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) 2001-2014, International Business Machines |
7 | * Corporation and others. All Rights Reserved. |
8 | * |
9 | ****************************************************************************** |
10 | * file name: utrie2.h |
11 | * encoding: UTF-8 |
12 | * tab size: 8 (not used) |
13 | * indentation:4 |
14 | * |
15 | * created on: 2008aug16 (starting from a copy of utrie.h) |
16 | * created by: Markus W. Scherer |
17 | */ |
18 | |
19 | #ifndef __UTRIE2_H__ |
20 | #define __UTRIE2_H__ |
21 | |
22 | #include "unicode/utypes.h" |
23 | #include "unicode/utf8.h" |
24 | #include "putilimp.h" |
25 | |
26 | U_CDECL_BEGIN |
27 | |
28 | struct UTrie; /* forward declaration */ |
29 | #ifndef __UTRIE_H__ |
30 | typedef struct UTrie UTrie; |
31 | #endif |
32 | |
33 | /** |
34 | * \file |
35 | * |
36 | * This is a common implementation of a Unicode trie. |
37 | * It is a kind of compressed, serializable table of 16- or 32-bit values associated with |
38 | * Unicode code points (0..0x10ffff). (A map from code points to integers.) |
39 | * |
40 | * This is the second common version of a Unicode trie (hence the name UTrie2). |
41 | * Compared with UTrie version 1: |
42 | * - Still splitting BMP code points 11:5 bits for index and data table lookups. |
43 | * - Still separate data for lead surrogate code _units_ vs. code _points_, |
44 | * but the lead surrogate code unit values are not required any more |
45 | * for data lookup for supplementary code points. |
46 | * - The "folding" mechanism is removed. In UTrie version 1, this somewhat |
47 | * hard-to-explain mechanism was meant to be used for optimized UTF-16 |
48 | * processing, with application-specific encoding of indexing bits |
49 | * in the lead surrogate data for the associated supplementary code points. |
50 | * - For the last single-value code point range (ending with U+10ffff), |
51 | * the starting code point ("highStart") and the value are stored. |
52 | * - For supplementary code points U+10000..highStart-1 a three-table lookup |
53 | * (two index tables and one data table) is used. The first index |
54 | * is truncated, omitting both the BMP portion and the high range. |
55 | * - There is a special small index for 2-byte UTF-8, and the initial data |
56 | * entries are designed for fast 1/2-byte UTF-8 lookup. |
57 | * Starting with ICU 60, C0 and C1 are not recognized as UTF-8 lead bytes any more at all, |
58 | * and the associated 2-byte indexes are unused. |
59 | */ |
60 | |
61 | /** |
62 | * Trie structure. |
63 | * Use only with public API macros and functions. |
64 | */ |
65 | struct UTrie2; |
66 | typedef struct UTrie2 UTrie2; |
67 | |
68 | /* Public UTrie2 API functions: read-only access ---------------------------- */ |
69 | |
70 | /** |
71 | * Selectors for the width of a UTrie2 data value. |
72 | */ |
73 | enum UTrie2ValueBits { |
74 | /** 16 bits per UTrie2 data value. */ |
75 | UTRIE2_16_VALUE_BITS, |
76 | /** 32 bits per UTrie2 data value. */ |
77 | UTRIE2_32_VALUE_BITS, |
78 | /** Number of selectors for the width of UTrie2 data values. */ |
79 | UTRIE2_COUNT_VALUE_BITS |
80 | }; |
81 | typedef enum UTrie2ValueBits UTrie2ValueBits; |
82 | |
83 | /** |
84 | * Open a frozen trie from its serialized from, stored in 32-bit-aligned memory. |
85 | * Inverse of utrie2_serialize(). |
86 | * The memory must remain valid and unchanged as long as the trie is used. |
87 | * You must utrie2_close() the trie once you are done using it. |
88 | * |
89 | * @param valueBits selects the data entry size; results in an |
90 | * U_INVALID_FORMAT_ERROR if it does not match the serialized form |
91 | * @param data a pointer to 32-bit-aligned memory containing the serialized form of a UTrie2 |
92 | * @param length the number of bytes available at data; |
93 | * can be more than necessary |
94 | * @param pActualLength receives the actual number of bytes at data taken up by the trie data; |
95 | * can be NULL |
96 | * @param pErrorCode an in/out ICU UErrorCode |
97 | * @return the unserialized trie |
98 | * |
99 | * @see utrie2_open |
100 | * @see utrie2_serialize |
101 | */ |
102 | U_CAPI UTrie2 * U_EXPORT2 |
103 | utrie2_openFromSerialized(UTrie2ValueBits valueBits, |
104 | const void *data, int32_t length, int32_t *pActualLength, |
105 | UErrorCode *pErrorCode); |
106 | |
107 | /** |
108 | * Open a frozen, empty "dummy" trie. |
109 | * A dummy trie is an empty trie, used when a real data trie cannot |
110 | * be loaded. Equivalent to calling utrie2_open() and utrie2_freeze(), |
111 | * but without internally creating and compacting/serializing the |
112 | * builder data structure. |
113 | * |
114 | * The trie always returns the initialValue, |
115 | * or the errorValue for out-of-range code points and illegal UTF-8. |
116 | * |
117 | * You must utrie2_close() the trie once you are done using it. |
118 | * |
119 | * @param valueBits selects the data entry size |
120 | * @param initialValue the initial value that is set for all code points |
121 | * @param errorValue the value for out-of-range code points and illegal UTF-8 |
122 | * @param pErrorCode an in/out ICU UErrorCode |
123 | * @return the dummy trie |
124 | * |
125 | * @see utrie2_openFromSerialized |
126 | * @see utrie2_open |
127 | */ |
128 | U_CAPI UTrie2 * U_EXPORT2 |
129 | utrie2_openDummy(UTrie2ValueBits valueBits, |
130 | uint32_t initialValue, uint32_t errorValue, |
131 | UErrorCode *pErrorCode); |
132 | |
133 | /** |
134 | * Get a value from a code point as stored in the trie. |
135 | * Easier to use than UTRIE2_GET16() and UTRIE2_GET32() but slower. |
136 | * Easier to use because, unlike the macros, this function works on all UTrie2 |
137 | * objects, frozen or not, holding 16-bit or 32-bit data values. |
138 | * |
139 | * @param trie the trie |
140 | * @param c the code point |
141 | * @return the value |
142 | */ |
143 | U_CAPI uint32_t U_EXPORT2 |
144 | utrie2_get32(const UTrie2 *trie, UChar32 c); |
145 | |
146 | /* enumeration callback types */ |
147 | |
148 | /** |
149 | * Callback from utrie2_enum(), extracts a uint32_t value from a |
150 | * trie value. This value will be passed on to the UTrie2EnumRange function. |
151 | * |
152 | * @param context an opaque pointer, as passed into utrie2_enum() |
153 | * @param value a value from the trie |
154 | * @return the value that is to be passed on to the UTrie2EnumRange function |
155 | */ |
156 | typedef uint32_t U_CALLCONV |
157 | UTrie2EnumValue(const void *context, uint32_t value); |
158 | |
159 | /** |
160 | * Callback from utrie2_enum(), is called for each contiguous range |
161 | * of code points with the same value as retrieved from the trie and |
162 | * transformed by the UTrie2EnumValue function. |
163 | * |
164 | * The callback function can stop the enumeration by returning false. |
165 | * |
166 | * @param context an opaque pointer, as passed into utrie2_enum() |
167 | * @param start the first code point in a contiguous range with value |
168 | * @param end the last code point in a contiguous range with value (inclusive) |
169 | * @param value the value that is set for all code points in [start..end] |
170 | * @return false to stop the enumeration |
171 | */ |
172 | typedef UBool U_CALLCONV |
173 | UTrie2EnumRange(const void *context, UChar32 start, UChar32 end, uint32_t value); |
174 | |
175 | /** |
176 | * Enumerate efficiently all values in a trie. |
177 | * Do not modify the trie during the enumeration. |
178 | * |
179 | * For each entry in the trie, the value to be delivered is passed through |
180 | * the UTrie2EnumValue function. |
181 | * The value is unchanged if that function pointer is NULL. |
182 | * |
183 | * For each contiguous range of code points with a given (transformed) value, |
184 | * the UTrie2EnumRange function is called. |
185 | * |
186 | * @param trie a pointer to the trie |
187 | * @param enumValue a pointer to a function that may transform the trie entry value, |
188 | * or NULL if the values from the trie are to be used directly |
189 | * @param enumRange a pointer to a function that is called for each contiguous range |
190 | * of code points with the same (transformed) value |
191 | * @param context an opaque pointer that is passed on to the callback functions |
192 | */ |
193 | U_CAPI void U_EXPORT2 |
194 | utrie2_enum(const UTrie2 *trie, |
195 | UTrie2EnumValue *enumValue, UTrie2EnumRange *enumRange, const void *context); |
196 | |
197 | /* Building a trie ---------------------------------------------------------- */ |
198 | |
199 | /** |
200 | * Open an empty, writable trie. At build time, 32-bit data values are used. |
201 | * utrie2_freeze() takes a valueBits parameter |
202 | * which determines the data value width in the serialized and frozen forms. |
203 | * You must utrie2_close() the trie once you are done using it. |
204 | * |
205 | * @param initialValue the initial value that is set for all code points |
206 | * @param errorValue the value for out-of-range code points and illegal UTF-8 |
207 | * @param pErrorCode an in/out ICU UErrorCode |
208 | * @return a pointer to the allocated and initialized new trie |
209 | */ |
210 | U_CAPI UTrie2 * U_EXPORT2 |
211 | utrie2_open(uint32_t initialValue, uint32_t errorValue, UErrorCode *pErrorCode); |
212 | |
213 | /** |
214 | * Clone a trie. |
215 | * You must utrie2_close() the clone once you are done using it. |
216 | * |
217 | * @param other the trie to clone |
218 | * @param pErrorCode an in/out ICU UErrorCode |
219 | * @return a pointer to the new trie clone |
220 | */ |
221 | U_CAPI UTrie2 * U_EXPORT2 |
222 | utrie2_clone(const UTrie2 *other, UErrorCode *pErrorCode); |
223 | |
224 | /** |
225 | * Clone a trie. The clone will be mutable/writable even if the other trie |
226 | * is frozen. (See utrie2_freeze().) |
227 | * You must utrie2_close() the clone once you are done using it. |
228 | * |
229 | * @param other the trie to clone |
230 | * @param pErrorCode an in/out ICU UErrorCode |
231 | * @return a pointer to the new trie clone |
232 | */ |
233 | U_CAPI UTrie2 * U_EXPORT2 |
234 | utrie2_cloneAsThawed(const UTrie2 *other, UErrorCode *pErrorCode); |
235 | |
236 | /** |
237 | * Close a trie and release associated memory. |
238 | * |
239 | * @param trie the trie |
240 | */ |
241 | U_CAPI void U_EXPORT2 |
242 | utrie2_close(UTrie2 *trie); |
243 | |
244 | /** |
245 | * Set a value for a code point. |
246 | * |
247 | * @param trie the unfrozen trie |
248 | * @param c the code point |
249 | * @param value the value |
250 | * @param pErrorCode an in/out ICU UErrorCode; among other possible error codes: |
251 | * - U_NO_WRITE_PERMISSION if the trie is frozen |
252 | */ |
253 | U_CAPI void U_EXPORT2 |
254 | utrie2_set32(UTrie2 *trie, UChar32 c, uint32_t value, UErrorCode *pErrorCode); |
255 | |
256 | /** |
257 | * Set a value in a range of code points [start..end]. |
258 | * All code points c with start<=c<=end will get the value if |
259 | * overwrite is true or if the old value is the initial value. |
260 | * |
261 | * @param trie the unfrozen trie |
262 | * @param start the first code point to get the value |
263 | * @param end the last code point to get the value (inclusive) |
264 | * @param value the value |
265 | * @param overwrite flag for whether old non-initial values are to be overwritten |
266 | * @param pErrorCode an in/out ICU UErrorCode; among other possible error codes: |
267 | * - U_NO_WRITE_PERMISSION if the trie is frozen |
268 | */ |
269 | U_CAPI void U_EXPORT2 |
270 | utrie2_setRange32(UTrie2 *trie, |
271 | UChar32 start, UChar32 end, |
272 | uint32_t value, UBool overwrite, |
273 | UErrorCode *pErrorCode); |
274 | |
275 | /** |
276 | * Freeze a trie. Make it immutable (read-only) and compact it, |
277 | * ready for serialization and for use with fast macros. |
278 | * Functions to set values will fail after serializing. |
279 | * |
280 | * A trie can be frozen only once. If this function is called again with different |
281 | * valueBits then it will set a U_ILLEGAL_ARGUMENT_ERROR. |
282 | * |
283 | * @param trie the trie |
284 | * @param valueBits selects the data entry size; if smaller than 32 bits, then |
285 | * the values stored in the trie will be truncated |
286 | * @param pErrorCode an in/out ICU UErrorCode; among other possible error codes: |
287 | * - U_INDEX_OUTOFBOUNDS_ERROR if the compacted index or data arrays are too long |
288 | * for serialization |
289 | * (the trie will be immutable and usable, |
290 | * but not frozen and not usable with the fast macros) |
291 | * |
292 | * @see utrie2_cloneAsThawed |
293 | */ |
294 | U_CAPI void U_EXPORT2 |
295 | utrie2_freeze(UTrie2 *trie, UTrie2ValueBits valueBits, UErrorCode *pErrorCode); |
296 | |
297 | /** |
298 | * Test if the trie is frozen. (See utrie2_freeze().) |
299 | * |
300 | * @param trie the trie |
301 | * @return true if the trie is frozen, that is, immutable, ready for serialization |
302 | * and for use with fast macros |
303 | */ |
304 | U_CAPI UBool U_EXPORT2 |
305 | utrie2_isFrozen(const UTrie2 *trie); |
306 | |
307 | /** |
308 | * Serialize a frozen trie into 32-bit aligned memory. |
309 | * If the trie is not frozen, then the function returns with a U_ILLEGAL_ARGUMENT_ERROR. |
310 | * A trie can be serialized multiple times. |
311 | * |
312 | * @param trie the frozen trie |
313 | * @param data a pointer to 32-bit-aligned memory to be filled with the trie data, |
314 | * can be NULL if capacity==0 |
315 | * @param capacity the number of bytes available at data, |
316 | * or 0 for preflighting |
317 | * @param pErrorCode an in/out ICU UErrorCode; among other possible error codes: |
318 | * - U_BUFFER_OVERFLOW_ERROR if the data storage block is too small for serialization |
319 | * - U_ILLEGAL_ARGUMENT_ERROR if the trie is not frozen or the data and capacity |
320 | * parameters are bad |
321 | * @return the number of bytes written or needed for the trie |
322 | * |
323 | * @see utrie2_openFromSerialized() |
324 | */ |
325 | U_CAPI int32_t U_EXPORT2 |
326 | utrie2_serialize(const UTrie2 *trie, |
327 | void *data, int32_t capacity, |
328 | UErrorCode *pErrorCode); |
329 | |
330 | /* Public UTrie2 API: miscellaneous functions ------------------------------- */ |
331 | |
332 | /** |
333 | * Build a UTrie2 (version 2) from a UTrie (version 1). |
334 | * Enumerates all values in the UTrie and builds a UTrie2 with the same values. |
335 | * The resulting UTrie2 will be frozen. |
336 | * |
337 | * @param trie1 the runtime UTrie structure to be enumerated |
338 | * @param errorValue the value for out-of-range code points and illegal UTF-8 |
339 | * @param pErrorCode an in/out ICU UErrorCode |
340 | * @return The frozen UTrie2 with the same values as the UTrie. |
341 | */ |
342 | U_CAPI UTrie2 * U_EXPORT2 |
343 | utrie2_fromUTrie(const UTrie *trie1, uint32_t errorValue, UErrorCode *pErrorCode); |
344 | |
345 | /* Public UTrie2 API macros ------------------------------------------------- */ |
346 | |
347 | /* |
348 | * These macros provide fast data lookup from a frozen trie. |
349 | * They will crash when used on an unfrozen trie. |
350 | */ |
351 | |
352 | /** |
353 | * Return a 16-bit trie value from a code point, with range checking. |
354 | * Returns trie->errorValue if c is not in the range 0..U+10ffff. |
355 | * |
356 | * @param trie (const UTrie2 *, in) a frozen trie |
357 | * @param c (UChar32, in) the input code point |
358 | * @return (uint16_t) The code point's trie value. |
359 | */ |
360 | #define UTRIE2_GET16(trie, c) _UTRIE2_GET((trie), index, (trie)->indexLength, (c)) |
361 | |
362 | /** |
363 | * Return a 32-bit trie value from a code point, with range checking. |
364 | * Returns trie->errorValue if c is not in the range 0..U+10ffff. |
365 | * |
366 | * @param trie (const UTrie2 *, in) a frozen trie |
367 | * @param c (UChar32, in) the input code point |
368 | * @return (uint32_t) The code point's trie value. |
369 | */ |
370 | #define UTRIE2_GET32(trie, c) _UTRIE2_GET((trie), data32, 0, (c)) |
371 | |
372 | /** |
373 | * UTF-16: Get the next code point (UChar32 c, out), post-increment src, |
374 | * and get a 16-bit value from the trie. |
375 | * |
376 | * @param trie (const UTrie2 *, in) a frozen trie |
377 | * @param src (const UChar *, in/out) the source text pointer |
378 | * @param limit (const UChar *, in) the limit pointer for the text, or NULL if NUL-terminated |
379 | * @param c (UChar32, out) variable for the code point |
380 | * @param result (uint16_t, out) uint16_t variable for the trie lookup result |
381 | */ |
382 | #define UTRIE2_U16_NEXT16(trie, src, limit, c, result) _UTRIE2_U16_NEXT(trie, index, src, limit, c, result) |
383 | |
384 | /** |
385 | * UTF-16: Get the next code point (UChar32 c, out), post-increment src, |
386 | * and get a 32-bit value from the trie. |
387 | * |
388 | * @param trie (const UTrie2 *, in) a frozen trie |
389 | * @param src (const UChar *, in/out) the source text pointer |
390 | * @param limit (const UChar *, in) the limit pointer for the text, or NULL if NUL-terminated |
391 | * @param c (UChar32, out) variable for the code point |
392 | * @param result (uint32_t, out) uint32_t variable for the trie lookup result |
393 | */ |
394 | #define UTRIE2_U16_NEXT32(trie, src, limit, c, result) _UTRIE2_U16_NEXT(trie, data32, src, limit, c, result) |
395 | |
396 | /** |
397 | * UTF-16: Get the previous code point (UChar32 c, out), pre-decrement src, |
398 | * and get a 16-bit value from the trie. |
399 | * |
400 | * @param trie (const UTrie2 *, in) a frozen trie |
401 | * @param start (const UChar *, in) the start pointer for the text |
402 | * @param src (const UChar *, in/out) the source text pointer |
403 | * @param c (UChar32, out) variable for the code point |
404 | * @param result (uint16_t, out) uint16_t variable for the trie lookup result |
405 | */ |
406 | #define UTRIE2_U16_PREV16(trie, start, src, c, result) _UTRIE2_U16_PREV(trie, index, start, src, c, result) |
407 | |
408 | /** |
409 | * UTF-16: Get the previous code point (UChar32 c, out), pre-decrement src, |
410 | * and get a 32-bit value from the trie. |
411 | * |
412 | * @param trie (const UTrie2 *, in) a frozen trie |
413 | * @param start (const UChar *, in) the start pointer for the text |
414 | * @param src (const UChar *, in/out) the source text pointer |
415 | * @param c (UChar32, out) variable for the code point |
416 | * @param result (uint32_t, out) uint32_t variable for the trie lookup result |
417 | */ |
418 | #define UTRIE2_U16_PREV32(trie, start, src, c, result) _UTRIE2_U16_PREV(trie, data32, start, src, c, result) |
419 | |
420 | /** |
421 | * UTF-8: Post-increment src and get a 16-bit value from the trie. |
422 | * |
423 | * @param trie (const UTrie2 *, in) a frozen trie |
424 | * @param src (const char *, in/out) the source text pointer |
425 | * @param limit (const char *, in) the limit pointer for the text (must not be NULL) |
426 | * @param result (uint16_t, out) uint16_t variable for the trie lookup result |
427 | */ |
428 | #define UTRIE2_U8_NEXT16(trie, src, limit, result)\ |
429 | _UTRIE2_U8_NEXT(trie, data16, index, src, limit, result) |
430 | |
431 | /** |
432 | * UTF-8: Post-increment src and get a 32-bit value from the trie. |
433 | * |
434 | * @param trie (const UTrie2 *, in) a frozen trie |
435 | * @param src (const char *, in/out) the source text pointer |
436 | * @param limit (const char *, in) the limit pointer for the text (must not be NULL) |
437 | * @param result (uint16_t, out) uint32_t variable for the trie lookup result |
438 | */ |
439 | #define UTRIE2_U8_NEXT32(trie, src, limit, result) \ |
440 | _UTRIE2_U8_NEXT(trie, data32, data32, src, limit, result) |
441 | |
442 | /** |
443 | * UTF-8: Pre-decrement src and get a 16-bit value from the trie. |
444 | * |
445 | * @param trie (const UTrie2 *, in) a frozen trie |
446 | * @param start (const char *, in) the start pointer for the text |
447 | * @param src (const char *, in/out) the source text pointer |
448 | * @param result (uint16_t, out) uint16_t variable for the trie lookup result |
449 | */ |
450 | #define UTRIE2_U8_PREV16(trie, start, src, result) \ |
451 | _UTRIE2_U8_PREV(trie, data16, index, start, src, result) |
452 | |
453 | /** |
454 | * UTF-8: Pre-decrement src and get a 32-bit value from the trie. |
455 | * |
456 | * @param trie (const UTrie2 *, in) a frozen trie |
457 | * @param start (const char *, in) the start pointer for the text |
458 | * @param src (const char *, in/out) the source text pointer |
459 | * @param result (uint16_t, out) uint32_t variable for the trie lookup result |
460 | */ |
461 | #define UTRIE2_U8_PREV32(trie, start, src, result) \ |
462 | _UTRIE2_U8_PREV(trie, data32, data32, start, src, result) |
463 | |
464 | /* Public UTrie2 API: optimized UTF-16 access ------------------------------- */ |
465 | |
466 | /* |
467 | * The following functions and macros are used for highly optimized UTF-16 |
468 | * text processing. The UTRIE2_U16_NEXTxy() macros do not depend on these. |
469 | * |
470 | * A UTrie2 stores separate values for lead surrogate code _units_ vs. code _points_. |
471 | * UTF-16 text processing can be optimized by detecting surrogate pairs and |
472 | * assembling supplementary code points only when there is non-trivial data |
473 | * available. |
474 | * |
475 | * At build-time, use utrie2_enumForLeadSurrogate() to see if there |
476 | * is non-trivial (non-initialValue) data for any of the supplementary |
477 | * code points associated with a lead surrogate. |
478 | * If so, then set a special (application-specific) value for the |
479 | * lead surrogate code _unit_, with utrie2_set32ForLeadSurrogateCodeUnit(). |
480 | * |
481 | * At runtime, use UTRIE2_GET16_FROM_U16_SINGLE_LEAD() or |
482 | * UTRIE2_GET32_FROM_U16_SINGLE_LEAD() per code unit. If there is non-trivial |
483 | * data and the code unit is a lead surrogate, then check if a trail surrogate |
484 | * follows. If so, assemble the supplementary code point with |
485 | * U16_GET_SUPPLEMENTARY() and look up its value with UTRIE2_GET16_FROM_SUPP() |
486 | * or UTRIE2_GET32_FROM_SUPP(); otherwise reset the lead |
487 | * surrogate's value or do a code point lookup for it. |
488 | * |
489 | * If there is only trivial data for lead and trail surrogates, then processing |
490 | * can often skip them. For example, in normalization or case mapping |
491 | * all characters that do not have any mappings are simply copied as is. |
492 | */ |
493 | |
494 | /** |
495 | * Get a value from a lead surrogate code unit as stored in the trie. |
496 | * |
497 | * @param trie the trie |
498 | * @param c the code unit (U+D800..U+DBFF) |
499 | * @return the value |
500 | */ |
501 | U_CAPI uint32_t U_EXPORT2 |
502 | utrie2_get32FromLeadSurrogateCodeUnit(const UTrie2 *trie, UChar32 c); |
503 | |
504 | /** |
505 | * Enumerate the trie values for the 1024=0x400 code points |
506 | * corresponding to a given lead surrogate. |
507 | * For example, for the lead surrogate U+D87E it will enumerate the values |
508 | * for [U+2F800..U+2FC00[. |
509 | * Used by data builder code that sets special lead surrogate code unit values |
510 | * for optimized UTF-16 string processing. |
511 | * |
512 | * Do not modify the trie during the enumeration. |
513 | * |
514 | * Except for the limited code point range, this functions just like utrie2_enum(): |
515 | * For each entry in the trie, the value to be delivered is passed through |
516 | * the UTrie2EnumValue function. |
517 | * The value is unchanged if that function pointer is NULL. |
518 | * |
519 | * For each contiguous range of code points with a given (transformed) value, |
520 | * the UTrie2EnumRange function is called. |
521 | * |
522 | * @param trie a pointer to the trie |
523 | * @param enumValue a pointer to a function that may transform the trie entry value, |
524 | * or NULL if the values from the trie are to be used directly |
525 | * @param enumRange a pointer to a function that is called for each contiguous range |
526 | * of code points with the same (transformed) value |
527 | * @param context an opaque pointer that is passed on to the callback functions |
528 | */ |
529 | U_CAPI void U_EXPORT2 |
530 | utrie2_enumForLeadSurrogate(const UTrie2 *trie, UChar32 lead, |
531 | UTrie2EnumValue *enumValue, UTrie2EnumRange *enumRange, |
532 | const void *context); |
533 | |
534 | /** |
535 | * Set a value for a lead surrogate code unit. |
536 | * |
537 | * @param trie the unfrozen trie |
538 | * @param lead the lead surrogate code unit (U+D800..U+DBFF) |
539 | * @param value the value |
540 | * @param pErrorCode an in/out ICU UErrorCode; among other possible error codes: |
541 | * - U_NO_WRITE_PERMISSION if the trie is frozen |
542 | */ |
543 | U_CAPI void U_EXPORT2 |
544 | utrie2_set32ForLeadSurrogateCodeUnit(UTrie2 *trie, |
545 | UChar32 lead, uint32_t value, |
546 | UErrorCode *pErrorCode); |
547 | |
548 | /** |
549 | * Return a 16-bit trie value from a UTF-16 single/lead code unit (<=U+ffff). |
550 | * Same as UTRIE2_GET16() if c is a BMP code point except for lead surrogates, |
551 | * but smaller and faster. |
552 | * |
553 | * @param trie (const UTrie2 *, in) a frozen trie |
554 | * @param c (UChar32, in) the input code unit, must be 0<=c<=U+ffff |
555 | * @return (uint16_t) The code unit's trie value. |
556 | */ |
557 | #define UTRIE2_GET16_FROM_U16_SINGLE_LEAD(trie, c) _UTRIE2_GET_FROM_U16_SINGLE_LEAD((trie), index, c) |
558 | |
559 | /** |
560 | * Return a 32-bit trie value from a UTF-16 single/lead code unit (<=U+ffff). |
561 | * Same as UTRIE2_GET32() if c is a BMP code point except for lead surrogates, |
562 | * but smaller and faster. |
563 | * |
564 | * @param trie (const UTrie2 *, in) a frozen trie |
565 | * @param c (UChar32, in) the input code unit, must be 0<=c<=U+ffff |
566 | * @return (uint32_t) The code unit's trie value. |
567 | */ |
568 | #define UTRIE2_GET32_FROM_U16_SINGLE_LEAD(trie, c) _UTRIE2_GET_FROM_U16_SINGLE_LEAD((trie), data32, c) |
569 | |
570 | /** |
571 | * Return a 16-bit trie value from a supplementary code point (U+10000..U+10ffff). |
572 | * |
573 | * @param trie (const UTrie2 *, in) a frozen trie |
574 | * @param c (UChar32, in) the input code point, must be U+10000<=c<=U+10ffff |
575 | * @return (uint16_t) The code point's trie value. |
576 | */ |
577 | #define UTRIE2_GET16_FROM_SUPP(trie, c) _UTRIE2_GET_FROM_SUPP((trie), index, c) |
578 | |
579 | /** |
580 | * Return a 32-bit trie value from a supplementary code point (U+10000..U+10ffff). |
581 | * |
582 | * @param trie (const UTrie2 *, in) a frozen trie |
583 | * @param c (UChar32, in) the input code point, must be U+10000<=c<=U+10ffff |
584 | * @return (uint32_t) The code point's trie value. |
585 | */ |
586 | #define UTRIE2_GET32_FROM_SUPP(trie, c) _UTRIE2_GET_FROM_SUPP((trie), data32, c) |
587 | |
588 | U_CDECL_END |
589 | |
590 | /* C++ convenience wrappers ------------------------------------------------- */ |
591 | |
592 | #ifdef __cplusplus |
593 | |
594 | #include "unicode/uobject.h" |
595 | #include "unicode/utf.h" |
596 | |
597 | U_NAMESPACE_BEGIN |
598 | |
599 | // Use the Forward/Backward subclasses below. |
600 | class UTrie2StringIterator : public UMemory { |
601 | public: |
602 | UTrie2StringIterator(const UTrie2 *t, const char16_t *p) : |
603 | trie(t), codePointStart(p), codePointLimit(p), codePoint(U_SENTINEL) {} |
604 | |
605 | const UTrie2 *trie; |
606 | const char16_t *codePointStart, *codePointLimit; |
607 | UChar32 codePoint; |
608 | }; |
609 | |
610 | class BackwardUTrie2StringIterator : public UTrie2StringIterator { |
611 | public: |
612 | BackwardUTrie2StringIterator(const UTrie2 *t, const char16_t *s, const char16_t *p) : |
613 | UTrie2StringIterator(t, p), start(s) {} |
614 | |
615 | uint16_t previous16(); |
616 | |
617 | const char16_t *start; |
618 | }; |
619 | |
620 | class ForwardUTrie2StringIterator : public UTrie2StringIterator { |
621 | public: |
622 | // Iteration limit l can be nullptr. |
623 | // In that case, the caller must detect c==0 and stop. |
624 | ForwardUTrie2StringIterator(const UTrie2 *t, const char16_t *p, const char16_t *l) : |
625 | UTrie2StringIterator(t, p), limit(l) {} |
626 | |
627 | uint16_t next16(); |
628 | |
629 | const char16_t *limit; |
630 | }; |
631 | |
632 | U_NAMESPACE_END |
633 | |
634 | #endif |
635 | |
636 | /* Internal definitions ----------------------------------------------------- */ |
637 | |
638 | U_CDECL_BEGIN |
639 | |
640 | /** Build-time trie structure. */ |
641 | struct UNewTrie2; |
642 | typedef struct UNewTrie2 UNewTrie2; |
643 | |
644 | /* |
645 | * Trie structure definition. |
646 | * |
647 | * Either the data table is 16 bits wide and accessed via the index |
648 | * pointer, with each index item increased by indexLength; |
649 | * in this case, data32==NULL, and data16 is used for direct ASCII access. |
650 | * |
651 | * Or the data table is 32 bits wide and accessed via the data32 pointer. |
652 | */ |
653 | struct UTrie2 { |
654 | /* protected: used by macros and functions for reading values */ |
655 | const uint16_t *index; |
656 | const uint16_t *data16; /* for fast UTF-8 ASCII access, if 16b data */ |
657 | const uint32_t *data32; /* NULL if 16b data is used via index */ |
658 | |
659 | int32_t indexLength, dataLength; |
660 | uint16_t index2NullOffset; /* 0xffff if there is no dedicated index-2 null block */ |
661 | uint16_t dataNullOffset; |
662 | uint32_t initialValue; |
663 | /** Value returned for out-of-range code points and illegal UTF-8. */ |
664 | uint32_t errorValue; |
665 | |
666 | /* Start of the last range which ends at U+10ffff, and its value. */ |
667 | UChar32 highStart; |
668 | int32_t highValueIndex; |
669 | |
670 | /* private: used by builder and unserialization functions */ |
671 | void *memory; /* serialized bytes; NULL if not frozen yet */ |
672 | int32_t length; /* number of serialized bytes at memory; 0 if not frozen yet */ |
673 | UBool isMemoryOwned; /* true if the trie owns the memory */ |
674 | UBool padding1; |
675 | int16_t padding2; |
676 | UNewTrie2 *newTrie; /* builder object; NULL when frozen */ |
677 | |
678 | #ifdef UTRIE2_DEBUG |
679 | const char *name; |
680 | #endif |
681 | }; |
682 | |
683 | /** |
684 | * Trie constants, defining shift widths, index array lengths, etc. |
685 | * |
686 | * These are needed for the runtime macros but users can treat these as |
687 | * implementation details and skip to the actual public API further below. |
688 | */ |
689 | enum { |
690 | /** Shift size for getting the index-1 table offset. */ |
691 | UTRIE2_SHIFT_1=6+5, |
692 | |
693 | /** Shift size for getting the index-2 table offset. */ |
694 | UTRIE2_SHIFT_2=5, |
695 | |
696 | /** |
697 | * Difference between the two shift sizes, |
698 | * for getting an index-1 offset from an index-2 offset. 6=11-5 |
699 | */ |
700 | UTRIE2_SHIFT_1_2=UTRIE2_SHIFT_1-UTRIE2_SHIFT_2, |
701 | |
702 | /** |
703 | * Number of index-1 entries for the BMP. 32=0x20 |
704 | * This part of the index-1 table is omitted from the serialized form. |
705 | */ |
706 | UTRIE2_OMITTED_BMP_INDEX_1_LENGTH=0x10000>>UTRIE2_SHIFT_1, |
707 | |
708 | /** Number of code points per index-1 table entry. 2048=0x800 */ |
709 | UTRIE2_CP_PER_INDEX_1_ENTRY=1<<UTRIE2_SHIFT_1, |
710 | |
711 | /** Number of entries in an index-2 block. 64=0x40 */ |
712 | UTRIE2_INDEX_2_BLOCK_LENGTH=1<<UTRIE2_SHIFT_1_2, |
713 | |
714 | /** Mask for getting the lower bits for the in-index-2-block offset. */ |
715 | UTRIE2_INDEX_2_MASK=UTRIE2_INDEX_2_BLOCK_LENGTH-1, |
716 | |
717 | /** Number of entries in a data block. 32=0x20 */ |
718 | UTRIE2_DATA_BLOCK_LENGTH=1<<UTRIE2_SHIFT_2, |
719 | |
720 | /** Mask for getting the lower bits for the in-data-block offset. */ |
721 | UTRIE2_DATA_MASK=UTRIE2_DATA_BLOCK_LENGTH-1, |
722 | |
723 | /** |
724 | * Shift size for shifting left the index array values. |
725 | * Increases possible data size with 16-bit index values at the cost |
726 | * of compactability. |
727 | * This requires data blocks to be aligned by UTRIE2_DATA_GRANULARITY. |
728 | */ |
729 | UTRIE2_INDEX_SHIFT=2, |
730 | |
731 | /** The alignment size of a data block. Also the granularity for compaction. */ |
732 | UTRIE2_DATA_GRANULARITY=1<<UTRIE2_INDEX_SHIFT, |
733 | |
734 | /* Fixed layout of the first part of the index array. ------------------- */ |
735 | |
736 | /** |
737 | * The BMP part of the index-2 table is fixed and linear and starts at offset 0. |
738 | * Length=2048=0x800=0x10000>>UTRIE2_SHIFT_2. |
739 | */ |
740 | UTRIE2_INDEX_2_OFFSET=0, |
741 | |
742 | /** |
743 | * The part of the index-2 table for U+D800..U+DBFF stores values for |
744 | * lead surrogate code _units_ not code _points_. |
745 | * Values for lead surrogate code _points_ are indexed with this portion of the table. |
746 | * Length=32=0x20=0x400>>UTRIE2_SHIFT_2. (There are 1024=0x400 lead surrogates.) |
747 | */ |
748 | UTRIE2_LSCP_INDEX_2_OFFSET=0x10000>>UTRIE2_SHIFT_2, |
749 | UTRIE2_LSCP_INDEX_2_LENGTH=0x400>>UTRIE2_SHIFT_2, |
750 | |
751 | /** Count the lengths of both BMP pieces. 2080=0x820 */ |
752 | UTRIE2_INDEX_2_BMP_LENGTH=UTRIE2_LSCP_INDEX_2_OFFSET+UTRIE2_LSCP_INDEX_2_LENGTH, |
753 | |
754 | /** |
755 | * The 2-byte UTF-8 version of the index-2 table follows at offset 2080=0x820. |
756 | * Length 32=0x20 for lead bytes C0..DF, regardless of UTRIE2_SHIFT_2. |
757 | */ |
758 | UTRIE2_UTF8_2B_INDEX_2_OFFSET=UTRIE2_INDEX_2_BMP_LENGTH, |
759 | UTRIE2_UTF8_2B_INDEX_2_LENGTH=0x800>>6, /* U+0800 is the first code point after 2-byte UTF-8 */ |
760 | |
761 | /** |
762 | * The index-1 table, only used for supplementary code points, at offset 2112=0x840. |
763 | * Variable length, for code points up to highStart, where the last single-value range starts. |
764 | * Maximum length 512=0x200=0x100000>>UTRIE2_SHIFT_1. |
765 | * (For 0x100000 supplementary code points U+10000..U+10ffff.) |
766 | * |
767 | * The part of the index-2 table for supplementary code points starts |
768 | * after this index-1 table. |
769 | * |
770 | * Both the index-1 table and the following part of the index-2 table |
771 | * are omitted completely if there is only BMP data. |
772 | */ |
773 | UTRIE2_INDEX_1_OFFSET=UTRIE2_UTF8_2B_INDEX_2_OFFSET+UTRIE2_UTF8_2B_INDEX_2_LENGTH, |
774 | UTRIE2_MAX_INDEX_1_LENGTH=0x100000>>UTRIE2_SHIFT_1, |
775 | |
776 | /* |
777 | * Fixed layout of the first part of the data array. ----------------------- |
778 | * Starts with 4 blocks (128=0x80 entries) for ASCII. |
779 | */ |
780 | |
781 | /** |
782 | * The illegal-UTF-8 data block follows the ASCII block, at offset 128=0x80. |
783 | * Used with linear access for single bytes 0..0xbf for simple error handling. |
784 | * Length 64=0x40, not UTRIE2_DATA_BLOCK_LENGTH. |
785 | */ |
786 | UTRIE2_BAD_UTF8_DATA_OFFSET=0x80, |
787 | |
788 | /** The start of non-linear-ASCII data blocks, at offset 192=0xc0. */ |
789 | UTRIE2_DATA_START_OFFSET=0xc0 |
790 | }; |
791 | |
792 | /* Internal functions and macros -------------------------------------------- */ |
793 | |
794 | /** |
795 | * Internal function for part of the UTRIE2_U8_NEXTxx() macro implementations. |
796 | * Do not call directly. |
797 | * @internal |
798 | */ |
799 | U_CAPI int32_t U_EXPORT2 |
800 | utrie2_internalU8NextIndex(const UTrie2 *trie, UChar32 c, |
801 | const uint8_t *src, const uint8_t *limit); |
802 | |
803 | /** |
804 | * Internal function for part of the UTRIE2_U8_PREVxx() macro implementations. |
805 | * Do not call directly. |
806 | * @internal |
807 | */ |
808 | U_CAPI int32_t U_EXPORT2 |
809 | utrie2_internalU8PrevIndex(const UTrie2 *trie, UChar32 c, |
810 | const uint8_t *start, const uint8_t *src); |
811 | |
812 | |
813 | /** Internal low-level trie getter. Returns a data index. */ |
814 | #define _UTRIE2_INDEX_RAW(offset, trieIndex, c) \ |
815 | (((int32_t)((trieIndex)[(offset)+((c)>>UTRIE2_SHIFT_2)]) \ |
816 | <<UTRIE2_INDEX_SHIFT)+ \ |
817 | ((c)&UTRIE2_DATA_MASK)) |
818 | |
819 | /** Internal trie getter from a UTF-16 single/lead code unit. Returns the data index. */ |
820 | #define _UTRIE2_INDEX_FROM_U16_SINGLE_LEAD(trieIndex, c) _UTRIE2_INDEX_RAW(0, trieIndex, c) |
821 | |
822 | /** Internal trie getter from a lead surrogate code point (D800..DBFF). Returns the data index. */ |
823 | #define _UTRIE2_INDEX_FROM_LSCP(trieIndex, c) \ |
824 | _UTRIE2_INDEX_RAW(UTRIE2_LSCP_INDEX_2_OFFSET-(0xd800>>UTRIE2_SHIFT_2), trieIndex, c) |
825 | |
826 | /** Internal trie getter from a BMP code point. Returns the data index. */ |
827 | #define _UTRIE2_INDEX_FROM_BMP(trieIndex, c) \ |
828 | _UTRIE2_INDEX_RAW(U_IS_LEAD(c) ? UTRIE2_LSCP_INDEX_2_OFFSET-(0xd800>>UTRIE2_SHIFT_2) : 0, \ |
829 | trieIndex, c) |
830 | |
831 | /** Internal trie getter from a supplementary code point below highStart. Returns the data index. */ |
832 | #define _UTRIE2_INDEX_FROM_SUPP(trieIndex, c) \ |
833 | (((int32_t)((trieIndex)[ \ |
834 | (trieIndex)[(UTRIE2_INDEX_1_OFFSET-UTRIE2_OMITTED_BMP_INDEX_1_LENGTH)+ \ |
835 | ((c)>>UTRIE2_SHIFT_1)]+ \ |
836 | (((c)>>UTRIE2_SHIFT_2)&UTRIE2_INDEX_2_MASK)]) \ |
837 | <<UTRIE2_INDEX_SHIFT)+ \ |
838 | ((c)&UTRIE2_DATA_MASK)) |
839 | |
840 | /** |
841 | * Internal trie getter from a code point, with checking that c is in 0..10FFFF. |
842 | * Returns the data index. |
843 | */ |
844 | #define _UTRIE2_INDEX_FROM_CP(trie, asciiOffset, c) \ |
845 | ((uint32_t)(c)<0xd800 ? \ |
846 | _UTRIE2_INDEX_RAW(0, (trie)->index, c) : \ |
847 | (uint32_t)(c)<=0xffff ? \ |
848 | _UTRIE2_INDEX_RAW( \ |
849 | (c)<=0xdbff ? UTRIE2_LSCP_INDEX_2_OFFSET-(0xd800>>UTRIE2_SHIFT_2) : 0, \ |
850 | (trie)->index, c) : \ |
851 | (uint32_t)(c)>0x10ffff ? \ |
852 | (asciiOffset)+UTRIE2_BAD_UTF8_DATA_OFFSET : \ |
853 | (c)>=(trie)->highStart ? \ |
854 | (trie)->highValueIndex : \ |
855 | _UTRIE2_INDEX_FROM_SUPP((trie)->index, c)) |
856 | |
857 | /** Internal trie getter from a UTF-16 single/lead code unit. Returns the data. */ |
858 | #define _UTRIE2_GET_FROM_U16_SINGLE_LEAD(trie, data, c) \ |
859 | (trie)->data[_UTRIE2_INDEX_FROM_U16_SINGLE_LEAD((trie)->index, c)] |
860 | |
861 | /** Internal trie getter from a supplementary code point. Returns the data. */ |
862 | #define _UTRIE2_GET_FROM_SUPP(trie, data, c) \ |
863 | (trie)->data[(c)>=(trie)->highStart ? (trie)->highValueIndex : \ |
864 | _UTRIE2_INDEX_FROM_SUPP((trie)->index, c)] |
865 | |
866 | /** |
867 | * Internal trie getter from a code point, with checking that c is in 0..10FFFF. |
868 | * Returns the data. |
869 | */ |
870 | #define _UTRIE2_GET(trie, data, asciiOffset, c) \ |
871 | (trie)->data[_UTRIE2_INDEX_FROM_CP(trie, asciiOffset, c)] |
872 | |
873 | /** Internal next-post-increment: get the next code point (c) and its data. */ |
874 | #define _UTRIE2_U16_NEXT(trie, data, src, limit, c, result) UPRV_BLOCK_MACRO_BEGIN { \ |
875 | { \ |
876 | uint16_t __c2; \ |
877 | (c)=*(src)++; \ |
878 | if(!U16_IS_LEAD(c)) { \ |
879 | (result)=_UTRIE2_GET_FROM_U16_SINGLE_LEAD(trie, data, c); \ |
880 | } else if((src)==(limit) || !U16_IS_TRAIL(__c2=*(src))) { \ |
881 | (result)=(trie)->data[_UTRIE2_INDEX_FROM_LSCP((trie)->index, c)]; \ |
882 | } else { \ |
883 | ++(src); \ |
884 | (c)=U16_GET_SUPPLEMENTARY((c), __c2); \ |
885 | (result)=_UTRIE2_GET_FROM_SUPP((trie), data, (c)); \ |
886 | } \ |
887 | } \ |
888 | } UPRV_BLOCK_MACRO_END |
889 | |
890 | /** Internal pre-decrement-previous: get the previous code point (c) and its data */ |
891 | #define _UTRIE2_U16_PREV(trie, data, start, src, c, result) UPRV_BLOCK_MACRO_BEGIN { \ |
892 | { \ |
893 | uint16_t __c2; \ |
894 | (c)=*--(src); \ |
895 | if(!U16_IS_TRAIL(c) || (src)==(start) || !U16_IS_LEAD(__c2=*((src)-1))) { \ |
896 | (result)=(trie)->data[_UTRIE2_INDEX_FROM_BMP((trie)->index, c)]; \ |
897 | } else { \ |
898 | --(src); \ |
899 | (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \ |
900 | (result)=_UTRIE2_GET_FROM_SUPP((trie), data, (c)); \ |
901 | } \ |
902 | } \ |
903 | } UPRV_BLOCK_MACRO_END |
904 | |
905 | /** Internal UTF-8 next-post-increment: get the next code point's data. */ |
906 | #define _UTRIE2_U8_NEXT(trie, ascii, data, src, limit, result) UPRV_BLOCK_MACRO_BEGIN { \ |
907 | uint8_t __lead=(uint8_t)*(src)++; \ |
908 | if(U8_IS_SINGLE(__lead)) { \ |
909 | (result)=(trie)->ascii[__lead]; \ |
910 | } else { \ |
911 | uint8_t __t1, __t2; \ |
912 | if( /* handle U+0800..U+FFFF inline */ \ |
913 | 0xe0<=__lead && __lead<0xf0 && ((src)+1)<(limit) && \ |
914 | U8_IS_VALID_LEAD3_AND_T1(__lead, __t1=(uint8_t)*(src)) && \ |
915 | (__t2=(uint8_t)(*((src)+1)-0x80))<= 0x3f \ |
916 | ) { \ |
917 | (src)+=2; \ |
918 | (result)=(trie)->data[ \ |
919 | ((int32_t)((trie)->index[((__lead-0xe0)<<(12-UTRIE2_SHIFT_2))+ \ |
920 | ((__t1&0x3f)<<(6-UTRIE2_SHIFT_2))+(__t2>>UTRIE2_SHIFT_2)]) \ |
921 | <<UTRIE2_INDEX_SHIFT)+ \ |
922 | (__t2&UTRIE2_DATA_MASK)]; \ |
923 | } else if( /* handle U+0080..U+07FF inline */ \ |
924 | __lead<0xe0 && __lead>=0xc2 && (src)<(limit) && \ |
925 | (__t1=(uint8_t)(*(src)-0x80))<=0x3f \ |
926 | ) { \ |
927 | ++(src); \ |
928 | (result)=(trie)->data[ \ |
929 | (trie)->index[(UTRIE2_UTF8_2B_INDEX_2_OFFSET-0xc0)+__lead]+ \ |
930 | __t1]; \ |
931 | } else { \ |
932 | int32_t __index=utrie2_internalU8NextIndex((trie), __lead, (const uint8_t *)(src), \ |
933 | (const uint8_t *)(limit)); \ |
934 | (src)+=__index&7; \ |
935 | (result)=(trie)->data[__index>>3]; \ |
936 | } \ |
937 | } \ |
938 | } UPRV_BLOCK_MACRO_END |
939 | |
940 | /** Internal UTF-8 pre-decrement-previous: get the previous code point's data. */ |
941 | #define _UTRIE2_U8_PREV(trie, ascii, data, start, src, result) UPRV_BLOCK_MACRO_BEGIN { \ |
942 | uint8_t __b=(uint8_t)*--(src); \ |
943 | if(U8_IS_SINGLE(__b)) { \ |
944 | (result)=(trie)->ascii[__b]; \ |
945 | } else { \ |
946 | int32_t __index=utrie2_internalU8PrevIndex((trie), __b, (const uint8_t *)(start), \ |
947 | (const uint8_t *)(src)); \ |
948 | (src)-=__index&7; \ |
949 | (result)=(trie)->data[__index>>3]; \ |
950 | } \ |
951 | } UPRV_BLOCK_MACRO_END |
952 | |
953 | U_CDECL_END |
954 | |
955 | #endif |
956 | |