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 #include "unicode/ucpmap.h" 12 #include "unicode/utf8.h" 13 14 #if U_SHOW_CPLUSPLUS_API 15 #include "unicode/localpointer.h" 16 #endif // U_SHOW_CPLUSPLUS_API 17 18 U_CDECL_BEGIN 19 20 /** 21 * \file 22 * \brief C API: This file defines an immutable Unicode code point trie. 23 * 24 * @see UCPTrie 25 * @see UMutableCPTrie 26 */ 27 28 #ifndef U_IN_DOXYGEN 29 /** @internal */ 30 typedef union UCPTrieData { 31 /** @internal */ 32 const void *ptr0; 33 /** @internal */ 34 const uint16_t *ptr16; 35 /** @internal */ 36 const uint32_t *ptr32; 37 /** @internal */ 38 const uint8_t *ptr8; 39 } UCPTrieData; 40 #endif 41 42 /** 43 * Immutable Unicode code point trie structure. 44 * Fast, reasonably compact, map from Unicode code points (U+0000..U+10FFFF) to integer values. 45 * For details see https://icu.unicode.org/design/struct/utrie 46 * 47 * Do not access UCPTrie fields directly; use public functions and macros. 48 * Functions are easy to use: They support all trie types and value widths. 49 * 50 * When performance is really important, macros provide faster access. 51 * Most macros are specific to either "fast" or "small" tries, see UCPTrieType. 52 * There are "fast" macros for special optimized use cases. 53 * 54 * The macros will return bogus values, or may crash, if used on the wrong type or value width. 55 * 56 * @see UMutableCPTrie 57 * @stable ICU 63 58 */ 59 struct UCPTrie { 60 #ifndef U_IN_DOXYGEN 61 /** @internal */ 62 const uint16_t *index; 63 /** @internal */ 64 UCPTrieData data; 65 66 /** @internal */ 67 int32_t indexLength; 68 /** @internal */ 69 int32_t dataLength; 70 /** Start of the last range which ends at U+10FFFF. @internal */ 71 UChar32 highStart; 72 /** highStart>>12 @internal */ 73 uint16_t shifted12HighStart; 74 75 /** @internal */ 76 int8_t type; // UCPTrieType 77 /** @internal */ 78 int8_t valueWidth; // UCPTrieValueWidth 79 80 /** padding/reserved @internal */ 81 uint32_t reserved32; 82 /** padding/reserved @internal */ 83 uint16_t reserved16; 84 85 /** 86 * Internal index-3 null block offset. 87 * Set to an impossibly high value (e.g., 0xffff) if there is no dedicated index-3 null block. 88 * @internal 89 */ 90 uint16_t index3NullOffset; 91 /** 92 * Internal data null block offset, not shifted. 93 * Set to an impossibly high value (e.g., 0xfffff) if there is no dedicated data null block. 94 * @internal 95 */ 96 int32_t dataNullOffset; 97 /** @internal */ 98 uint32_t nullValue; 99 100 #ifdef UCPTRIE_DEBUG 101 /** @internal */ 102 const char *name; 103 #endif 104 #endif 105 }; 106 #ifndef U_IN_DOXYGEN 107 typedef struct UCPTrie UCPTrie; 108 #endif 109 110 /** 111 * Selectors for the type of a UCPTrie. 112 * Different trade-offs for size vs. speed. 113 * 114 * @see umutablecptrie_buildImmutable 115 * @see ucptrie_openFromBinary 116 * @see ucptrie_getType 117 * @stable ICU 63 118 */ 119 enum UCPTrieType { 120 /** 121 * For ucptrie_openFromBinary() to accept any type. 122 * ucptrie_getType() will return the actual type. 123 * @stable ICU 63 124 */ 125 UCPTRIE_TYPE_ANY = -1, 126 /** 127 * Fast/simple/larger BMP data structure. Use functions and "fast" macros. 128 * @stable ICU 63 129 */ 130 UCPTRIE_TYPE_FAST, 131 /** 132 * Small/slower BMP data structure. Use functions and "small" macros. 133 * @stable ICU 63 134 */ 135 UCPTRIE_TYPE_SMALL 136 }; 137 #ifndef U_IN_DOXYGEN 138 typedef enum UCPTrieType UCPTrieType; 139 #endif 140 141 /** 142 * Selectors for the number of bits in a UCPTrie data value. 143 * 144 * @see umutablecptrie_buildImmutable 145 * @see ucptrie_openFromBinary 146 * @see ucptrie_getValueWidth 147 * @stable ICU 63 148 */ 149 enum UCPTrieValueWidth { 150 /** 151 * For ucptrie_openFromBinary() to accept any data value width. 152 * ucptrie_getValueWidth() will return the actual data value width. 153 * @stable ICU 63 154 */ 155 UCPTRIE_VALUE_BITS_ANY = -1, 156 /** 157 * The trie stores 16 bits per data value. 158 * It returns them as unsigned values 0..0xffff=65535. 159 * @stable ICU 63 160 */ 161 UCPTRIE_VALUE_BITS_16, 162 /** 163 * The trie stores 32 bits per data value. 164 * @stable ICU 63 165 */ 166 UCPTRIE_VALUE_BITS_32, 167 /** 168 * The trie stores 8 bits per data value. 169 * It returns them as unsigned values 0..0xff=255. 170 * @stable ICU 63 171 */ 172 UCPTRIE_VALUE_BITS_8 173 }; 174 #ifndef U_IN_DOXYGEN 175 typedef enum UCPTrieValueWidth UCPTrieValueWidth; 176 #endif 177 178 /** 179 * Opens a trie from its binary form, stored in 32-bit-aligned memory. 180 * Inverse of ucptrie_toBinary(). 181 * 182 * The memory must remain valid and unchanged as long as the trie is used. 183 * You must ucptrie_close() the trie once you are done using it. 184 * 185 * @param type selects the trie type; results in an 186 * U_INVALID_FORMAT_ERROR if it does not match the binary data; 187 * use UCPTRIE_TYPE_ANY to accept any type 188 * @param valueWidth selects the number of bits in a data value; results in an 189 * U_INVALID_FORMAT_ERROR if it does not match the binary data; 190 * use UCPTRIE_VALUE_BITS_ANY to accept any data value width 191 * @param data a pointer to 32-bit-aligned memory containing the binary data of a UCPTrie 192 * @param length the number of bytes available at data; 193 * can be more than necessary 194 * @param pActualLength receives the actual number of bytes at data taken up by the trie data; 195 * can be NULL 196 * @param pErrorCode an in/out ICU UErrorCode 197 * @return the trie 198 * 199 * @see umutablecptrie_open 200 * @see umutablecptrie_buildImmutable 201 * @see ucptrie_toBinary 202 * @stable ICU 63 203 */ 204 U_CAPI UCPTrie * U_EXPORT2 205 ucptrie_openFromBinary(UCPTrieType type, UCPTrieValueWidth valueWidth, 206 const void *data, int32_t length, int32_t *pActualLength, 207 UErrorCode *pErrorCode); 208 209 /** 210 * Closes a trie and releases associated memory. 211 * 212 * @param trie the trie 213 * @stable ICU 63 214 */ 215 U_CAPI void U_EXPORT2 216 ucptrie_close(UCPTrie *trie); 217 218 /** 219 * Returns the trie type. 220 * 221 * @param trie the trie 222 * @return the trie type 223 * @see ucptrie_openFromBinary 224 * @see UCPTRIE_TYPE_ANY 225 * @stable ICU 63 226 */ 227 U_CAPI UCPTrieType U_EXPORT2 228 ucptrie_getType(const UCPTrie *trie); 229 230 /** 231 * Returns the number of bits in a trie data value. 232 * 233 * @param trie the trie 234 * @return the number of bits in a trie data value 235 * @see ucptrie_openFromBinary 236 * @see UCPTRIE_VALUE_BITS_ANY 237 * @stable ICU 63 238 */ 239 U_CAPI UCPTrieValueWidth U_EXPORT2 240 ucptrie_getValueWidth(const UCPTrie *trie); 241 242 /** 243 * Returns the value for a code point as stored in the trie, with range checking. 244 * Returns the trie error value if c is not in the range 0..U+10FFFF. 245 * 246 * Easier to use than UCPTRIE_FAST_GET() and similar macros but slower. 247 * Easier to use because, unlike the macros, this function works on all UCPTrie 248 * objects, for all types and value widths. 249 * 250 * @param trie the trie 251 * @param c the code point 252 * @return the trie value, 253 * or the trie error value if the code point is not in the range 0..U+10FFFF 254 * @stable ICU 63 255 */ 256 U_CAPI uint32_t U_EXPORT2 257 ucptrie_get(const UCPTrie *trie, UChar32 c); 258 259 /** 260 * Returns the last code point such that all those from start to there have the same value. 261 * Can be used to efficiently iterate over all same-value ranges in a trie. 262 * (This is normally faster than iterating over code points and get()ting each value, 263 * but much slower than a data structure that stores ranges directly.) 264 * 265 * If the UCPMapValueFilter function pointer is not NULL, then 266 * the value to be delivered is passed through that function, and the return value is the end 267 * of the range where all values are modified to the same actual value. 268 * The value is unchanged if that function pointer is NULL. 269 * 270 * Example: 271 * \code 272 * UChar32 start = 0, end; 273 * uint32_t value; 274 * while ((end = ucptrie_getRange(trie, start, UCPMAP_RANGE_NORMAL, 0, 275 * NULL, NULL, &value)) >= 0) { 276 * // Work with the range start..end and its value. 277 * start = end + 1; 278 * } 279 * \endcode 280 * 281 * @param trie the trie 282 * @param start range start 283 * @param option defines whether surrogates are treated normally, 284 * or as having the surrogateValue; usually UCPMAP_RANGE_NORMAL 285 * @param surrogateValue value for surrogates; ignored if option==UCPMAP_RANGE_NORMAL 286 * @param filter a pointer to a function that may modify the trie data value, 287 * or NULL if the values from the trie are to be used unmodified 288 * @param context an opaque pointer that is passed on to the filter function 289 * @param pValue if not NULL, receives the value that every code point start..end has; 290 * may have been modified by filter(context, trie value) 291 * if that function pointer is not NULL 292 * @return the range end code point, or -1 if start is not a valid code point 293 * @stable ICU 63 294 */ 295 U_CAPI UChar32 U_EXPORT2 296 ucptrie_getRange(const UCPTrie *trie, UChar32 start, 297 UCPMapRangeOption option, uint32_t surrogateValue, 298 UCPMapValueFilter *filter, const void *context, uint32_t *pValue); 299 300 /** 301 * Writes a memory-mappable form of the trie into 32-bit aligned memory. 302 * Inverse of ucptrie_openFromBinary(). 303 * 304 * @param trie the trie 305 * @param data a pointer to 32-bit-aligned memory to be filled with the trie data; 306 * can be NULL if capacity==0 307 * @param capacity the number of bytes available at data, or 0 for pure preflighting 308 * @param pErrorCode an in/out ICU UErrorCode; 309 * U_BUFFER_OVERFLOW_ERROR if the capacity is too small 310 * @return the number of bytes written or (if buffer overflow) needed for the trie 311 * 312 * @see ucptrie_openFromBinary() 313 * @stable ICU 63 314 */ 315 U_CAPI int32_t U_EXPORT2 316 ucptrie_toBinary(const UCPTrie *trie, void *data, int32_t capacity, UErrorCode *pErrorCode); 317 318 /** 319 * Macro parameter value for a trie with 16-bit data values. 320 * Use the name of this macro as a "dataAccess" parameter in other macros. 321 * Do not use this macro in any other way. 322 * 323 * @see UCPTRIE_VALUE_BITS_16 324 * @stable ICU 63 325 */ 326 #define UCPTRIE_16(trie, i) ((trie)->data.ptr16[i]) 327 328 /** 329 * Macro parameter value for a trie with 32-bit data values. 330 * Use the name of this macro as a "dataAccess" parameter in other macros. 331 * Do not use this macro in any other way. 332 * 333 * @see UCPTRIE_VALUE_BITS_32 334 * @stable ICU 63 335 */ 336 #define UCPTRIE_32(trie, i) ((trie)->data.ptr32[i]) 337 338 /** 339 * Macro parameter value for a trie with 8-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_8 344 * @stable ICU 63 345 */ 346 #define UCPTRIE_8(trie, i) ((trie)->data.ptr8[i]) 347 348 /** 349 * Returns a trie value for a code point, with range checking. 350 * Returns the trie error value if c is not in the range 0..U+10FFFF. 351 * 352 * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST 353 * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the trie’s value width 354 * @param c (UChar32, in) the input code point 355 * @return The code point's trie value. 356 * @stable ICU 63 357 */ 358 #define UCPTRIE_FAST_GET(trie, dataAccess, c) dataAccess(trie, _UCPTRIE_CP_INDEX(trie, 0xffff, c)) 359 360 /** 361 * Returns a 16-bit trie value for a code point, with range checking. 362 * Returns the trie error value if c is not in the range U+0000..U+10FFFF. 363 * 364 * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_SMALL 365 * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the trie’s value width 366 * @param c (UChar32, in) the input code point 367 * @return The code point's trie value. 368 * @stable ICU 63 369 */ 370 #define UCPTRIE_SMALL_GET(trie, dataAccess, c) \ 371 dataAccess(trie, _UCPTRIE_CP_INDEX(trie, UCPTRIE_SMALL_MAX, c)) 372 373 /** 374 * UTF-16: Reads the next code point (UChar32 c, out), post-increments src, 375 * and gets a value from the trie. 376 * Sets the trie error value if c is an unpaired surrogate. 377 * 378 * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST 379 * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the trie’s value width 380 * @param src (const UChar *, in/out) the source text pointer 381 * @param limit (const UChar *, in) the limit pointer for the text, or NULL if NUL-terminated 382 * @param c (UChar32, out) variable for the code point 383 * @param result (out) variable for the trie lookup result 384 * @stable ICU 63 385 */ 386 #define UCPTRIE_FAST_U16_NEXT(trie, dataAccess, src, limit, c, result) UPRV_BLOCK_MACRO_BEGIN { \ 387 (c) = *(src)++; \ 388 int32_t __index; \ 389 if (!U16_IS_SURROGATE(c)) { \ 390 __index = _UCPTRIE_FAST_INDEX(trie, c); \ 391 } else { \ 392 uint16_t __c2; \ 393 if (U16_IS_SURROGATE_LEAD(c) && (src) != (limit) && U16_IS_TRAIL(__c2 = *(src))) { \ 394 ++(src); \ 395 (c) = U16_GET_SUPPLEMENTARY((c), __c2); \ 396 __index = _UCPTRIE_SMALL_INDEX(trie, c); \ 397 } else { \ 398 __index = (trie)->dataLength - UCPTRIE_ERROR_VALUE_NEG_DATA_OFFSET; \ 399 } \ 400 } \ 401 (result) = dataAccess(trie, __index); \ 402 } UPRV_BLOCK_MACRO_END 403 404 /** 405 * UTF-16: Reads the previous code point (UChar32 c, out), pre-decrements src, 406 * and gets a value from the trie. 407 * Sets the trie error value if c is an unpaired surrogate. 408 * 409 * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST 410 * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the trie’s value width 411 * @param start (const UChar *, in) the start pointer for the text 412 * @param src (const UChar *, in/out) the source text pointer 413 * @param c (UChar32, out) variable for the code point 414 * @param result (out) variable for the trie lookup result 415 * @stable ICU 63 416 */ 417 #define UCPTRIE_FAST_U16_PREV(trie, dataAccess, start, src, c, result) UPRV_BLOCK_MACRO_BEGIN { \ 418 (c) = *--(src); \ 419 int32_t __index; \ 420 if (!U16_IS_SURROGATE(c)) { \ 421 __index = _UCPTRIE_FAST_INDEX(trie, c); \ 422 } else { \ 423 uint16_t __c2; \ 424 if (U16_IS_SURROGATE_TRAIL(c) && (src) != (start) && U16_IS_LEAD(__c2 = *((src) - 1))) { \ 425 --(src); \ 426 (c) = U16_GET_SUPPLEMENTARY(__c2, (c)); \ 427 __index = _UCPTRIE_SMALL_INDEX(trie, c); \ 428 } else { \ 429 __index = (trie)->dataLength - UCPTRIE_ERROR_VALUE_NEG_DATA_OFFSET; \ 430 } \ 431 } \ 432 (result) = dataAccess(trie, __index); \ 433 } UPRV_BLOCK_MACRO_END 434 435 /** 436 * UTF-8: Post-increments src and gets a value from the trie. 437 * Sets the trie error value for an ill-formed byte sequence. 438 * 439 * Unlike UCPTRIE_FAST_U16_NEXT() this UTF-8 macro does not provide the code point 440 * because it would be more work to do so and is often not needed. 441 * If the trie value differs from the error value, then the byte sequence is well-formed, 442 * and the code point can be assembled without revalidation. 443 * 444 * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST 445 * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the trie’s value width 446 * @param src (const char *, in/out) the source text pointer 447 * @param limit (const char *, in) the limit pointer for the text (must not be NULL) 448 * @param result (out) variable for the trie lookup result 449 * @stable ICU 63 450 */ 451 #define UCPTRIE_FAST_U8_NEXT(trie, dataAccess, src, limit, result) UPRV_BLOCK_MACRO_BEGIN { \ 452 int32_t __lead = (uint8_t)*(src)++; \ 453 if (!U8_IS_SINGLE(__lead)) { \ 454 uint8_t __t1, __t2, __t3; \ 455 if ((src) != (limit) && \ 456 (__lead >= 0xe0 ? \ 457 __lead < 0xf0 ? /* U+0800..U+FFFF except surrogates */ \ 458 U8_LEAD3_T1_BITS[__lead &= 0xf] & (1 << ((__t1 = *(src)) >> 5)) && \ 459 ++(src) != (limit) && (__t2 = *(src) - 0x80) <= 0x3f && \ 460 (__lead = ((int32_t)(trie)->index[(__lead << 6) + (__t1 & 0x3f)]) + __t2, 1) \ 461 : /* U+10000..U+10FFFF */ \ 462 (__lead -= 0xf0) <= 4 && \ 463 U8_LEAD4_T1_BITS[(__t1 = *(src)) >> 4] & (1 << __lead) && \ 464 (__lead = (__lead << 6) | (__t1 & 0x3f), ++(src) != (limit)) && \ 465 (__t2 = *(src) - 0x80) <= 0x3f && \ 466 ++(src) != (limit) && (__t3 = *(src) - 0x80) <= 0x3f && \ 467 (__lead = __lead >= (trie)->shifted12HighStart ? \ 468 (trie)->dataLength - UCPTRIE_HIGH_VALUE_NEG_DATA_OFFSET : \ 469 ucptrie_internalSmallU8Index((trie), __lead, __t2, __t3), 1) \ 470 : /* U+0080..U+07FF */ \ 471 __lead >= 0xc2 && (__t1 = *(src) - 0x80) <= 0x3f && \ 472 (__lead = (int32_t)(trie)->index[__lead & 0x1f] + __t1, 1))) { \ 473 ++(src); \ 474 } else { \ 475 __lead = (trie)->dataLength - UCPTRIE_ERROR_VALUE_NEG_DATA_OFFSET; /* ill-formed*/ \ 476 } \ 477 } \ 478 (result) = dataAccess(trie, __lead); \ 479 } UPRV_BLOCK_MACRO_END 480 481 /** 482 * UTF-8: Pre-decrements src and gets a value from the trie. 483 * Sets the trie error value for an ill-formed byte sequence. 484 * 485 * Unlike UCPTRIE_FAST_U16_PREV() this UTF-8 macro does not provide the code point 486 * because it would be more work to do so and is often not needed. 487 * If the trie value differs from the error value, then the byte sequence is well-formed, 488 * and the code point can be assembled without revalidation. 489 * 490 * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST 491 * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the trie’s value width 492 * @param start (const char *, in) the start pointer for the text 493 * @param src (const char *, in/out) the source text pointer 494 * @param result (out) variable for the trie lookup result 495 * @stable ICU 63 496 */ 497 #define UCPTRIE_FAST_U8_PREV(trie, dataAccess, start, src, result) UPRV_BLOCK_MACRO_BEGIN { \ 498 int32_t __index = (uint8_t)*--(src); \ 499 if (!U8_IS_SINGLE(__index)) { \ 500 __index = ucptrie_internalU8PrevIndex((trie), __index, (const uint8_t *)(start), \ 501 (const uint8_t *)(src)); \ 502 (src) -= __index & 7; \ 503 __index >>= 3; \ 504 } \ 505 (result) = dataAccess(trie, __index); \ 506 } UPRV_BLOCK_MACRO_END 507 508 /** 509 * Returns a trie value for an ASCII code point, without range checking. 510 * 511 * @param trie (const UCPTrie *, in) the trie (of either fast or small type) 512 * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the trie’s value width 513 * @param c (UChar32, in) the input code point; must be U+0000..U+007F 514 * @return The ASCII code point's trie value. 515 * @stable ICU 63 516 */ 517 #define UCPTRIE_ASCII_GET(trie, dataAccess, c) dataAccess(trie, c) 518 519 /** 520 * Returns a trie value for a BMP code point (U+0000..U+FFFF), without range checking. 521 * Can be used to look up a value for a UTF-16 code unit if other parts of 522 * the string processing check for surrogates. 523 * 524 * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST 525 * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the trie’s value width 526 * @param c (UChar32, in) the input code point, must be U+0000..U+FFFF 527 * @return The BMP code point's trie value. 528 * @stable ICU 63 529 */ 530 #define UCPTRIE_FAST_BMP_GET(trie, dataAccess, c) dataAccess(trie, _UCPTRIE_FAST_INDEX(trie, c)) 531 532 /** 533 * Returns a trie value for a supplementary code point (U+10000..U+10FFFF), 534 * without range checking. 535 * 536 * @param trie (const UCPTrie *, in) the trie; must have type UCPTRIE_TYPE_FAST 537 * @param dataAccess UCPTRIE_16, UCPTRIE_32, or UCPTRIE_8 according to the trie’s value width 538 * @param c (UChar32, in) the input code point, must be U+10000..U+10FFFF 539 * @return The supplementary code point's trie value. 540 * @stable ICU 63 541 */ 542 #define UCPTRIE_FAST_SUPP_GET(trie, dataAccess, c) dataAccess(trie, _UCPTRIE_SMALL_INDEX(trie, c)) 543 544 /* Internal definitions ----------------------------------------------------- */ 545 546 #ifndef U_IN_DOXYGEN 547 548 /** 549 * Internal implementation constants. 550 * These are needed for the API macros, but users should not use these directly. 551 * @internal 552 */ 553 enum { 554 /** @internal */ 555 UCPTRIE_FAST_SHIFT = 6, 556 557 /** Number of entries in a data block for code points below the fast limit. 64=0x40 @internal */ 558 UCPTRIE_FAST_DATA_BLOCK_LENGTH = 1 << UCPTRIE_FAST_SHIFT, 559 560 /** Mask for getting the lower bits for the in-fast-data-block offset. @internal */ 561 UCPTRIE_FAST_DATA_MASK = UCPTRIE_FAST_DATA_BLOCK_LENGTH - 1, 562 563 /** @internal */ 564 UCPTRIE_SMALL_MAX = 0xfff, 565 566 /** 567 * Offset from dataLength (to be subtracted) for fetching the 568 * value returned for out-of-range code points and ill-formed UTF-8/16. 569 * @internal 570 */ 571 UCPTRIE_ERROR_VALUE_NEG_DATA_OFFSET = 1, 572 /** 573 * Offset from dataLength (to be subtracted) for fetching the 574 * value returned for code points highStart..U+10FFFF. 575 * @internal 576 */ 577 UCPTRIE_HIGH_VALUE_NEG_DATA_OFFSET = 2 578 }; 579 580 /* Internal functions and macros -------------------------------------------- */ 581 // Do not conditionalize with #ifndef U_HIDE_INTERNAL_API, needed for public API 582 583 /** @internal */ 584 U_CAPI int32_t U_EXPORT2 585 ucptrie_internalSmallIndex(const UCPTrie *trie, UChar32 c); 586 587 /** @internal */ 588 U_CAPI int32_t U_EXPORT2 589 ucptrie_internalSmallU8Index(const UCPTrie *trie, int32_t lt1, uint8_t t2, uint8_t t3); 590 591 /** 592 * Internal function for part of the UCPTRIE_FAST_U8_PREVxx() macro implementations. 593 * Do not call directly. 594 * @internal 595 */ 596 U_CAPI int32_t U_EXPORT2 597 ucptrie_internalU8PrevIndex(const UCPTrie *trie, UChar32 c, 598 const uint8_t *start, const uint8_t *src); 599 600 /** Internal trie getter for a code point below the fast limit. Returns the data index. @internal */ 601 #define _UCPTRIE_FAST_INDEX(trie, c) \ 602 ((int32_t)(trie)->index[(c) >> UCPTRIE_FAST_SHIFT] + ((c) & UCPTRIE_FAST_DATA_MASK)) 603 604 /** Internal trie getter for a code point at or above the fast limit. Returns the data index. @internal */ 605 #define _UCPTRIE_SMALL_INDEX(trie, c) \ 606 ((c) >= (trie)->highStart ? \ 607 (trie)->dataLength - UCPTRIE_HIGH_VALUE_NEG_DATA_OFFSET : \ 608 ucptrie_internalSmallIndex(trie, c)) 609 610 /** 611 * Internal trie getter for a code point, with checking that c is in U+0000..10FFFF. 612 * Returns the data index. 613 * @internal 614 */ 615 #define _UCPTRIE_CP_INDEX(trie, fastMax, c) \ 616 ((uint32_t)(c) <= (uint32_t)(fastMax) ? \ 617 _UCPTRIE_FAST_INDEX(trie, c) : \ 618 (uint32_t)(c) <= 0x10ffff ? \ 619 _UCPTRIE_SMALL_INDEX(trie, c) : \ 620 (trie)->dataLength - UCPTRIE_ERROR_VALUE_NEG_DATA_OFFSET) 621 622 U_CDECL_END 623 624 #endif // U_IN_DOXYGEN 625 626 #if U_SHOW_CPLUSPLUS_API 627 628 U_NAMESPACE_BEGIN 629 630 /** 631 * \class LocalUCPTriePointer 632 * "Smart pointer" class, closes a UCPTrie via ucptrie_close(). 633 * For most methods see the LocalPointerBase base class. 634 * 635 * @see LocalPointerBase 636 * @see LocalPointer 637 * @stable ICU 63 638 */ 639 U_DEFINE_LOCAL_OPEN_POINTER(LocalUCPTriePointer, UCPTrie, ucptrie_close); 640 641 U_NAMESPACE_END 642 643 #endif // U_SHOW_CPLUSPLUS_API 644 645 #endif 646