) |
| 338 | { |
| 339 | MultiByteToWideChar(cp_in, 0, |
| 340 | str, -1, path_buf_wide, path_buf_wide_len); |
| 341 | |
| 342 | if (*path_buf_wide) |
| 343 | { |
| 344 | int path_buf_len = WideCharToMultiByte(cp_out, 0, |
| 345 | path_buf_wide, -1, NULL, 0, NULL, NULL); |
| 346 | |
| 347 | if (path_buf_len) |
| 348 | { |
| 349 | char *path_buf = (char*) |
| 350 | calloc(path_buf_len + sizeof(char), sizeof(char)); |
| 351 | |
| 352 | if (path_buf) |
| 353 | { |
| 354 | WideCharToMultiByte(cp_out, 0, |
| 355 | path_buf_wide, -1, path_buf, |
| 356 | path_buf_len, NULL, NULL); |
| 357 | |
| 358 | free(path_buf_wide); |
| 359 | |
| 360 | if (*path_buf) |
| 361 | return path_buf; |
| 362 | |
| 363 | free(path_buf); |
| 364 | return NULL; |
| 365 | } |
| 366 | } |
| 367 | else |
| 368 | { |
| 369 | free(path_buf_wide); |
| 370 | return strdup(str); |
| 371 | } |
| 372 | } |
| 373 | |
| 374 | free(path_buf_wide); |
| 375 | } |
| 376 | |
| 377 | return NULL; |
| 378 | } |
| 379 | #endif |
| 380 | |
| 381 | /** |
| 382 | * utf8_to_local_string_alloc: |
| 383 | * |
| 384 | * @return Returned pointer MUST be freed by the caller if non-NULL. |
| 385 | **/ |
| 386 | char* utf8_to_local_string_alloc(const char *str) |
| 387 | { |
| 388 | if (str && *str) |
| 389 | #if defined(_WIN32) && !defined(_XBOX) && !defined(UNICODE) |
| 390 | return mb_to_mb_string_alloc(str, CODEPAGE_UTF8, CODEPAGE_LOCAL); |
| 391 | #else |
| 392 | return strdup(str); /* Assume string needs no modification if not on Windows */ |
| 393 | #endif |
| 394 | return NULL; |
| 395 | } |
| 396 | |
| 397 | /** |
| 398 | * local_to_utf8_string_alloc: |
| 399 | * |
| 400 | * @return Returned pointer MUST be freed by the caller if non-NULL. |
| 401 | **/ |
| 402 | char *local_to_utf8_string_alloc(const char *str) |
| 403 | { |
| 404 | if (str && *str) |
| 405 | #if defined(_WIN32) && !defined(_XBOX) && !defined(UNICODE) |
| 406 | return mb_to_mb_string_alloc(str, CODEPAGE_LOCAL, CODEPAGE_UTF8); |
| 407 | #else |
| 408 | return strdup(str); /* Assume string needs no modification if not on Windows */ |
| 409 | #endif |
| 410 | return NULL; |
| 411 | } |
| 412 | |
| 413 | /** |
| 414 | * utf8_to_utf16_string_alloc: |
| 415 | * |
| 416 | * @return Returned pointer MUST be freed by the caller if non-NULL. |
| 417 | **/ |
| 418 | wchar_t* utf8_to_utf16_string_alloc(const char *str) |
| 419 | { |
| 420 | #ifdef _WIN32 |
| 421 | int len = 0; |
| 422 | #else |
| 423 | size_t len = 0; |
| 424 | #endif |
| 425 | wchar_t *buf = NULL; |
| 426 | |
| 427 | if (!str || !*str) |
| 428 | return NULL; |
| 429 | |
| 430 | #ifdef _WIN32 |
| 431 | if ((len = MultiByteToWideChar(CP_UTF8, 0, str, -1, NULL, 0))) |
| 432 | { |
| 433 | if (!(buf = (wchar_t*)calloc(len, sizeof(wchar_t)))) |
| 434 | return NULL; |
| 435 | |
| 436 | if ((MultiByteToWideChar(CP_UTF8, 0, str, -1, buf, len)) < 0) |
| 437 | { |
| 438 | free(buf); |
| 439 | return NULL; |
| 440 | } |
| 441 | } |
| 442 | else |
| 443 | { |
| 444 | /* Fallback to ANSI codepage instead */ |
| 445 | if ((len = MultiByteToWideChar(CP_ACP, 0, str, -1, NULL, 0))) |
| 446 | { |
| 447 | if (!(buf = (wchar_t*)calloc(len, sizeof(wchar_t)))) |
| 448 | return NULL; |
| 449 | |
| 450 | if ((MultiByteToWideChar(CP_ACP, 0, str, -1, buf, len)) < 0) |
| 451 | { |
| 452 | free(buf); |
| 453 | return NULL; |
| 454 | } |
| 455 | } |
| 456 | } |
| 457 | #else |
| 458 | /* NOTE: For now, assume non-Windows platforms' locale is already UTF-8. */ |
| 459 | if ((len = mbstowcs(NULL, str, 0) + 1)) |
| 460 | { |
| 461 | if (!(buf = (wchar_t*)calloc(len, sizeof(wchar_t)))) |
| 462 | return NULL; |
| 463 | |
| 464 | if ((mbstowcs(buf, str, len)) == (size_t)-1) |
| 465 | { |
| 466 | free(buf); |
| 467 | return NULL; |
| 468 | } |
| 469 | } |
| 470 | #endif |
| 471 | |
| 472 | return buf; |
| 473 | } |
| 474 | |
| 475 | /** |
| 476 | * utf16_to_utf8_string_alloc: |
| 477 | * |
| 478 | * @return Returned pointer MUST be freed by the caller if non-NULL. |
| 479 | **/ |
| 480 | char* utf16_to_utf8_string_alloc(const wchar_t *str) |
| 481 | { |
| 482 | #ifdef _WIN32 |
| 483 | int len = 0; |
| 484 | #else |
| 485 | size_t len = 0; |
| 486 | #endif |
| 487 | char *buf = NULL; |
| 488 | |
| 489 | if (!str || !*str) |
| 490 | return NULL; |
| 491 | |
| 492 | #ifdef _WIN32 |
| 493 | { |
| 494 | UINT code_page = CP_UTF8; |
| 495 | |
| 496 | /* fallback to ANSI codepage instead */ |
| 497 | if (!(len = WideCharToMultiByte(code_page, |
| 498 | 0, str, -1, NULL, 0, NULL, NULL))) |
| 499 | { |
| 500 | code_page = CP_ACP; |
| 501 | len = WideCharToMultiByte(code_page, |
| 502 | 0, str, -1, NULL, 0, NULL, NULL); |
| 503 | } |
| 504 | |
| 505 | if (!(buf = (char*)calloc(len, sizeof(char)))) |
| 506 | return NULL; |
| 507 | |
| 508 | if (WideCharToMultiByte(code_page, |
| 509 | 0, str, -1, buf, len, NULL, NULL) < 0) |
| 510 | { |
| 511 | free(buf); |
| 512 | return NULL; |
| 513 | } |
| 514 | } |
| 515 | #else |
| 516 | /* NOTE: For now, assume non-Windows platforms' |
| 517 | * locale is already UTF-8. */ |
| 518 | if ((len = wcstombs(NULL, str, 0) + 1)) |
| 519 | { |
| 520 | if (!(buf = (char*)calloc(len, sizeof(char)))) |
| 521 | return NULL; |
| 522 | |
| 523 | if (wcstombs(buf, str, len) == (size_t)-1) |
| 524 | { |
| 525 | free(buf); |
| 526 | return NULL; |
| 527 | } |
| 528 | } |
| 529 | #endif |
| 530 | |
| 531 | return buf; |
| 532 | } |