1 // © 2016 and later: Unicode, Inc. and others. 2 // License & terms of use: http://www.unicode.org/copyright.html 3 /* 4 ******************************************************************************* 5 * 6 * Copyright (C) 1999-2012, International Business Machines 7 * Corporation and others. All Rights Reserved. 8 * 9 ******************************************************************************* 10 * file name: utf16.h 11 * encoding: UTF-8 12 * tab size: 8 (not used) 13 * indentation:4 14 * 15 * created on: 1999sep09 16 * created by: Markus W. Scherer 17 */ 18 19 /** 20 * \file 21 * \brief C API: 16-bit Unicode handling macros 22 * 23 * This file defines macros to deal with 16-bit Unicode (UTF-16) code units and strings. 24 * 25 * For more information see utf.h and the ICU User Guide Strings chapter 26 * (http://userguide.icu-project.org/strings). 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 #include "unicode/umachine.h" 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 * Assuming c is a surrogate code point (U16_IS_SURROGATE(c)), 87 * is it a trail surrogate? 88 * @param c 16-bit code unit 89 * @return TRUE or FALSE 90 * @stable ICU 4.2 91 */ 92 #define U16_IS_SURROGATE_TRAIL(c) (((c)&0x400)!=0) 93 94 /** 95 * Helper constant for U16_GET_SUPPLEMENTARY. 96 * @internal 97 */ 98 #define U16_SURROGATE_OFFSET ((0xd800<<10UL)+0xdc00-0x10000) 99 100 /** 101 * Get a supplementary code point value (U+10000..U+10ffff) 102 * from its lead and trail surrogates. 103 * The result is undefined if the input values are not 104 * lead and trail surrogates. 105 * 106 * @param lead lead surrogate (U+d800..U+dbff) 107 * @param trail trail surrogate (U+dc00..U+dfff) 108 * @return supplementary code point (U+10000..U+10ffff) 109 * @stable ICU 2.4 110 */ 111 #define U16_GET_SUPPLEMENTARY(lead, trail) \ 112 (((UChar32)(lead)<<10UL)+(UChar32)(trail)-U16_SURROGATE_OFFSET) 113 114 115 /** 116 * Get the lead surrogate (0xd800..0xdbff) for a 117 * supplementary code point (0x10000..0x10ffff). 118 * @param supplementary 32-bit code point (U+10000..U+10ffff) 119 * @return lead surrogate (U+d800..U+dbff) for supplementary 120 * @stable ICU 2.4 121 */ 122 #define U16_LEAD(supplementary) (UChar)(((supplementary)>>10)+0xd7c0) 123 124 /** 125 * Get the trail surrogate (0xdc00..0xdfff) for a 126 * supplementary code point (0x10000..0x10ffff). 127 * @param supplementary 32-bit code point (U+10000..U+10ffff) 128 * @return trail surrogate (U+dc00..U+dfff) for supplementary 129 * @stable ICU 2.4 130 */ 131 #define U16_TRAIL(supplementary) (UChar)(((supplementary)&0x3ff)|0xdc00) 132 133 /** 134 * How many 16-bit code units are used to encode this Unicode code point? (1 or 2) 135 * The result is not defined if c is not a Unicode code point (U+0000..U+10ffff). 136 * @param c 32-bit code point 137 * @return 1 or 2 138 * @stable ICU 2.4 139 */ 140 #define U16_LENGTH(c) ((uint32_t)(c)<=0xffff ? 1 : 2) 141 142 /** 143 * The maximum number of 16-bit code units per Unicode code point (U+0000..U+10ffff). 144 * @return 2 145 * @stable ICU 2.4 146 */ 147 #define U16_MAX_LENGTH 2 148 149 /** 150 * Get a code point from a string at a random-access offset, 151 * without changing the offset. 152 * "Unsafe" macro, assumes well-formed UTF-16. 153 * 154 * The offset may point to either the lead or trail surrogate unit 155 * for a supplementary code point, in which case the macro will read 156 * the adjacent matching surrogate as well. 157 * The result is undefined if the offset points to a single, unpaired surrogate. 158 * Iteration through a string is more efficient with U16_NEXT_UNSAFE or U16_NEXT. 159 * 160 * @param s const UChar * string 161 * @param i string offset 162 * @param c output UChar32 variable 163 * @see U16_GET 164 * @stable ICU 2.4 165 */ 166 #define U16_GET_UNSAFE(s, i, c) { \ 167 (c)=(s)[i]; \ 168 if(U16_IS_SURROGATE(c)) { \ 169 if(U16_IS_SURROGATE_LEAD(c)) { \ 170 (c)=U16_GET_SUPPLEMENTARY((c), (s)[(i)+1]); \ 171 } else { \ 172 (c)=U16_GET_SUPPLEMENTARY((s)[(i)-1], (c)); \ 173 } \ 174 } \ 175 } 176 177 /** 178 * Get a code point from a string at a random-access offset, 179 * without changing the offset. 180 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 181 * 182 * The offset may point to either the lead or trail surrogate unit 183 * for a supplementary code point, in which case the macro will read 184 * the adjacent matching surrogate as well. 185 * 186 * The length can be negative for a NUL-terminated string. 187 * 188 * If the offset points to a single, unpaired surrogate, then 189 * c is set to that unpaired surrogate. 190 * Iteration through a string is more efficient with U16_NEXT_UNSAFE or U16_NEXT. 191 * 192 * @param s const UChar * string 193 * @param start starting string offset (usually 0) 194 * @param i string offset, must be start<=i<length 195 * @param length string length 196 * @param c output UChar32 variable 197 * @see U16_GET_UNSAFE 198 * @stable ICU 2.4 199 */ 200 #define U16_GET(s, start, i, length, c) { \ 201 (c)=(s)[i]; \ 202 if(U16_IS_SURROGATE(c)) { \ 203 uint16_t __c2; \ 204 if(U16_IS_SURROGATE_LEAD(c)) { \ 205 if((i)+1!=(length) && U16_IS_TRAIL(__c2=(s)[(i)+1])) { \ 206 (c)=U16_GET_SUPPLEMENTARY((c), __c2); \ 207 } \ 208 } else { \ 209 if((i)>(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \ 210 (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \ 211 } \ 212 } \ 213 } \ 214 } 215 216 #ifndef U_HIDE_DRAFT_API 217 218 /** 219 * Get a code point from a string at a random-access offset, 220 * without changing the offset. 221 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 222 * 223 * The offset may point to either the lead or trail surrogate unit 224 * for a supplementary code point, in which case the macro will read 225 * the adjacent matching surrogate as well. 226 * 227 * The length can be negative for a NUL-terminated string. 228 * 229 * If the offset points to a single, unpaired surrogate, then 230 * c is set to U+FFFD. 231 * Iteration through a string is more efficient with U16_NEXT_UNSAFE or U16_NEXT_OR_FFFD. 232 * 233 * @param s const UChar * string 234 * @param start starting string offset (usually 0) 235 * @param i string offset, must be start<=i<length 236 * @param length string length 237 * @param c output UChar32 variable 238 * @see U16_GET_UNSAFE 239 * @draft ICU 60 240 */ 241 #define U16_GET_OR_FFFD(s, start, i, length, c) { \ 242 (c)=(s)[i]; \ 243 if(U16_IS_SURROGATE(c)) { \ 244 uint16_t __c2; \ 245 if(U16_IS_SURROGATE_LEAD(c)) { \ 246 if((i)+1!=(length) && U16_IS_TRAIL(__c2=(s)[(i)+1])) { \ 247 (c)=U16_GET_SUPPLEMENTARY((c), __c2); \ 248 } else { \ 249 (c)=0xfffd; \ 250 } \ 251 } else { \ 252 if((i)>(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \ 253 (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \ 254 } else { \ 255 (c)=0xfffd; \ 256 } \ 257 } \ 258 } \ 259 } 260 261 #endif // U_HIDE_DRAFT_API 262 263 /* definitions with forward iteration --------------------------------------- */ 264 265 /** 266 * Get a code point from a string at a code point boundary offset, 267 * and advance the offset to the next code point boundary. 268 * (Post-incrementing forward iteration.) 269 * "Unsafe" macro, assumes well-formed UTF-16. 270 * 271 * The offset may point to the lead surrogate unit 272 * for a supplementary code point, in which case the macro will read 273 * the following trail surrogate as well. 274 * If the offset points to a trail surrogate, then that itself 275 * will be returned as the code point. 276 * The result is undefined if the offset points to a single, unpaired lead surrogate. 277 * 278 * @param s const UChar * string 279 * @param i string offset 280 * @param c output UChar32 variable 281 * @see U16_NEXT 282 * @stable ICU 2.4 283 */ 284 #define U16_NEXT_UNSAFE(s, i, c) { \ 285 (c)=(s)[(i)++]; \ 286 if(U16_IS_LEAD(c)) { \ 287 (c)=U16_GET_SUPPLEMENTARY((c), (s)[(i)++]); \ 288 } \ 289 } 290 291 /** 292 * Get a code point from a string at a code point boundary offset, 293 * and advance the offset to the next code point boundary. 294 * (Post-incrementing forward iteration.) 295 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 296 * 297 * The length can be negative for a NUL-terminated string. 298 * 299 * The offset may point to the lead surrogate unit 300 * for a supplementary code point, in which case the macro will read 301 * the following trail surrogate as well. 302 * If the offset points to a trail surrogate or 303 * to a single, unpaired lead surrogate, then c is set to that unpaired surrogate. 304 * 305 * @param s const UChar * string 306 * @param i string offset, must be i<length 307 * @param length string length 308 * @param c output UChar32 variable 309 * @see U16_NEXT_UNSAFE 310 * @stable ICU 2.4 311 */ 312 #define U16_NEXT(s, i, length, c) { \ 313 (c)=(s)[(i)++]; \ 314 if(U16_IS_LEAD(c)) { \ 315 uint16_t __c2; \ 316 if((i)!=(length) && U16_IS_TRAIL(__c2=(s)[(i)])) { \ 317 ++(i); \ 318 (c)=U16_GET_SUPPLEMENTARY((c), __c2); \ 319 } \ 320 } \ 321 } 322 323 #ifndef U_HIDE_DRAFT_API 324 325 /** 326 * Get a code point from a string at a code point boundary offset, 327 * and advance the offset to the next code point boundary. 328 * (Post-incrementing forward iteration.) 329 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 330 * 331 * The length can be negative for a NUL-terminated string. 332 * 333 * The offset may point to the lead surrogate unit 334 * for a supplementary code point, in which case the macro will read 335 * the following trail surrogate as well. 336 * If the offset points to a trail surrogate or 337 * to a single, unpaired lead surrogate, then c is set to U+FFFD. 338 * 339 * @param s const UChar * string 340 * @param i string offset, must be i<length 341 * @param length string length 342 * @param c output UChar32 variable 343 * @see U16_NEXT_UNSAFE 344 * @draft ICU 60 345 */ 346 #define U16_NEXT_OR_FFFD(s, i, length, c) { \ 347 (c)=(s)[(i)++]; \ 348 if(U16_IS_SURROGATE(c)) { \ 349 uint16_t __c2; \ 350 if(U16_IS_SURROGATE_LEAD(c) && (i)!=(length) && U16_IS_TRAIL(__c2=(s)[(i)])) { \ 351 ++(i); \ 352 (c)=U16_GET_SUPPLEMENTARY((c), __c2); \ 353 } else { \ 354 (c)=0xfffd; \ 355 } \ 356 } \ 357 } 358 359 #endif // U_HIDE_DRAFT_API 360 361 /** 362 * Append a code point to a string, overwriting 1 or 2 code units. 363 * The offset points to the current end of the string contents 364 * and is advanced (post-increment). 365 * "Unsafe" macro, assumes a valid code point and sufficient space in the string. 366 * Otherwise, the result is undefined. 367 * 368 * @param s const UChar * string buffer 369 * @param i string offset 370 * @param c code point to append 371 * @see U16_APPEND 372 * @stable ICU 2.4 373 */ 374 #define U16_APPEND_UNSAFE(s, i, c) { \ 375 if((uint32_t)(c)<=0xffff) { \ 376 (s)[(i)++]=(uint16_t)(c); \ 377 } else { \ 378 (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \ 379 (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \ 380 } \ 381 } 382 383 /** 384 * Append a code point to a string, overwriting 1 or 2 code units. 385 * The offset points to the current end of the string contents 386 * and is advanced (post-increment). 387 * "Safe" macro, checks for a valid code point. 388 * If a surrogate pair is written, checks for sufficient space in the string. 389 * If the code point is not valid or a trail surrogate does not fit, 390 * then isError is set to TRUE. 391 * 392 * @param s const UChar * string buffer 393 * @param i string offset, must be i<capacity 394 * @param capacity size of the string buffer 395 * @param c code point to append 396 * @param isError output UBool set to TRUE if an error occurs, otherwise not modified 397 * @see U16_APPEND_UNSAFE 398 * @stable ICU 2.4 399 */ 400 #define U16_APPEND(s, i, capacity, c, isError) { \ 401 if((uint32_t)(c)<=0xffff) { \ 402 (s)[(i)++]=(uint16_t)(c); \ 403 } else if((uint32_t)(c)<=0x10ffff && (i)+1<(capacity)) { \ 404 (s)[(i)++]=(uint16_t)(((c)>>10)+0xd7c0); \ 405 (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \ 406 } else /* c>0x10ffff or not enough space */ { \ 407 (isError)=TRUE; \ 408 } \ 409 } 410 411 /** 412 * Advance the string offset from one code point boundary to the next. 413 * (Post-incrementing iteration.) 414 * "Unsafe" macro, assumes well-formed UTF-16. 415 * 416 * @param s const UChar * string 417 * @param i string offset 418 * @see U16_FWD_1 419 * @stable ICU 2.4 420 */ 421 #define U16_FWD_1_UNSAFE(s, i) { \ 422 if(U16_IS_LEAD((s)[(i)++])) { \ 423 ++(i); \ 424 } \ 425 } 426 427 /** 428 * Advance the string offset from one code point boundary to the next. 429 * (Post-incrementing iteration.) 430 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 431 * 432 * The length can be negative for a NUL-terminated string. 433 * 434 * @param s const UChar * string 435 * @param i string offset, must be i<length 436 * @param length string length 437 * @see U16_FWD_1_UNSAFE 438 * @stable ICU 2.4 439 */ 440 #define U16_FWD_1(s, i, length) { \ 441 if(U16_IS_LEAD((s)[(i)++]) && (i)!=(length) && U16_IS_TRAIL((s)[i])) { \ 442 ++(i); \ 443 } \ 444 } 445 446 /** 447 * Advance the string offset from one code point boundary to the n-th next one, 448 * i.e., move forward by n code points. 449 * (Post-incrementing iteration.) 450 * "Unsafe" macro, assumes well-formed UTF-16. 451 * 452 * @param s const UChar * string 453 * @param i string offset 454 * @param n number of code points to skip 455 * @see U16_FWD_N 456 * @stable ICU 2.4 457 */ 458 #define U16_FWD_N_UNSAFE(s, i, n) { \ 459 int32_t __N=(n); \ 460 while(__N>0) { \ 461 U16_FWD_1_UNSAFE(s, i); \ 462 --__N; \ 463 } \ 464 } 465 466 /** 467 * Advance the string offset from one code point boundary to the n-th next one, 468 * i.e., move forward by n code points. 469 * (Post-incrementing iteration.) 470 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 471 * 472 * The length can be negative for a NUL-terminated string. 473 * 474 * @param s const UChar * string 475 * @param i int32_t string offset, must be i<length 476 * @param length int32_t string length 477 * @param n number of code points to skip 478 * @see U16_FWD_N_UNSAFE 479 * @stable ICU 2.4 480 */ 481 #define U16_FWD_N(s, i, length, n) { \ 482 int32_t __N=(n); \ 483 while(__N>0 && ((i)<(length) || ((length)<0 && (s)[i]!=0))) { \ 484 U16_FWD_1(s, i, length); \ 485 --__N; \ 486 } \ 487 } 488 489 /** 490 * Adjust a random-access offset to a code point boundary 491 * at the start of a code point. 492 * If the offset points to the trail surrogate of a surrogate pair, 493 * then the offset is decremented. 494 * Otherwise, it is not modified. 495 * "Unsafe" macro, assumes well-formed UTF-16. 496 * 497 * @param s const UChar * string 498 * @param i string offset 499 * @see U16_SET_CP_START 500 * @stable ICU 2.4 501 */ 502 #define U16_SET_CP_START_UNSAFE(s, i) { \ 503 if(U16_IS_TRAIL((s)[i])) { \ 504 --(i); \ 505 } \ 506 } 507 508 /** 509 * Adjust a random-access offset to a code point boundary 510 * at the start of a code point. 511 * If the offset points to the trail surrogate of a surrogate pair, 512 * then the offset is decremented. 513 * Otherwise, it is not modified. 514 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 515 * 516 * @param s const UChar * string 517 * @param start starting string offset (usually 0) 518 * @param i string offset, must be start<=i 519 * @see U16_SET_CP_START_UNSAFE 520 * @stable ICU 2.4 521 */ 522 #define U16_SET_CP_START(s, start, i) { \ 523 if(U16_IS_TRAIL((s)[i]) && (i)>(start) && U16_IS_LEAD((s)[(i)-1])) { \ 524 --(i); \ 525 } \ 526 } 527 528 /* definitions with backward iteration -------------------------------------- */ 529 530 /** 531 * Move the string offset from one code point boundary to the previous one 532 * and get the code point between them. 533 * (Pre-decrementing backward iteration.) 534 * "Unsafe" macro, assumes well-formed UTF-16. 535 * 536 * The input offset may be the same as the string length. 537 * If the offset is behind a trail surrogate unit 538 * for a supplementary code point, then the macro will read 539 * the preceding lead surrogate as well. 540 * If the offset is behind a lead surrogate, then that itself 541 * will be returned as the code point. 542 * The result is undefined if the offset is behind a single, unpaired trail surrogate. 543 * 544 * @param s const UChar * string 545 * @param i string offset 546 * @param c output UChar32 variable 547 * @see U16_PREV 548 * @stable ICU 2.4 549 */ 550 #define U16_PREV_UNSAFE(s, i, c) { \ 551 (c)=(s)[--(i)]; \ 552 if(U16_IS_TRAIL(c)) { \ 553 (c)=U16_GET_SUPPLEMENTARY((s)[--(i)], (c)); \ 554 } \ 555 } 556 557 /** 558 * Move the string offset from one code point boundary to the previous one 559 * and get the code point between them. 560 * (Pre-decrementing backward iteration.) 561 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 562 * 563 * The input offset may be the same as the string length. 564 * If the offset is behind a trail surrogate unit 565 * for a supplementary code point, then the macro will read 566 * the preceding lead surrogate as well. 567 * If the offset is behind a lead surrogate or behind a single, unpaired 568 * trail surrogate, then c is set to that unpaired surrogate. 569 * 570 * @param s const UChar * string 571 * @param start starting string offset (usually 0) 572 * @param i string offset, must be start<i 573 * @param c output UChar32 variable 574 * @see U16_PREV_UNSAFE 575 * @stable ICU 2.4 576 */ 577 #define U16_PREV(s, start, i, c) { \ 578 (c)=(s)[--(i)]; \ 579 if(U16_IS_TRAIL(c)) { \ 580 uint16_t __c2; \ 581 if((i)>(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \ 582 --(i); \ 583 (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \ 584 } \ 585 } \ 586 } 587 588 #ifndef U_HIDE_DRAFT_API 589 590 /** 591 * Move the string offset from one code point boundary to the previous one 592 * and get the code point between them. 593 * (Pre-decrementing backward iteration.) 594 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 595 * 596 * The input offset may be the same as the string length. 597 * If the offset is behind a trail surrogate unit 598 * for a supplementary code point, then the macro will read 599 * the preceding lead surrogate as well. 600 * If the offset is behind a lead surrogate or behind a single, unpaired 601 * trail surrogate, then c is set to U+FFFD. 602 * 603 * @param s const UChar * string 604 * @param start starting string offset (usually 0) 605 * @param i string offset, must be start<i 606 * @param c output UChar32 variable 607 * @see U16_PREV_UNSAFE 608 * @draft ICU 60 609 */ 610 #define U16_PREV_OR_FFFD(s, start, i, c) { \ 611 (c)=(s)[--(i)]; \ 612 if(U16_IS_SURROGATE(c)) { \ 613 uint16_t __c2; \ 614 if(U16_IS_SURROGATE_TRAIL(c) && (i)>(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \ 615 --(i); \ 616 (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \ 617 } else { \ 618 (c)=0xfffd; \ 619 } \ 620 } \ 621 } 622 623 #endif // U_HIDE_DRAFT_API 624 625 /** 626 * Move the string offset from one code point boundary to the previous one. 627 * (Pre-decrementing backward iteration.) 628 * The input offset may be the same as the string length. 629 * "Unsafe" macro, assumes well-formed UTF-16. 630 * 631 * @param s const UChar * string 632 * @param i string offset 633 * @see U16_BACK_1 634 * @stable ICU 2.4 635 */ 636 #define U16_BACK_1_UNSAFE(s, i) { \ 637 if(U16_IS_TRAIL((s)[--(i)])) { \ 638 --(i); \ 639 } \ 640 } 641 642 /** 643 * Move the string offset from one code point boundary to the previous one. 644 * (Pre-decrementing backward iteration.) 645 * The input offset may be the same as the string length. 646 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 647 * 648 * @param s const UChar * string 649 * @param start starting string offset (usually 0) 650 * @param i string offset, must be start<i 651 * @see U16_BACK_1_UNSAFE 652 * @stable ICU 2.4 653 */ 654 #define U16_BACK_1(s, start, i) { \ 655 if(U16_IS_TRAIL((s)[--(i)]) && (i)>(start) && U16_IS_LEAD((s)[(i)-1])) { \ 656 --(i); \ 657 } \ 658 } 659 660 /** 661 * Move the string offset from one code point boundary to the n-th one before it, 662 * i.e., move backward by n code points. 663 * (Pre-decrementing backward iteration.) 664 * The input offset may be the same as the string length. 665 * "Unsafe" macro, assumes well-formed UTF-16. 666 * 667 * @param s const UChar * string 668 * @param i string offset 669 * @param n number of code points to skip 670 * @see U16_BACK_N 671 * @stable ICU 2.4 672 */ 673 #define U16_BACK_N_UNSAFE(s, i, n) { \ 674 int32_t __N=(n); \ 675 while(__N>0) { \ 676 U16_BACK_1_UNSAFE(s, i); \ 677 --__N; \ 678 } \ 679 } 680 681 /** 682 * Move the string offset from one code point boundary to the n-th one before it, 683 * i.e., move backward by n code points. 684 * (Pre-decrementing backward iteration.) 685 * The input offset may be the same as the string length. 686 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 687 * 688 * @param s const UChar * string 689 * @param start start of string 690 * @param i string offset, must be start<i 691 * @param n number of code points to skip 692 * @see U16_BACK_N_UNSAFE 693 * @stable ICU 2.4 694 */ 695 #define U16_BACK_N(s, start, i, n) { \ 696 int32_t __N=(n); \ 697 while(__N>0 && (i)>(start)) { \ 698 U16_BACK_1(s, start, i); \ 699 --__N; \ 700 } \ 701 } 702 703 /** 704 * Adjust a random-access offset to a code point boundary after a code point. 705 * If the offset is behind the lead surrogate of a surrogate pair, 706 * then the offset is incremented. 707 * Otherwise, it is not modified. 708 * The input offset may be the same as the string length. 709 * "Unsafe" macro, assumes well-formed UTF-16. 710 * 711 * @param s const UChar * string 712 * @param i string offset 713 * @see U16_SET_CP_LIMIT 714 * @stable ICU 2.4 715 */ 716 #define U16_SET_CP_LIMIT_UNSAFE(s, i) { \ 717 if(U16_IS_LEAD((s)[(i)-1])) { \ 718 ++(i); \ 719 } \ 720 } 721 722 /** 723 * Adjust a random-access offset to a code point boundary after a code point. 724 * If the offset is behind the lead surrogate of a surrogate pair, 725 * then the offset is incremented. 726 * Otherwise, it is not modified. 727 * The input offset may be the same as the string length. 728 * "Safe" macro, handles unpaired surrogates and checks for string boundaries. 729 * 730 * The length can be negative for a NUL-terminated string. 731 * 732 * @param s const UChar * string 733 * @param start int32_t starting string offset (usually 0) 734 * @param i int32_t string offset, start<=i<=length 735 * @param length int32_t string length 736 * @see U16_SET_CP_LIMIT_UNSAFE 737 * @stable ICU 2.4 738 */ 739 #define U16_SET_CP_LIMIT(s, start, i, length) { \ 740 if((start)<(i) && ((i)<(length) || (length)<0) && U16_IS_LEAD((s)[(i)-1]) && U16_IS_TRAIL((s)[i])) { \ 741 ++(i); \ 742 } \ 743 } 744 745 #endif 746