1 // 2017 and later: Unicode, Inc. and others. 2 // License & terms of use: http://www.unicode.org/copyright.html 3 4 // ucptrie.h (modified from utrie2.h) 5 // created: 2017dec29 Markus W. Scherer 6 7 #ifndef __UCPTRIE_H__ 8 #define __UCPTRIE_H__ 9 10 #include "unicode/utypes.h" 11 12 #ifndef U_HIDE_DRAFT_API 13 14 #include "unicode/localpointer.h" 15 #include "unicode/ucpmap.h" 16 #include "unicode/utf8.h" 17 18 U_CDECL_BEGIN 19 20 /** 21 * \file 22 * 23 * This file defines an immutable Unicode code point trie. 24 * 25 * @see UCPTrie 26 * @see UMutableCPTrie 27 */ 28 29 #ifndef U_IN_DOXYGEN 30 /** @internal */ 31 typedef union UCPTrieData { 32 /** @internal */ 33 const void *ptr0; 34 /** @internal */ 35 const uint16_t *ptr16; 36 /** @internal */ 37 const uint32_t *ptr32; 38 /** @internal */ 39 const uint8_t *ptr8; 40 } UCPTrieData; 41 #endif 42 43 /** 44 * Immutable Unicode code point trie structure. 45 * Fast, reasonably compact, map from Unicode code points (U+0000..U+10FFFF) to integer values. 46 * For details see http://site.icu-project.org/design/struct/utrie 47 * 48 * Do not access UCPTrie fields directly; use public functions and macros. 49 * Functions are easy to use: They support all trie types and value widths. 50 * 51 * When performance is really important, macros provide faster access. 52 * Most macros are specific to either "fast" or "small" tries, see UCPTrieType. 53 * There are "fast" macros for special optimized use cases. 54 * 55 * The macros will return bogus values, or may crash, if used on the wrong type or value width. 56 * 57 * @see UMutableCPTrie 58 * @draft ICU 63 59 */ 60 struct UCPTrie { 61 #ifndef U_IN_DOXYGEN 62 /** @internal */ 63 const uint16_t *index; 64 /** @internal */ 65 UCPTrieData data; 66 67 /** @internal */ 68 int32_t indexLength; 69 /** @internal */ 70 int32_t dataLength; 71 /** Start of the last range which ends at U+10FFFF. @internal */ 72 UChar32 highStart; 73 /** highStart>>12 @internal */ 74 uint16_t shifted12HighStart; 75 76 /** @internal */ 77 int8_t type; // UCPTrieType 78 /** @internal */ 79 int8_t valueWidth; // UCPTrieValueWidth 80 81 /** padding/reserved @internal */ 82 uint32_t reserved32; 83 /** padding/reserved @internal */ 84 uint16_t reserved16; 85 86 /** 87 * Internal index-3 null block offset. 88 * Set to an impossibly high value (e.g., 0xffff) if there is no dedicated index-3 null block. 89 * @internal 90 */ 91 uint16_t index3NullOffset; 92 /** 93 * Internal data null block offset, not shifted. 94 * Set to an impossibly high value (e.g., 0xfffff) if there is no dedicated data null block. 95 * @internal 96 */ 97 int32_t dataNullOffset; 98 /** @internal */ 99 uint32_t nullValue; 100 101 #ifdef UCPTRIE_DEBUG 102 /** @internal */ 103 const char *name; 104 #endif 105 #endif 106 }; 107 #ifndef U_IN_DOXYGEN 108 typedef struct UCPTrie UCPTrie; 109 #endif 110 111 /** 112 * Selectors for the type of a UCPTrie. 113 * Different trade-offs for size vs. speed. 114 * 115 * @see umutablecptrie_buildImmutable 116 * @see ucptrie_openFromBinary 117 * @see ucptrie_getType 118 * @draft ICU 63 119 */ 120 enum UCPTrieType { 121 /** 122 * For ucptrie_openFromBinary() to accept any type. 123 * ucptrie_getType() will return the actual type. 124 * @draft ICU 63 125 */ 126 UCPTRIE_TYPE_ANY = -1, 127 /** 128 * Fast/simple/larger BMP data structure. Use functions and "fast" macros. 129 * @draft ICU 63 130 */ 131 UCPTRIE_TYPE_FAST, 132 /** 133 * Small/slower BMP data structure. Use functions and "small" macros. 134 * @draft ICU 63 135 */ 136 UCPTRIE_TYPE_SMALL 137 }; 138 #ifndef U_IN_DOXYGEN 139 typedef enum UCPTrieType UCPTrieType; 140 #endif 141 142 /** 143 * Selectors for the number of bits in a UCPTrie data value. 144 * 145 * @see umutablecptrie_buildImmutable 146 * @see ucptrie_openFromBinary 147 * @see ucptrie_getValueWidth 148 * @draft ICU 63 149 */ 150 enum UCPTrieValueWidth { 151 /** 152 * For ucptrie_openFromBinary() to accept any data value width. 153 * ucptrie_getValueWidth() will return the actual data value width. 154 * @draft ICU 63 155 */ 156 UCPTRIE_VALUE_BITS_ANY = -1, 157 /** 158 * The trie stores 16 bits per data value. 159 * It returns them as unsigned values 0..0xffff=65535. 160 * @draft ICU 63 161 */ 162 UCPTRIE_VALUE_BITS_16, 163 /** 164 * The trie stores 32 bits per data value. 165 * @draft ICU 63 166 */ 167 UCPTRIE_VALUE_BITS_32, 168 /** 169 * The trie stores 8 bits per data value. 170 * It returns them as unsigned values 0..0xff=255. 171 * @draft ICU 63 172 */ 173 UCPTRIE_VALUE_BITS_8 174 }; 175 #ifndef U_IN_DOXYGEN 176 typedef enum UCPTrieValueWidth UCPTrieValueWidth; 177 #endif 178 179 /** 180 * Opens a trie from its binary form, stored in 32-bit-aligned memory. 181 * Inverse of ucptrie_toBinary(). 182 * 183 * The memory must remain valid and unchanged as long as the trie is used. 184 * You must ucptrie_close() the trie once you are done using it. 185 * 186 * @param type selects the trie type; results in an 187 * U_INVALID_FORMAT_ERROR if it does not match the binary data; 188 * use UCPTRIE_TYPE_ANY to accept any type 189 * @param valueWidth selects the number of bits in a data value; results in an 190 * U_INVALID_FORMAT_ERROR if it does not match the binary data; 191 * use UCPTRIE_VALUE_BITS_ANY to accept any data value width 192 * @param data a pointer to 32-bit-aligned memory containing the binary data of a UCPTrie 193 * @param length the number of bytes available at data; 194 * can be more than necessary 195 * @param pActualLength receives the actual number of bytes at data taken up by the trie data; 196 * can be NULL 197 * @param pErrorCode an in/out ICU UErrorCode 198 * @return the trie 199 * 200 * @see umutablecptrie_open 201 * @see umutablecptrie_buildImmutable 202 * @see ucptrie_toBinary 203 * @draft ICU 63 204 */ 205 U_CAPI UCPTrie * U_EXPORT2 206 ucptrie_openFromBinary(UCPTrieType type, UCPTrieValueWidth valueWidth, 207 const void *data, int32_t length, int32_t *pActualLength, 208 UErrorCode *pErrorCode); 209 210 /** 211 * Closes a trie and releases associated memory. 212 * 213 * @param trie the trie 214 * @draft ICU 63 215 */ 216 U_CAPI void U_EXPORT2 217 ucptrie_close(UCPTrie *trie); 218 219 #if U_SHOW_CPLUSPLUS_API 220 221 U_NAMESPACE_BEGIN 222 223 /** 224 * \class LocalUCPTriePointer 225 * "Smart pointer" class, closes a UCPTrie via ucptrie_close(). 226 * For most methods see the LocalPointerBase base class. 227 * 228 * @see LocalPointerBase 229 * @see LocalPointer 230 * @draft ICU 63 231 */ 232 U_DEFINE_LOCAL_OPEN_POINTER(LocalUCPTriePointer, UCPTrie, ucptrie_close); 233 234 U_NAMESPACE_END 235 236 #endif 237 238 /** 239 * Returns the trie type. 240 * 241 * @param trie the trie 242 * @return the trie type 243 * @see ucptrie_openFromBinary 244 * @see UCPTRIE_TYPE_ANY 245 * @draft ICU 63 246 */ 247 U_CAPI UCPTrieType U_EXPORT2 248 ucptrie_getType(const UCPTrie *trie); 249 250 /** 251 * Returns the number of bits in a trie data value. 252 * 253 * @param trie the trie 254 * @return the number of bits in a trie data value 255 * @see ucptrie_openFromBinary 256 * @see UCPTRIE_VALUE_BITS_ANY 257 * @draft ICU 63 258 */ 259 U_CAPI UCPTrieValueWidth U_EXPORT2 260 ucptrie_getValueWidth(const UCPTrie *trie); 261 262 /** 263 * Returns the value for a code point as stored in the trie, with range checking. 264 * Returns the trie error value if c is not in the range 0..U+10FFFF. 265 * 266 * Easier to use than UCPTRIE_FAST_GET() and similar macros but slower. 267 * Easier to use because, unlike the macros, this function works on all UCPTrie 268 * objects, for all types and value widths. 269 * 270 * @param trie the trie 271 * @param c the code point 272 * @return the trie value, 273 * or the trie error value if the code point is not in the range 0..U+10FFFF 274 * @draft ICU 63 275 */ 276 U_CAPI uint32_t U_EXPORT2 277 ucptrie_get(const UCPTrie *trie, UChar32 c); 278 279 /** 280 * Returns the last code point such that all those from start to there have the same value. 281 * Can be used to efficiently iterate over all same-value ranges in a trie. 282 * (This is normally faster than iterating over code points and get()ting each value, 283 * but much slower than a data structure that stores ranges directly.) 284 * 285 * If the UCPMapValueFilter function pointer is not NULL, then 286 * the value to be delivered is passed through that function, and the return value is the end 287 * of the range where all values are modified to the same actual value. 288 * The value is unchanged if that function pointer is NULL. 289 * 290 * Example: 291 * \code 292 * UChar32 start = 0, end; 293 * uint32_t value; 294 * while ((end = ucptrie_getRange(trie, start, UCPMAP_RANGE_NORMAL, 0, 295 * NULL, NULL, &value)) >= 0) { 296 * // Work with the range start..end and its value. 297 * start = end + 1; 298 * } 299 * \endcode 300 * 301 * @param trie the trie 302 * @param start range start 303 * @param option defines whether surrogates are treated normally, 304 * or as having the surrogateValue; usually UCPMAP_RANGE_NORMAL 305 * @param surrogateValue value for surrogates; ignored if option==UCPMAP_RANGE_NORMAL 306 * @param filter a pointer to a function that may modify the trie data value, 307 * or NULL if the values from the trie are to be used unmodified 308 * @param context an opaque pointer that is passed on to the filter function 309 * @param pValue if not NULL, receives the value that every code point start..end has; 310 * may have been modified by filter(context, trie value) 311 * if that function pointer is not NULL 312 * @return the range end code point, or -1 if start is not a valid code point 313 * @draft ICU 63 314 */ 315 U_CAPI UChar32 U_EXPORT2 316 ucptrie_getRange(const UCPTrie *trie, UChar32 start, 317 UCPMapRangeOption option, uint32_t surrogateValue, 318 UCPMapValueFilter *filter, const void *context, uint32_t *pValue); 319 320 /** 321 * Writes a memory-mappable form of the trie into 32-bit aligned memory. 322 * Inverse of ucptrie_openFromBinary(). 323 * 324 * @param trie the trie 325 * @param data a pointer to 32-bit-aligned memory to be filled with the trie data; 326 * can be NULL if capacity==0 327 * @param capacity the number of bytes available at data, or 0 for pure preflighting 328 * @param pErrorCode an in/out ICU UErrorCode; 329 * U_BUFFER_OVERFLOW_ERROR if the capacity is too small 330 * @return the number of bytes written or (if buffer overflow) needed for the trie 331 * 332 * @see ucptrie_openFromBinary() 333 * @draft ICU 63 334 */ 335 U_CAPI int32_t U_EXPORT2 336 ucptrie_toBinary(const UCPTrie *trie, void *data, int32_t capacity, UErrorCode *pErrorCode); 337 338 /** 339 * Macro parameter value for a trie with 16-bit data values. 340 * Use the name of this macro as a "dataAccess" parameter in other macros. 341 * Do not use this macro in any other way. 342 * 343 * @see UCPTRIE_VALUE_BITS_16 344 * @draft ICU 63 345 */ 346 #define UCPTRIE_16(trie, i) ((trie)->data.ptr16[i]) 347 348 /** 349 * Macro parameter value for a trie with 32-bit data values. 350 * Use the name of this macro as a "dataAccess" parameter in other macros. 351 * Do not use this macro in any other way. 352 * 353 * @see UCPTRIE_VALUE_BITS_32 354 * @draft ICU 63 355 */ 356 #define UCPTRIE_32(trie, i) ((trie)->data.ptr32[i]) 357 358 /** 359 * Macro parameter value for a trie with 8-bit data values. 360 * Use the name of this macro as a "dataAccess" parameter in other macros. 361 * Do not use this macro in any other way. 362 * 363 * @see UCPTRIE_VALUE_BITS_8 364 * @draft ICU 63 365 */ 366 #define UCPTRIE_8(trie, i) ((trie)->data.ptr8[i]) 367 368 /** 369 * Returns a trie value for a code point, with range checking. 370 * Returns the trie error value if c is not in the range 0..U+10FFFF. 371 * 372 * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST 373 * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the tries value width 374 * @param c (UChar32, in) the input code point 375 * @return The code point's trie value. 376 * @draft ICU 63 377 */ 378 #define UCPTRIE_FAST_GET(trie, dataAccess, c) dataAccess(trie, _UCPTRIE_CP_INDEX(trie, 0xffff, c)) 379 380 /** 381 * Returns a 16-bit trie value for a code point, with range checking. 382 * Returns the trie error value if c is not in the range U+0000..U+10FFFF. 383 * 384 * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_SMALL 385 * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the tries value width 386 * @param c (UChar32, in) the input code point 387 * @return The code point's trie value. 388 * @draft ICU 63 389 */ 390 #define UCPTRIE_SMALL_GET(trie, dataAccess, c) \ 391 dataAccess(trie, _UCPTRIE_CP_INDEX(trie, UCPTRIE_SMALL_MAX, c)) 392 393 /** 394 * UTF-16: Reads the next code point (UChar32 c, out), post-increments src, 395 * and gets a value from the trie. 396 * Sets the trie error value if c is an unpaired surrogate. 397 * 398 * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST 399 * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the tries value width 400 * @param src (const UChar *, in/out) the source text pointer 401 * @param limit (const UChar *, in) the limit pointer for the text, or NULL if NUL-terminated 402 * @param c (UChar32, out) variable for the code point 403 * @param result (out) variable for the trie lookup result 404 * @draft ICU 63 405 */ 406 #define UCPTRIE_FAST_U16_NEXT(trie, dataAccess, src, limit, c, result) { \ 407 (c) = *(src)++; \ 408 int32_t __index; \ 409 if (!U16_IS_SURROGATE(c)) { \ 410 __index = _UCPTRIE_FAST_INDEX(trie, c); \ 411 } else { \ 412 uint16_t __c2; \ 413 if (U16_IS_SURROGATE_LEAD(c) && (src) != (limit) && U16_IS_TRAIL(__c2 = *(src))) { \ 414 ++(src); \ 415 (c) = U16_GET_SUPPLEMENTARY((c), __c2); \ 416 __index = _UCPTRIE_SMALL_INDEX(trie, c); \ 417 } else { \ 418 __index = (trie)->dataLength - UCPTRIE_ERROR_VALUE_NEG_DATA_OFFSET; \ 419 } \ 420 } \ 421 (result) = dataAccess(trie, __index); \ 422 } 423 424 /** 425 * UTF-16: Reads the previous code point (UChar32 c, out), pre-decrements src, 426 * and gets a value from the trie. 427 * Sets the trie error value if c is an unpaired surrogate. 428 * 429 * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST 430 * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the tries value width 431 * @param start (const UChar *, in) the start pointer for the text 432 * @param src (const UChar *, in/out) the source text pointer 433 * @param c (UChar32, out) variable for the code point 434 * @param result (out) variable for the trie lookup result 435 * @draft ICU 63 436 */ 437 #define UCPTRIE_FAST_U16_PREV(trie, dataAccess, start, src, c, result) { \ 438 (c) = *--(src); \ 439 int32_t __index; \ 440 if (!U16_IS_SURROGATE(c)) { \ 441 __index = _UCPTRIE_FAST_INDEX(trie, c); \ 442 } else { \ 443 uint16_t __c2; \ 444 if (U16_IS_SURROGATE_TRAIL(c) && (src) != (start) && U16_IS_LEAD(__c2 = *((src) - 1))) { \ 445 --(src); \ 446 (c) = U16_GET_SUPPLEMENTARY(__c2, (c)); \ 447 __index = _UCPTRIE_SMALL_INDEX(trie, c); \ 448 } else { \ 449 __index = (trie)->dataLength - UCPTRIE_ERROR_VALUE_NEG_DATA_OFFSET; \ 450 } \ 451 } \ 452 (result) = dataAccess(trie, __index); \ 453 } 454 455 /** 456 * UTF-8: Post-increments src and gets a value from the trie. 457 * Sets the trie error value for an ill-formed byte sequence. 458 * 459 * Unlike UCPTRIE_FAST_U16_NEXT() this UTF-8 macro does not provide the code point 460 * because it would be more work to do so and is often not needed. 461 * If the trie value differs from the error value, then the byte sequence is well-formed, 462 * and the code point can be assembled without revalidation. 463 * 464 * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST 465 * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the tries value width 466 * @param src (const char *, in/out) the source text pointer 467 * @param limit (const char *, in) the limit pointer for the text (must not be NULL) 468 * @param result (out) variable for the trie lookup result 469 * @draft ICU 63 470 */ 471 #define UCPTRIE_FAST_U8_NEXT(trie, dataAccess, src, limit, result) { \ 472 int32_t __lead = (uint8_t)*(src)++; \ 473 if (!U8_IS_SINGLE(__lead)) { \ 474 uint8_t __t1, __t2, __t3; \ 475 if ((src) != (limit) && \ 476 (__lead >= 0xe0 ? \ 477 __lead < 0xf0 ? /* U+0800..U+FFFF except surrogates */ \ 478 U8_LEAD3_T1_BITS[__lead &= 0xf] & (1 << ((__t1 = *(src)) >> 5)) && \ 479 ++(src) != (limit) && (__t2 = *(src) - 0x80) <= 0x3f && \ 480 (__lead = ((int32_t)(trie)->index[(__lead << 6) + (__t1 & 0x3f)]) + __t2, 1) \ 481 : /* U+10000..U+10FFFF */ \ 482 (__lead -= 0xf0) <= 4 && \ 483 U8_LEAD4_T1_BITS[(__t1 = *(src)) >> 4] & (1 << __lead) && \ 484 (__lead = (__lead << 6) | (__t1 & 0x3f), ++(src) != (limit)) && \ 485 (__t2 = *(src) - 0x80) <= 0x3f && \ 486 ++(src) != (limit) && (__t3 = *(src) - 0x80) <= 0x3f && \ 487 (__lead = __lead >= (trie)->shifted12HighStart ? \ 488 (trie)->dataLength - UCPTRIE_HIGH_VALUE_NEG_DATA_OFFSET : \ 489 ucptrie_internalSmallU8Index((trie), __lead, __t2, __t3), 1) \ 490 : /* U+0080..U+07FF */ \ 491 __lead >= 0xc2 && (__t1 = *(src) - 0x80) <= 0x3f && \ 492 (__lead = (int32_t)(trie)->index[__lead & 0x1f] + __t1, 1))) { \ 493 ++(src); \ 494 } else { \ 495 __lead = (trie)->dataLength - UCPTRIE_ERROR_VALUE_NEG_DATA_OFFSET; /* ill-formed*/ \ 496 } \ 497 } \ 498 (result) = dataAccess(trie, __lead); \ 499 } 500 501 /** 502 * UTF-8: Pre-decrements src and gets a value from the trie. 503 * Sets the trie error value for an ill-formed byte sequence. 504 * 505 * Unlike UCPTRIE_FAST_U16_PREV() this UTF-8 macro does not provide the code point 506 * because it would be more work to do so and is often not needed. 507 * If the trie value differs from the error value, then the byte sequence is well-formed, 508 * and the code point can be assembled without revalidation. 509 * 510 * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST 511 * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the tries value width 512 * @param start (const char *, in) the start pointer for the text 513 * @param src (const char *, in/out) the source text pointer 514 * @param result (out) variable for the trie lookup result 515 * @draft ICU 63 516 */ 517 #define UCPTRIE_FAST_U8_PREV(trie, dataAccess, start, src, result) { \ 518 int32_t __index = (uint8_t)*--(src); \ 519 if (!U8_IS_SINGLE(__index)) { \ 520 __index = ucptrie_internalU8PrevIndex((trie), __index, (const uint8_t *)(start), \ 521 (const uint8_t *)(src)); \ 522 (src) -= __index & 7; \ 523 __index >>= 3; \ 524 } \ 525 (result) = dataAccess(trie, __index); \ 526 } 527 528 /** 529 * Returns a trie value for an ASCII code point, without range checking. 530 * 531 * @param trie (const UCPTrie *, in) the trie (of either fast or small type) 532 * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the tries value width 533 * @param c (UChar32, in) the input code point; must be U+0000..U+007F 534 * @return The ASCII code point's trie value. 535 * @draft ICU 63 536 */ 537 #define UCPTRIE_ASCII_GET(trie, dataAccess, c) dataAccess(trie, c) 538 539 /** 540 * Returns a trie value for a BMP code point (U+0000..U+FFFF), without range checking. 541 * Can be used to look up a value for a UTF-16 code unit if other parts of 542 * the string processing check for surrogates. 543 * 544 * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST 545 * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the tries value width 546 * @param c (UChar32, in) the input code point, must be U+0000..U+FFFF 547 * @return The BMP code point's trie value. 548 * @draft ICU 63 549 */ 550 #define UCPTRIE_FAST_BMP_GET(trie, dataAccess, c) dataAccess(trie, _UCPTRIE_FAST_INDEX(trie, c)) 551 552 /** 553 * Returns a trie value for a supplementary code point (U+10000..U+10FFFF), 554 * without range checking. 555 * 556 * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST 557 * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the tries value width 558 * @param c (UChar32, in) the input code point, must be U+10000..U+10FFFF 559 * @return The supplementary code point's trie value. 560 * @draft ICU 63 561 */ 562 #define UCPTRIE_FAST_SUPP_GET(trie, dataAccess, c) dataAccess(trie, _UCPTRIE_SMALL_INDEX(trie, c)) 563 564 /* Internal definitions ----------------------------------------------------- */ 565 566 #ifndef U_IN_DOXYGEN 567 568 /** 569 * Internal implementation constants. 570 * These are needed for the API macros, but users should not use these directly. 571 * @internal 572 */ 573 enum { 574 /** @internal */ 575 UCPTRIE_FAST_SHIFT = 6, 576 577 /** Number of entries in a data block for code points below the fast limit. 64=0x40 @internal */ 578 UCPTRIE_FAST_DATA_BLOCK_LENGTH = 1 << UCPTRIE_FAST_SHIFT, 579 580 /** Mask for getting the lower bits for the in-fast-data-block offset. @internal */ 581 UCPTRIE_FAST_DATA_MASK = UCPTRIE_FAST_DATA_BLOCK_LENGTH - 1, 582 583 /** @internal */ 584 UCPTRIE_SMALL_MAX = 0xfff, 585 586 /** 587 * Offset from dataLength (to be subtracted) for fetching the 588 * value returned for out-of-range code points and ill-formed UTF-8/16. 589 * @internal 590 */ 591 UCPTRIE_ERROR_VALUE_NEG_DATA_OFFSET = 1, 592 /** 593 * Offset from dataLength (to be subtracted) for fetching the 594 * value returned for code points highStart..U+10FFFF. 595 * @internal 596 */ 597 UCPTRIE_HIGH_VALUE_NEG_DATA_OFFSET = 2 598 }; 599 600 /* Internal functions and macros -------------------------------------------- */ 601 // Do not conditionalize with #ifndef U_HIDE_INTERNAL_API, needed for public API 602 603 /** @internal */ 604 U_INTERNAL int32_t U_EXPORT2 605 ucptrie_internalSmallIndex(const UCPTrie *trie, UChar32 c); 606 607 /** @internal */ 608 U_INTERNAL int32_t U_EXPORT2 609 ucptrie_internalSmallU8Index(const UCPTrie *trie, int32_t lt1, uint8_t t2, uint8_t t3); 610 611 /** 612 * Internal function for part of the UCPTRIE_FAST_U8_PREVxx() macro implementations. 613 * Do not call directly. 614 * @internal 615 */ 616 U_INTERNAL int32_t U_EXPORT2 617 ucptrie_internalU8PrevIndex(const UCPTrie *trie, UChar32 c, 618 const uint8_t *start, const uint8_t *src); 619 620 /** Internal trie getter for a code point below the fast limit. Returns the data index. @internal */ 621 #define _UCPTRIE_FAST_INDEX(trie, c) \ 622 ((int32_t)(trie)->index[(c) >> UCPTRIE_FAST_SHIFT] + ((c) & UCPTRIE_FAST_DATA_MASK)) 623 624 /** Internal trie getter for a code point at or above the fast limit. Returns the data index. @internal */ 625 #define _UCPTRIE_SMALL_INDEX(trie, c) \ 626 ((c) >= (trie)->highStart ? \ 627 (trie)->dataLength - UCPTRIE_HIGH_VALUE_NEG_DATA_OFFSET : \ 628 ucptrie_internalSmallIndex(trie, c)) 629 630 /** 631 * Internal trie getter for a code point, with checking that c is in U+0000..10FFFF. 632 * Returns the data index. 633 * @internal 634 */ 635 #define _UCPTRIE_CP_INDEX(trie, fastMax, c) \ 636 ((uint32_t)(c) <= (uint32_t)(fastMax) ? \ 637 _UCPTRIE_FAST_INDEX(trie, c) : \ 638 (uint32_t)(c) <= 0x10ffff ? \ 639 _UCPTRIE_SMALL_INDEX(trie, c) : \ 640 (trie)->dataLength - UCPTRIE_ERROR_VALUE_NEG_DATA_OFFSET) 641 642 U_CDECL_END 643 644 #endif // U_IN_DOXYGEN 645 #endif // U_HIDE_DRAFT_API 646 #endif 647