The Android Open Source Project | 9066cfe | 2009-03-03 19:31:44 -0800 | [diff] [blame] | 1 | /* |
| 2 | * Copyright (C) 2007 The Android Open Source Project |
| 3 | * |
| 4 | * Licensed under the Apache License, Version 2.0 (the "License"); |
| 5 | * you may not use this file except in compliance with the License. |
| 6 | * You may obtain a copy of the License at |
| 7 | * |
| 8 | * http://www.apache.org/licenses/LICENSE-2.0 |
| 9 | * |
| 10 | * Unless required by applicable law or agreed to in writing, software |
| 11 | * distributed under the License is distributed on an "AS IS" BASIS, |
| 12 | * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| 13 | * See the License for the specific language governing permissions and |
| 14 | * limitations under the License. |
| 15 | */ |
| 16 | |
| 17 | #include <objmng/drm_i18n.h> |
| 18 | |
| 19 | #define IS_GB2312_HIGH_BYTE(c) ((c) >= 0xA1 && (c) <= 0xF7) |
| 20 | #define IS_GB2312_LOW_BYTE(c) ((c) >= 0xA1 && (c) <= 0xFE) |
| 21 | #define IS_GBK_HIGH_BYTE(c) ((c) >= 0x81 && (c) <= 0xFE) |
| 22 | #define IS_GBK_LOW_BYTE(c) ((c) >= 0x40 && (c) <= 0xFE && (c) != 0x7F) |
| 23 | #define IS_BIG5_HIGH_BYTE(c) ((c) >= 0xA1 && (c) <= 0xF9) |
| 24 | #define IS_BIG5_LOW_BYTE(c) (((c) >= 0x40 && (c) <= 0x7E) \ |
| 25 | || ((c) >= 0xA1 && (c) <= 0xFE)) |
| 26 | #define IS_ASCII(c) ((c) <= 127) |
| 27 | |
| 28 | #define INVALID_UNICODE 0xFFFD |
| 29 | |
| 30 | #define I18N_LATIN1_SUPPORT |
| 31 | #define I18N_UTF8_UTF16_SUPPORT |
| 32 | |
| 33 | |
| 34 | /** |
| 35 | * Simply convert ISO 8859-1 (latin1) to unicode |
| 36 | */ |
| 37 | static int32_t latin1ToWcs(const uint8_t *mbs, int32_t mbsLen, |
| 38 | uint16_t *wcsBuf, int32_t bufSizeInWideChar, |
| 39 | int32_t *bytesConsumed); |
| 40 | |
| 41 | /** |
| 42 | * Convert one unicode char to ISO 8859-1 (latin1) byte |
| 43 | */ |
| 44 | static int32_t wcToLatin1(uint16_t wc, uint8_t * mbs, int32_t bufSize); |
| 45 | |
| 46 | /** |
| 47 | * Convert UTF-8 to unicode |
| 48 | */ |
| 49 | static int32_t utf8ToWcs(const uint8_t *mbs, int32_t mbsLen, |
| 50 | uint16_t *wcsBuf, int32_t bufSizeInWideChar, |
| 51 | int32_t *bytesConsumed); |
| 52 | |
| 53 | /** |
| 54 | * Convert one unicode char to UTF-8 bytes |
| 55 | */ |
| 56 | static int32_t wcToUtf8(uint16_t wc, uint8_t * mbs, int32_t bufSize); |
| 57 | |
| 58 | /** |
| 59 | * Convert UTF-16 BE to unicode |
| 60 | */ |
| 61 | static int32_t utf16beToWcs(const uint8_t *mbs, int32_t mbsLen, |
| 62 | uint16_t *wcsBuf, int32_t bufSizeInWideChar, |
| 63 | int32_t *bytesConsumed); |
| 64 | |
| 65 | /** |
| 66 | * Convert one unicode char to UTF-16 BE bytes |
| 67 | */ |
| 68 | static int32_t wcToUtf16be(uint16_t wc, uint8_t * mbs, int32_t bufSize); |
| 69 | |
| 70 | /** |
| 71 | * Convert UTF-16 LE to unicode |
| 72 | */ |
| 73 | static int32_t utf16leToWcs(const uint8_t *mbs, int32_t mbsLen, |
| 74 | uint16_t *wcsBuf, int32_t bufSizeInWideChar, |
| 75 | int32_t *bytesConsumed); |
| 76 | |
| 77 | /** |
| 78 | * Convert one unicode char to UTF-16 LE bytes |
| 79 | */ |
| 80 | static int32_t wcToUtf16le(uint16_t wc, uint8_t * mbs, int32_t bufSize); |
| 81 | |
| 82 | /* |
| 83 | * see drm_i18n.h |
| 84 | */ |
| 85 | int32_t DRM_i18n_mbsToWcs(DRM_Charset_t charset, |
| 86 | const uint8_t *mbs, int32_t mbsLen, |
| 87 | uint16_t *wcsBuf, int32_t bufSizeInWideChar, |
| 88 | int32_t *bytesConsumed) |
| 89 | { |
| 90 | switch (charset) |
| 91 | { |
| 92 | #ifdef I18N_GB2312_SUPPORT |
| 93 | case DRM_CHARSET_GB2312: |
| 94 | return gb2312ToWcs(mbs, mbsLen, wcsBuf, bufSizeInWideChar, bytesConsumed); |
| 95 | #endif |
| 96 | #ifdef I18N_GBK_SUPPORT |
| 97 | case DRM_CHARSET_GBK: |
| 98 | return gbkToWcs(mbs, mbsLen, wcsBuf, bufSizeInWideChar, bytesConsumed); |
| 99 | #endif |
| 100 | #ifdef I18N_BIG5_SUPPORT |
| 101 | case DRM_CHARSET_BIG5: |
| 102 | return big5ToWcs(mbs, mbsLen, wcsBuf, bufSizeInWideChar, bytesConsumed); |
| 103 | #endif |
| 104 | #ifdef I18N_LATIN1_SUPPORT |
| 105 | case DRM_CHARSET_LATIN1: |
| 106 | return latin1ToWcs(mbs, mbsLen, wcsBuf, bufSizeInWideChar, bytesConsumed); |
| 107 | #endif |
| 108 | #ifdef I18N_ISO8859X_SUPPORT |
| 109 | case DRM_CHARSET_LATIN2: |
| 110 | case DRM_CHARSET_LATIN3: |
| 111 | case DRM_CHARSET_LATIN4: |
| 112 | case DRM_CHARSET_CYRILLIC: |
| 113 | case DRM_CHARSET_ARABIC: |
| 114 | case DRM_CHARSET_GREEK: |
| 115 | case DRM_CHARSET_HEBREW: |
| 116 | case DRM_CHARSET_LATIN5: |
| 117 | case DRM_CHARSET_LATIN6: |
| 118 | case DRM_CHARSET_THAI: |
| 119 | case DRM_CHARSET_LATIN7: |
| 120 | case DRM_CHARSET_LATIN8: |
| 121 | case DRM_CHARSET_LATIN9: |
| 122 | case DRM_CHARSET_LATIN10: |
| 123 | return iso8859xToWcs(charset, mbs, mbsLen, wcsBuf, bufSizeInWideChar, bytesConsumed); |
| 124 | #endif |
| 125 | #ifdef I18N_UTF8_UTF16_SUPPORT |
| 126 | case DRM_CHARSET_UTF8: |
| 127 | return utf8ToWcs(mbs, mbsLen, wcsBuf, bufSizeInWideChar, bytesConsumed); |
| 128 | case DRM_CHARSET_UTF16BE: |
| 129 | return utf16beToWcs(mbs, mbsLen, wcsBuf, bufSizeInWideChar, bytesConsumed); |
| 130 | case DRM_CHARSET_UTF16LE: |
| 131 | return utf16leToWcs(mbs, mbsLen, wcsBuf, bufSizeInWideChar, bytesConsumed); |
| 132 | #endif |
| 133 | default: |
| 134 | return -1; |
| 135 | } |
| 136 | } |
| 137 | |
| 138 | /* |
| 139 | * see drm_i18n.h |
| 140 | */ |
| 141 | int32_t DRM_i18n_wcsToMbs(DRM_Charset_t charset, |
| 142 | const uint16_t *wcs, int32_t wcsLen, |
| 143 | uint8_t *mbsBuf, int32_t bufSizeInByte) |
| 144 | { |
| 145 | int32_t (* wcToMbFunc)(uint16_t, uint8_t *, int32_t); |
| 146 | int32_t charIndex = 0; |
| 147 | int32_t numMultiBytes = 0; |
| 148 | |
| 149 | switch (charset) |
| 150 | { |
| 151 | #ifdef I18N_LATIN1_SUPPORT |
| 152 | case DRM_CHARSET_LATIN1: |
| 153 | wcToMbFunc = wcToLatin1; |
| 154 | break; |
| 155 | #endif |
| 156 | #ifdef I18N_UTF8_UTF16_SUPPORT |
| 157 | case DRM_CHARSET_UTF8: |
| 158 | wcToMbFunc = wcToUtf8; |
| 159 | break; |
| 160 | case DRM_CHARSET_UTF16BE: |
| 161 | wcToMbFunc = wcToUtf16be; |
| 162 | break; |
| 163 | case DRM_CHARSET_UTF16LE: |
| 164 | wcToMbFunc = wcToUtf16le; |
| 165 | break; |
| 166 | #endif |
| 167 | #ifdef I18N_ISO8859X_SUPPORT |
| 168 | case DRM_CHARSET_LATIN2: |
| 169 | case DRM_CHARSET_LATIN3: |
| 170 | case DRM_CHARSET_LATIN4: |
| 171 | case DRM_CHARSET_CYRILLIC: |
| 172 | case DRM_CHARSET_ARABIC: |
| 173 | case DRM_CHARSET_GREEK: |
| 174 | case DRM_CHARSET_HEBREW: |
| 175 | case DRM_CHARSET_LATIN5: |
| 176 | case DRM_CHARSET_LATIN6: |
| 177 | case DRM_CHARSET_THAI: |
| 178 | case DRM_CHARSET_LATIN7: |
| 179 | case DRM_CHARSET_LATIN8: |
| 180 | case DRM_CHARSET_LATIN9: |
| 181 | case DRM_CHARSET_LATIN10: |
| 182 | return wcsToIso8859x(charset, wcs, wcsLen, mbsBuf, bufSizeInByte); |
| 183 | #endif |
| 184 | default: |
| 185 | return -1; |
| 186 | } |
| 187 | |
| 188 | if (mbsBuf) { |
| 189 | while (numMultiBytes < bufSizeInByte && charIndex < wcsLen) { |
| 190 | /* TODO: handle surrogate pair values here */ |
| 191 | int32_t mbLen = wcToMbFunc(wcs[charIndex], |
| 192 | &mbsBuf[numMultiBytes], bufSizeInByte - numMultiBytes); |
| 193 | |
| 194 | if (numMultiBytes + mbLen > bufSizeInByte) { |
| 195 | /* Insufficient buffer. Don't update numMultiBytes */ |
| 196 | break; |
| 197 | } |
| 198 | charIndex++; |
| 199 | numMultiBytes += mbLen; |
| 200 | } |
| 201 | } else { |
| 202 | while (charIndex < wcsLen) { |
| 203 | /* TODO: handle surrogate pair values here */ |
| 204 | numMultiBytes += wcToMbFunc(wcs[charIndex], NULL, 0); |
| 205 | charIndex++; |
| 206 | } |
| 207 | } |
| 208 | |
| 209 | return numMultiBytes; |
| 210 | } |
| 211 | |
| 212 | |
| 213 | #ifdef I18N_LATIN1_SUPPORT |
| 214 | |
| 215 | int32_t latin1ToWcs(const uint8_t *mbs, int32_t mbsLen, |
| 216 | uint16_t *wcsBuf, int32_t bufSizeInWideChar, |
| 217 | int32_t *bytesConsumed) |
| 218 | { |
| 219 | int32_t charsToConvert; |
| 220 | int32_t len; |
| 221 | |
| 222 | if (wcsBuf == NULL) { |
| 223 | return mbsLen; |
| 224 | } |
| 225 | |
| 226 | len = charsToConvert = mbsLen > bufSizeInWideChar ? bufSizeInWideChar : mbsLen; |
| 227 | if (len < 0) |
| 228 | return 0; |
| 229 | while (len--) { |
| 230 | *wcsBuf++ = *mbs++; |
| 231 | } |
| 232 | |
| 233 | if (bytesConsumed) |
| 234 | *bytesConsumed = charsToConvert; |
| 235 | |
| 236 | return charsToConvert; |
| 237 | } |
| 238 | |
| 239 | int32_t wcToLatin1(uint16_t wc, uint8_t * mbs, int32_t bufSize) |
| 240 | { |
| 241 | uint8_t ch; |
| 242 | |
| 243 | if (wc < 0x100) { |
| 244 | ch = (uint8_t)(wc & 0xff); |
| 245 | } else { |
| 246 | ch = '?'; |
| 247 | } |
| 248 | if (mbs && bufSize > 0) |
| 249 | *mbs = ch; |
| 250 | return 1; |
| 251 | } |
| 252 | |
| 253 | #endif /* I18N_LATIN1_SUPPORT */ |
| 254 | |
| 255 | #ifdef I18N_UTF8_UTF16_SUPPORT |
| 256 | |
| 257 | int32_t utf8ToWcs(const uint8_t *mbs, int32_t mbsLen, |
| 258 | uint16_t *wcsBuf, int32_t bufSizeInWideChar, |
| 259 | int32_t *bytesConsumed) |
| 260 | { |
| 261 | int32_t charsConverted = 0; |
| 262 | int32_t i = 0; |
| 263 | int32_t wideChar; |
| 264 | |
| 265 | if (wcsBuf == NULL) { |
| 266 | /* No conversion but we're still going to calculate bytesConsumed */ |
| 267 | bufSizeInWideChar = mbsLen * 2; |
| 268 | } |
| 269 | |
| 270 | while((i < mbsLen) && (charsConverted < bufSizeInWideChar)) { |
| 271 | uint8_t ch = mbs[i]; |
| 272 | uint8_t ch2, ch3, ch4; |
| 273 | |
| 274 | wideChar = -1; |
| 275 | |
| 276 | if(IS_ASCII(ch)) { |
| 277 | wideChar = ch; |
| 278 | i++; |
| 279 | } else if ((ch & 0xc0) == 0xc0) { |
| 280 | int utfStart = i; |
| 281 | if ((ch & 0xe0) == 0xc0) { |
| 282 | /* 2 byte sequence */ |
| 283 | if (i + 1 < mbsLen && ((ch2 = mbs[i + 1]) & 0xc0) == 0x80) { |
| 284 | wideChar = (uint16_t)(((ch & 0x1F) << 6) | (ch2 & 0x3F)); |
| 285 | i += 2; |
| 286 | } else { |
| 287 | /* skip incomplete sequence */ |
| 288 | i++; |
| 289 | } |
| 290 | } else if ((ch & 0xf0) == 0xe0) { |
| 291 | /* 3 byte sequence */ |
| 292 | if (i + 2 < mbsLen |
| 293 | && ((ch2 = mbs[i + 1]) & 0xc0) == 0x80 |
| 294 | && ((ch3 = mbs[i + 2]) & 0xc0) == 0x80) { |
| 295 | wideChar = (uint16_t)(((ch & 0x0F) << 12) | ((ch2 & 0x3F) << 6) | (ch3 & 0x3F)); |
| 296 | i += 3; |
| 297 | } else { |
| 298 | /* skip incomplete sequence (up to 2 bytes) */ |
| 299 | i++; |
| 300 | if (i < mbsLen && (mbs[i] & 0xc0) == 0x80) |
| 301 | i++; |
| 302 | } |
| 303 | } else if ((ch & 0xf8) == 0xf0) { |
| 304 | /* 4 byte sequence */ |
| 305 | if (i + 3 < mbsLen |
| 306 | && ((ch2 = mbs[i + 1]) & 0xc0) == 0x80 |
| 307 | && ((ch3 = mbs[i + 2]) & 0xc0) == 0x80 |
| 308 | && ((ch4 = mbs[i + 3]) & 0xc0) == 0x80) { |
| 309 | /* FIXME: we do NOT support U+10000 - U+10FFFF for now. |
| 310 | * leave it as 0xFFFD. */ |
| 311 | wideChar = INVALID_UNICODE; |
| 312 | i += 4; |
| 313 | } else { |
| 314 | /* skip incomplete sequence (up to 3 bytes) */ |
| 315 | i++; |
| 316 | if (i < mbsLen && (mbs[i] & 0xc0) == 0x80) { |
| 317 | i++; |
| 318 | if (i < mbsLen && (mbs[i] & 0xc0) == 0x80) { |
| 319 | i++; |
| 320 | } |
| 321 | } |
| 322 | } |
| 323 | } else { |
| 324 | /* invalid */ |
| 325 | i++; |
| 326 | } |
| 327 | if (i >= mbsLen && wideChar == -1) { |
| 328 | /* Possible incomplete UTF-8 sequence at the end of mbs. |
| 329 | * Leave it to the caller. |
| 330 | */ |
| 331 | i = utfStart; |
| 332 | break; |
| 333 | } |
| 334 | } else { |
| 335 | /* invalid */ |
| 336 | i++; |
| 337 | } |
| 338 | if(wcsBuf) { |
| 339 | if (wideChar == -1) |
| 340 | wideChar = INVALID_UNICODE; |
| 341 | wcsBuf[charsConverted] = (uint16_t)wideChar; |
| 342 | } |
| 343 | charsConverted++; |
| 344 | } |
| 345 | |
| 346 | if (bytesConsumed) |
| 347 | *bytesConsumed = i; |
| 348 | |
| 349 | return charsConverted; |
| 350 | } |
| 351 | |
| 352 | int32_t wcToUtf8(uint16_t wc, uint8_t * mbs, int32_t bufSize) |
| 353 | { |
| 354 | if (wc <= 0x7f) { |
| 355 | if (mbs && (bufSize >= 1)) { |
| 356 | *mbs = (uint8_t)wc; |
| 357 | } |
| 358 | return 1; |
| 359 | } else if (wc <= 0x7ff) { |
| 360 | if (mbs && (bufSize >= 2)) { |
| 361 | *mbs++ = (uint8_t)((wc >> 6) | 0xc0); |
| 362 | *mbs = (uint8_t)((wc & 0x3f) | 0x80); |
| 363 | } |
| 364 | return 2; |
| 365 | } else { |
| 366 | if (mbs && (bufSize >= 3)) { |
| 367 | *mbs++ = (uint8_t)((wc >> 12) | 0xe0); |
| 368 | *mbs++ = (uint8_t)(((wc >> 6) & 0x3f)| 0x80); |
| 369 | *mbs = (uint8_t)((wc & 0x3f) | 0x80); |
| 370 | } |
| 371 | return 3; |
| 372 | } |
| 373 | } |
| 374 | |
| 375 | int32_t utf16beToWcs(const uint8_t *mbs, int32_t mbsLen, |
| 376 | uint16_t *wcsBuf, int32_t bufSizeInWideChar, |
| 377 | int32_t *bytesConsumed) |
| 378 | { |
| 379 | int32_t charsToConvert; |
| 380 | int32_t len; |
| 381 | |
| 382 | if (wcsBuf == NULL) { |
| 383 | return mbsLen / 2; |
| 384 | } |
| 385 | |
| 386 | len = charsToConvert = (mbsLen / 2) > bufSizeInWideChar ? bufSizeInWideChar : (mbsLen / 2); |
| 387 | while (len--) { |
| 388 | /* TODO: handle surrogate pair values */ |
| 389 | *wcsBuf++ = (uint16_t)((*mbs << 8) | *(mbs + 1)); |
| 390 | mbs += 2; |
| 391 | } |
| 392 | |
| 393 | if (bytesConsumed) |
| 394 | *bytesConsumed = charsToConvert * 2; |
| 395 | |
| 396 | return charsToConvert; |
| 397 | } |
| 398 | |
| 399 | int32_t wcToUtf16be(uint16_t wc, uint8_t * mbs, int32_t bufSize) |
| 400 | { |
| 401 | if (mbs && bufSize >= 2) { |
| 402 | /* TODO: handle surrogate pair values */ |
| 403 | *mbs = (uint8_t)(wc >> 8); |
| 404 | *(mbs + 1) = (uint8_t)(wc & 0xff); |
| 405 | } |
| 406 | return 2; |
| 407 | } |
| 408 | |
| 409 | int32_t utf16leToWcs(const uint8_t *mbs, int32_t mbsLen, |
| 410 | uint16_t *wcsBuf, int32_t bufSizeInWideChar, |
| 411 | int32_t *bytesConsumed) |
| 412 | { |
| 413 | int32_t charsToConvert; |
| 414 | int32_t len; |
| 415 | |
| 416 | if (wcsBuf == NULL) { |
| 417 | return mbsLen / 2; |
| 418 | } |
| 419 | |
| 420 | len = charsToConvert = (mbsLen / 2) > bufSizeInWideChar ? bufSizeInWideChar : (mbsLen / 2); |
| 421 | while (len--) { |
| 422 | /* TODO: handle surrogate pair values */ |
| 423 | *wcsBuf++ = (uint16_t)(*mbs | (*(mbs + 1) << 8)); |
| 424 | mbs += 2; |
| 425 | } |
| 426 | |
| 427 | if (bytesConsumed) |
| 428 | *bytesConsumed = charsToConvert * 2; |
| 429 | |
| 430 | return charsToConvert; |
| 431 | } |
| 432 | |
| 433 | int32_t wcToUtf16le(uint16_t wc, uint8_t * mbs, int32_t bufSize) |
| 434 | { |
| 435 | if (mbs && bufSize >= 2) { |
| 436 | /* TODO: handle surrogate pair values */ |
| 437 | *mbs = (uint8_t)(wc & 0xff); |
| 438 | *(mbs + 1) = (uint8_t)(wc >> 8); |
| 439 | } |
| 440 | return 2; |
| 441 | } |
| 442 | |
| 443 | #endif /* I18N_UTF8_UTF16_SUPPORT */ |
| 444 | |