1 /* 2 ******************************************************************************* 3 * 4 * Copyright (C) 1999-2004, International Business Machines 5 * Corporation and others. All Rights Reserved. 6 * 7 ******************************************************************************* 8 * file name: utf16.h 9 * encoding: US-ASCII 10 * tab size: 8 (not used) 11 * indentation:4 12 * 13 * created on: 1999sep09 14 * created by: Markus W. Scherer 15 */ 16 17 /** 18 * \file 19 * \brief C API: 16-bit Unicode handling macros 20 * 21 * This file defines macros to deal with 16-bit Unicode (UTF-16) code units and strings. 22 * utf16.h is included by utf.h after unicode/umachine.h 23 * and some common definitions. 24 * 25 * For more information see utf.h and the ICU User Guide Strings chapter 26 * (http://oss.software.ibm.com/icu/userguide/). 27 * 28 * <em>Usage:</em> 29 * ICU coding guidelines for if() statements should be followed when using these macros. 30 * Compound statements (curly braces {}) must be used for if-else-while... 31 * bodies and all macro statements should be terminated with semicolon. 32 */ 33 34 #ifndef __UTF16_H__ 35 #define __UTF16_H__ 36 37 /* utf.h must be included first. */ 38 #ifndef __UTF_H__ 39 # include "unicode/utf.h" 40 #endif 41 42 /* single-code point definitions -------------------------------------------- */ 43 44 /** 45 * Does this code unit alone encode a code point (BMP, not a surrogate)? 46 * @param c 16-bit code unit 47 * @return TRUE or FALSE 48 * @stable ICU 2.4 49 */ 50 #define U16_IS_SINGLE(c) !U_IS_SURROGATE(c) 51 52 /** 53 * Is this code unit a lead surrogate (U+d800..U+dbff)? 54 * @param c 16-bit code unit 55 * @return TRUE or FALSE 56 * @stable ICU 2.4 57 */ 58 #define U16_IS_LEAD(c) (((c)&0xfffffc00)==0xd800) 59 60 /** 61 * Is this code unit a trail surrogate (U+dc00..U+dfff)? 62 * @param c 16-bit code unit 63 * @return TRUE or FALSE 64 * @stable ICU 2.4 65 */ 66 #define U16_IS_TRAIL(c) (((c)&0xfffffc00)==0xdc00) 67 68 /** 69 * Is this code unit a surrogate (U+d800..U+dfff)? 70 * @param c 16-bit code unit 71 * @return TRUE or FALSE 72 * @stable ICU 2.4 73 */ 74 #define U16_IS_SURROGATE(c) U_IS_SURROGATE(c) 75 76 /** 77 * Assuming c is a surrogate code point (U16_IS_SURROGATE(c)), 78 * is it a lead surrogate? 79 * @param c 16-bit code unit 80 * @return TRUE or FALSE 81 * @stable ICU 2.4 82 */ 83 #define U16_IS_SURROGATE_LEAD(c) (((c)&0x400)==0) 84 85 /** 86 * Helper constant for U16_GET_SUPPLEMENTARY. 87 * @internal 88 */ 89 #define U16_SURROGATE_OFFSET ((0xd800<<10UL)+0xdc00-0x10000) 90 91 /** 92 * Get a supplementary code point value (U+10000..U+10ffff) 93 * from its lead and trail surrogates. 94 * The result is undefined if the input values are not 95 * lead and trail surrogates. 96 * 97 * @param lead lead surrogate (U+d800..U+dbff) 98 * @param trail trail surrogate (U+dc00..U+dfff) 99 * @return supplementary code point (U+10000..U+10ffff) 100 * @stable ICU 2.4 101 */ 102 #define U16_GET_SUPPLEMENTARY(lead, trail) \ 103 (((UChar32)(lead)<<10UL)+(UChar32)(trail)-U16_SURROGATE_OFFSET) 104 105 106 /** 107 * Get the lead surrogate (0xd800..0xdbff) for a 108 * supplementary code point (0x10000..0x10ffff). 109 * @param supplementary 32-bit code point (U+10000..U+10ffff) 110 * @return lead surrogate (U+d800..U+dbff) for supplementary 111 * @stable ICU 2.4 112 */ 113 #define U16_LEAD(supplementary) (UChar)(((supplementary)>>10)+0xd7c0) 114 115 /** 116 * Get the trail surrogate (0xdc00..0xdfff) for a 117 * supplementary code point (0x10000..0x10ffff). 118 * @param supplementary 32-bit code point (U+10000..U+10ffff) 119 * @return trail surrogate (U+dc00..U+dfff) for supplementary 120 * @stable ICU 2.4 121 */ 122 #define U16_TRAIL(supplementary) (UChar)(((supplementary)&0x3ff)|0xdc00) 123 124 /** 125 * How many 16-bit code units are used to encode this Unicode code point? (1 or 2) 126 * The result is not defined if c is not a Unicode code point (U+0000..U+10ffff). 127 * @param c 32-bit code point 128 * @return 1 or 2 129 * @stable ICU 2.4 130 */ 131 #define U16_LENGTH(c) ((uint32_t)(c)<=0xffff ? 1 : 2) 132 133 /** 134 * The maximum number of 16-bit code units per Unicode code point (U+0000..U+10ffff). 135 * @return 2 136 * @stable ICU 2.4 137 */ 138 #define U16_MAX_LENGTH 2 139 140 /** 141 * Get a code point from a string at a random-access offset, 142 * without changing the offset. 143 * "Unsafe" macro, assumes well-formed UTF-16. 144 * 145 * The offset may point to either the lead or trail surrogate unit 146 * for a supplementary code point, in which case the macro will read 147 * the adjacent matching surrogate as well. 148 * The result is undefined if the offset points to a single, unpaired surrogate. 149 * Iteration through a string is more efficient with U16_NEXT_UNSAFE or U16_NEXT. 150 * 151 * @param s const UChar * string 152 * @param i string offset 153 * @param c output UChar32 variable 154 * @see U16_GET 155 * @stable ICU 2.4 156 */ 157 #define U16_GET_UNSAFE(s, i, c) { \ 158 (c)=(s)[i]; \ 159 if(U16_IS_SURROGATE(c)) { \ 160 if(U16_IS_SURROGATE_LEAD(c)) { \ 161 (c)=U16_GET_SUPPLEMENTARY((c), (s)[(i)+1]); \ 162 } else { \ 163 (c)=U16_GET_SUPPLEMENTARY((s)[(i)-1], (c)); \ 164 } \ 165 } \ 166 } 167 168 /** 169 * Get a code point from a string at a random-access offset, 170 * without changing the offset. 171 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 172 * 173 * The offset may point to either the lead or trail surrogate unit 174 * for a supplementary code point, in which case the macro will read 175 * the adjacent matching surrogate as well. 176 * If the offset points to a single, unpaired surrogate, then that itself 177 * will be returned as the code point. 178 * Iteration through a string is more efficient with U16_NEXT_UNSAFE or U16_NEXT. 179 * 180 * @param s const UChar * string 181 * @param start starting string offset (usually 0) 182 * @param i string offset, start<=i<length 183 * @param length string length 184 * @param c output UChar32 variable 185 * @see U16_GET_UNSAFE 186 * @stable ICU 2.4 187 */ 188 #define U16_GET(s, start, i, length, c) { \ 189 (c)=(s)[i]; \ 190 if(U16_IS_SURROGATE(c)) { \ 191 uint16_t __c2; \ 192 if(U16_IS_SURROGATE_LEAD(c)) { \ 193 if((i)+1<(length) && U16_IS_TRAIL(__c2=(s)[(i)+1])) { \ 194 (c)=U16_GET_SUPPLEMENTARY((c), __c2); \ 195 } \ 196 } else { \ 197 if((i)-1>=(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \ 198 (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \ 199 } \ 200 } \ 201 } \ 202 } 203 204 /* definitions with forward iteration --------------------------------------- */ 205 206 /** 207 * Get a code point from a string at a code point boundary offset, 208 * and advance the offset to the next code point boundary. 209 * (Post-incrementing forward iteration.) 210 * "Unsafe" macro, assumes well-formed UTF-16. 211 * 212 * The offset may point to the lead surrogate unit 213 * for a supplementary code point, in which case the macro will read 214 * the following trail surrogate as well. 215 * If the offset points to a trail surrogate, then that itself 216 * will be returned as the code point. 217 * The result is undefined if the offset points to a single, unpaired lead surrogate. 218 * 219 * @param s const UChar * string 220 * @param i string offset 221 * @param c output UChar32 variable 222 * @see U16_NEXT 223 * @stable ICU 2.4 224 */ 225 #define U16_NEXT_UNSAFE(s, i, c) { \ 226 (c)=(s)[(i)++]; \ 227 if(U16_IS_LEAD(c)) { \ 228 (c)=U16_GET_SUPPLEMENTARY((c), (s)[(i)++]); \ 229 } \ 230 } 231 232 /** 233 * Get a code point from a string at a code point boundary offset, 234 * and advance the offset to the next code point boundary. 235 * (Post-incrementing forward iteration.) 236 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 237 * 238 * The offset may point to the lead surrogate unit 239 * for a supplementary code point, in which case the macro will read 240 * the following trail surrogate as well. 241 * If the offset points to a trail surrogate or 242 * to a single, unpaired lead surrogate, then that itself 243 * will be returned as the code point. 244 * 245 * @param s const UChar * string 246 * @param i string offset, i<length 247 * @param length string length 248 * @param c output UChar32 variable 249 * @see U16_NEXT_UNSAFE 250 * @stable ICU 2.4 251 */ 252 #define U16_NEXT(s, i, length, c) { \ 253 (c)=(s)[(i)++]; \ 254 if(U16_IS_LEAD(c)) { \ 255 uint16_t __c2; \ 256 if((i)<(length) && U16_IS_TRAIL(__c2=(s)[(i)])) { \ 257 ++(i); \ 258 (c)=U16_GET_SUPPLEMENTARY((c), __c2); \ 259 } \ 260 } \ 261 } 262 263 /** 264 * Append a code point to a string, overwriting 1 or 2 code units. 265 * The offset points to the current end of the string contents 266 * and is advanced (post-increment). 267 * "Unsafe" macro, assumes a valid code point and sufficient space in the string. 268 * Otherwise, the result is undefined. 269 * 270 * @param s const UChar * string buffer 271 * @param i string offset 272 * @param c code point to append 273 * @see U16_APPEND 274 * @stable ICU 2.4 275 */ 276 #define U16_APPEND_UNSAFE(s, i, c) { \ 277 if((uint32_t)(c)<=0xffff) { \ 278 (s)[(i)++]=(uint16_t)(c); \ 279 } else { \ 280 (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \ 281 (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \ 282 } \ 283 } 284 285 /** 286 * Append a code point to a string, overwriting 1 or 2 code units. 287 * The offset points to the current end of the string contents 288 * and is advanced (post-increment). 289 * "Safe" macro, checks for a valid code point. 290 * If a surrogate pair is written, checks for sufficient space in the string. 291 * If the code point is not valid or a trail surrogate does not fit, 292 * then isError is set to TRUE. 293 * 294 * @param s const UChar * string buffer 295 * @param i string offset, i<length 296 * @param capacity size of the string buffer 297 * @param c code point to append 298 * @param isError output UBool set to TRUE if an error occurs, otherwise not modified 299 * @see U16_APPEND_UNSAFE 300 * @stable ICU 2.4 301 */ 302 #define U16_APPEND(s, i, capacity, c, isError) { \ 303 if((uint32_t)(c)<=0xffff) { \ 304 (s)[(i)++]=(uint16_t)(c); \ 305 } else if((uint32_t)(c)<=0x10ffff && (i)+1<(capacity)) { \ 306 (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \ 307 (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \ 308 } else /* c>0x10ffff or not enough space */ { \ 309 (isError)=TRUE; \ 310 } \ 311 } 312 313 /** 314 * Advance the string offset from one code point boundary to the next. 315 * (Post-incrementing iteration.) 316 * "Unsafe" macro, assumes well-formed UTF-16. 317 * 318 * @param s const UChar * string 319 * @param i string offset 320 * @see U16_FWD_1 321 * @stable ICU 2.4 322 */ 323 #define U16_FWD_1_UNSAFE(s, i) { \ 324 if(U16_IS_LEAD((s)[(i)++])) { \ 325 ++(i); \ 326 } \ 327 } 328 329 /** 330 * Advance the string offset from one code point boundary to the next. 331 * (Post-incrementing iteration.) 332 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 333 * 334 * @param s const UChar * string 335 * @param i string offset, i<length 336 * @param length string length 337 * @see U16_FWD_1_UNSAFE 338 * @stable ICU 2.4 339 */ 340 #define U16_FWD_1(s, i, length) { \ 341 if(U16_IS_LEAD((s)[(i)++]) && (i)<(length) && U16_IS_TRAIL((s)[i])) { \ 342 ++(i); \ 343 } \ 344 } 345 346 /** 347 * Advance the string offset from one code point boundary to the n-th next one, 348 * i.e., move forward by n code points. 349 * (Post-incrementing iteration.) 350 * "Unsafe" macro, assumes well-formed UTF-16. 351 * 352 * @param s const UChar * string 353 * @param i string offset 354 * @param n number of code points to skip 355 * @see U16_FWD_N 356 * @stable ICU 2.4 357 */ 358 #define U16_FWD_N_UNSAFE(s, i, n) { \ 359 int32_t __N=(n); \ 360 while(__N>0) { \ 361 U16_FWD_1_UNSAFE(s, i); \ 362 --__N; \ 363 } \ 364 } 365 366 /** 367 * Advance the string offset from one code point boundary to the n-th next one, 368 * i.e., move forward by n code points. 369 * (Post-incrementing iteration.) 370 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 371 * 372 * @param s const UChar * string 373 * @param i string offset, i<length 374 * @param length string length 375 * @param n number of code points to skip 376 * @see U16_FWD_N_UNSAFE 377 * @stable ICU 2.4 378 */ 379 #define U16_FWD_N(s, i, length, n) { \ 380 int32_t __N=(n); \ 381 while(__N>0 && (i)<(length)) { \ 382 U16_FWD_1(s, i, length); \ 383 --__N; \ 384 } \ 385 } 386 387 /** 388 * Adjust a random-access offset to a code point boundary 389 * at the start of a code point. 390 * If the offset points to the trail surrogate of a surrogate pair, 391 * then the offset is decremented. 392 * Otherwise, it is not modified. 393 * "Unsafe" macro, assumes well-formed UTF-16. 394 * 395 * @param s const UChar * string 396 * @param i string offset 397 * @see U16_SET_CP_START 398 * @stable ICU 2.4 399 */ 400 #define U16_SET_CP_START_UNSAFE(s, i) { \ 401 if(U16_IS_TRAIL((s)[i])) { \ 402 --(i); \ 403 } \ 404 } 405 406 /** 407 * Adjust a random-access offset to a code point boundary 408 * at the start of a code point. 409 * If the offset points to the trail surrogate of a surrogate pair, 410 * then the offset is decremented. 411 * Otherwise, it is not modified. 412 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 413 * 414 * @param s const UChar * string 415 * @param start starting string offset (usually 0) 416 * @param i string offset, start<=i 417 * @see U16_SET_CP_START_UNSAFE 418 * @stable ICU 2.4 419 */ 420 #define U16_SET_CP_START(s, start, i) { \ 421 if(U16_IS_TRAIL((s)[i]) && (i)>(start) && U16_IS_LEAD((s)[(i)-1])) { \ 422 --(i); \ 423 } \ 424 } 425 426 /* definitions with backward iteration -------------------------------------- */ 427 428 /** 429 * Move the string offset from one code point boundary to the previous one 430 * and get the code point between them. 431 * (Pre-decrementing backward iteration.) 432 * "Unsafe" macro, assumes well-formed UTF-16. 433 * 434 * The input offset may be the same as the string length. 435 * If the offset is behind a trail surrogate unit 436 * for a supplementary code point, then the macro will read 437 * the preceding lead surrogate as well. 438 * If the offset is behind a lead surrogate, then that itself 439 * will be returned as the code point. 440 * The result is undefined if the offset is behind a single, unpaired trail surrogate. 441 * 442 * @param s const UChar * string 443 * @param i string offset 444 * @param c output UChar32 variable 445 * @see U16_PREV 446 * @stable ICU 2.4 447 */ 448 #define U16_PREV_UNSAFE(s, i, c) { \ 449 (c)=(s)[--(i)]; \ 450 if(U16_IS_TRAIL(c)) { \ 451 (c)=U16_GET_SUPPLEMENTARY((s)[--(i)], (c)); \ 452 } \ 453 } 454 455 /** 456 * Move the string offset from one code point boundary to the previous one 457 * and get the code point between them. 458 * (Pre-decrementing backward iteration.) 459 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 460 * 461 * The input offset may be the same as the string length. 462 * If the offset is behind a trail surrogate unit 463 * for a supplementary code point, then the macro will read 464 * the preceding lead surrogate as well. 465 * If the offset is behind a lead surrogate or behind a single, unpaired 466 * trail surrogate, then that itself 467 * will be returned as the code point. 468 * 469 * @param s const UChar * string 470 * @param start starting string offset (usually 0) 471 * @param i string offset, start<=i 472 * @param c output UChar32 variable 473 * @see U16_PREV_UNSAFE 474 * @stable ICU 2.4 475 */ 476 #define U16_PREV(s, start, i, c) { \ 477 (c)=(s)[--(i)]; \ 478 if(U16_IS_TRAIL(c)) { \ 479 uint16_t __c2; \ 480 if((i)>(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \ 481 --(i); \ 482 (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \ 483 } \ 484 } \ 485 } 486 487 /** 488 * Move the string offset from one code point boundary to the previous one. 489 * (Pre-decrementing backward iteration.) 490 * The input offset may be the same as the string length. 491 * "Unsafe" macro, assumes well-formed UTF-16. 492 * 493 * @param s const UChar * string 494 * @param i string offset 495 * @see U16_BACK_1 496 * @stable ICU 2.4 497 */ 498 #define U16_BACK_1_UNSAFE(s, i) { \ 499 if(U16_IS_TRAIL((s)[--(i)])) { \ 500 --(i); \ 501 } \ 502 } 503 504 /** 505 * Move the string offset from one code point boundary to the previous one. 506 * (Pre-decrementing backward iteration.) 507 * The input offset may be the same as the string length. 508 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 509 * 510 * @param s const UChar * string 511 * @param start starting string offset (usually 0) 512 * @param i string offset, start<=i 513 * @see U16_BACK_1_UNSAFE 514 * @stable ICU 2.4 515 */ 516 #define U16_BACK_1(s, start, i) { \ 517 if(U16_IS_TRAIL((s)[--(i)]) && (i)>(start) && U16_IS_LEAD((s)[(i)-1])) { \ 518 --(i); \ 519 } \ 520 } 521 522 /** 523 * Move the string offset from one code point boundary to the n-th one before it, 524 * i.e., move backward by n code points. 525 * (Pre-decrementing backward iteration.) 526 * The input offset may be the same as the string length. 527 * "Unsafe" macro, assumes well-formed UTF-16. 528 * 529 * @param s const UChar * string 530 * @param i string offset 531 * @param n number of code points to skip 532 * @see U16_BACK_N 533 * @stable ICU 2.4 534 */ 535 #define U16_BACK_N_UNSAFE(s, i, n) { \ 536 int32_t __N=(n); \ 537 while(__N>0) { \ 538 U16_BACK_1_UNSAFE(s, i); \ 539 --__N; \ 540 } \ 541 } 542 543 /** 544 * Move the string offset from one code point boundary to the n-th one before it, 545 * i.e., move backward by n code points. 546 * (Pre-decrementing backward iteration.) 547 * The input offset may be the same as the string length. 548 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 549 * 550 * @param s const UChar * string 551 * @param start start of string 552 * @param i string offset, i<length 553 * @param n number of code points to skip 554 * @see U16_BACK_N_UNSAFE 555 * @stable ICU 2.4 556 */ 557 #define U16_BACK_N(s, start, i, n) { \ 558 int32_t __N=(n); \ 559 while(__N>0 && (i)>(start)) { \ 560 U16_BACK_1(s, start, i); \ 561 --__N; \ 562 } \ 563 } 564 565 /** 566 * Adjust a random-access offset to a code point boundary after a code point. 567 * If the offset is behind the lead surrogate of a surrogate pair, 568 * then the offset is incremented. 569 * Otherwise, it is not modified. 570 * The input offset may be the same as the string length. 571 * "Unsafe" macro, assumes well-formed UTF-16. 572 * 573 * @param s const UChar * string 574 * @param i string offset 575 * @see U16_SET_CP_LIMIT 576 * @stable ICU 2.4 577 */ 578 #define U16_SET_CP_LIMIT_UNSAFE(s, i) { \ 579 if(U16_IS_LEAD((s)[(i)-1])) { \ 580 ++(i); \ 581 } \ 582 } 583 584 /** 585 * Adjust a random-access offset to a code point boundary after a code point. 586 * If the offset is behind the lead surrogate of a surrogate pair, 587 * then the offset is incremented. 588 * Otherwise, it is not modified. 589 * The input offset may be the same as the string length. 590 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 591 * 592 * @param s const UChar * string 593 * @param start starting string offset (usually 0) 594 * @param i string offset, start<=i<=length 595 * @param length string length 596 * @see U16_SET_CP_LIMIT_UNSAFE 597 * @stable ICU 2.4 598 */ 599 #define U16_SET_CP_LIMIT(s, start, i, length) { \ 600 if((start)<(i) && (i)<(length) && U16_IS_LEAD((s)[(i)-1]) && U16_IS_TRAIL((s)[i])) { \ 601 ++(i); \ 602 } \ 603 } 604 605 #endif 606