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) 2009-2013, International Business Machines |
7 | * Corporation and others. All Rights Reserved. |
8 | * |
9 | ******************************************************************************* |
10 | * file name: normalizer2.h |
11 | * encoding: UTF-8 |
12 | * tab size: 8 (not used) |
13 | * indentation:4 |
14 | * |
15 | * created on: 2009nov22 |
16 | * created by: Markus W. Scherer |
17 | */ |
18 | |
19 | #ifndef __NORMALIZER2_H__ |
20 | #define __NORMALIZER2_H__ |
21 | |
22 | /** |
23 | * \file |
24 | * \brief C++ API: New API for Unicode Normalization. |
25 | */ |
26 | |
27 | #include "unicode/utypes.h" |
28 | |
29 | #if U_SHOW_CPLUSPLUS_API |
30 | |
31 | #if !UCONFIG_NO_NORMALIZATION |
32 | |
33 | #include "unicode/stringpiece.h" |
34 | #include "unicode/uniset.h" |
35 | #include "unicode/unistr.h" |
36 | #include "unicode/unorm2.h" |
37 | |
38 | U_NAMESPACE_BEGIN |
39 | |
40 | class ByteSink; |
41 | |
42 | /** |
43 | * Unicode normalization functionality for standard Unicode normalization or |
44 | * for using custom mapping tables. |
45 | * All instances of this class are unmodifiable/immutable. |
46 | * Instances returned by getInstance() are singletons that must not be deleted by the caller. |
47 | * The Normalizer2 class is not intended for public subclassing. |
48 | * |
49 | * The primary functions are to produce a normalized string and to detect whether |
50 | * a string is already normalized. |
51 | * The most commonly used normalization forms are those defined in |
52 | * http://www.unicode.org/unicode/reports/tr15/ |
53 | * However, this API supports additional normalization forms for specialized purposes. |
54 | * For example, NFKC_Casefold is provided via getInstance("nfkc_cf", COMPOSE) |
55 | * and can be used in implementations of UTS #46. |
56 | * |
57 | * Not only are the standard compose and decompose modes supplied, |
58 | * but additional modes are provided as documented in the Mode enum. |
59 | * |
60 | * Some of the functions in this class identify normalization boundaries. |
61 | * At a normalization boundary, the portions of the string |
62 | * before it and starting from it do not interact and can be handled independently. |
63 | * |
64 | * The spanQuickCheckYes() stops at a normalization boundary. |
65 | * When the goal is a normalized string, then the text before the boundary |
66 | * can be copied, and the remainder can be processed with normalizeSecondAndAppend(). |
67 | * |
68 | * The hasBoundaryBefore(), hasBoundaryAfter() and isInert() functions test whether |
69 | * a character is guaranteed to be at a normalization boundary, |
70 | * regardless of context. |
71 | * This is used for moving from one normalization boundary to the next |
72 | * or preceding boundary, and for performing iterative normalization. |
73 | * |
74 | * Iterative normalization is useful when only a small portion of a |
75 | * longer string needs to be processed. |
76 | * For example, in ICU, iterative normalization is used by the NormalizationTransliterator |
77 | * (to avoid replacing already-normalized text) and ucol_nextSortKeyPart() |
78 | * (to process only the substring for which sort key bytes are computed). |
79 | * |
80 | * The set of normalization boundaries returned by these functions may not be |
81 | * complete: There may be more boundaries that could be returned. |
82 | * Different functions may return different boundaries. |
83 | * @stable ICU 4.4 |
84 | */ |
85 | class U_COMMON_API Normalizer2 : public UObject { |
86 | public: |
87 | /** |
88 | * Destructor. |
89 | * @stable ICU 4.4 |
90 | */ |
91 | ~Normalizer2(); |
92 | |
93 | /** |
94 | * Returns a Normalizer2 instance for Unicode NFC normalization. |
95 | * Same as getInstance(NULL, "nfc", UNORM2_COMPOSE, errorCode). |
96 | * Returns an unmodifiable singleton instance. Do not delete it. |
97 | * @param errorCode Standard ICU error code. Its input value must |
98 | * pass the U_SUCCESS() test, or else the function returns |
99 | * immediately. Check for U_FAILURE() on output or use with |
100 | * function chaining. (See User Guide for details.) |
101 | * @return the requested Normalizer2, if successful |
102 | * @stable ICU 49 |
103 | */ |
104 | static const Normalizer2 * |
105 | getNFCInstance(UErrorCode &errorCode); |
106 | |
107 | /** |
108 | * Returns a Normalizer2 instance for Unicode NFD normalization. |
109 | * Same as getInstance(NULL, "nfc", UNORM2_DECOMPOSE, errorCode). |
110 | * Returns an unmodifiable singleton instance. Do not delete it. |
111 | * @param errorCode Standard ICU error code. Its input value must |
112 | * pass the U_SUCCESS() test, or else the function returns |
113 | * immediately. Check for U_FAILURE() on output or use with |
114 | * function chaining. (See User Guide for details.) |
115 | * @return the requested Normalizer2, if successful |
116 | * @stable ICU 49 |
117 | */ |
118 | static const Normalizer2 * |
119 | getNFDInstance(UErrorCode &errorCode); |
120 | |
121 | /** |
122 | * Returns a Normalizer2 instance for Unicode NFKC normalization. |
123 | * Same as getInstance(NULL, "nfkc", UNORM2_COMPOSE, errorCode). |
124 | * Returns an unmodifiable singleton instance. Do not delete it. |
125 | * @param errorCode Standard ICU error code. Its input value must |
126 | * pass the U_SUCCESS() test, or else the function returns |
127 | * immediately. Check for U_FAILURE() on output or use with |
128 | * function chaining. (See User Guide for details.) |
129 | * @return the requested Normalizer2, if successful |
130 | * @stable ICU 49 |
131 | */ |
132 | static const Normalizer2 * |
133 | getNFKCInstance(UErrorCode &errorCode); |
134 | |
135 | /** |
136 | * Returns a Normalizer2 instance for Unicode NFKD normalization. |
137 | * Same as getInstance(NULL, "nfkc", UNORM2_DECOMPOSE, errorCode). |
138 | * Returns an unmodifiable singleton instance. Do not delete it. |
139 | * @param errorCode Standard ICU error code. Its input value must |
140 | * pass the U_SUCCESS() test, or else the function returns |
141 | * immediately. Check for U_FAILURE() on output or use with |
142 | * function chaining. (See User Guide for details.) |
143 | * @return the requested Normalizer2, if successful |
144 | * @stable ICU 49 |
145 | */ |
146 | static const Normalizer2 * |
147 | getNFKDInstance(UErrorCode &errorCode); |
148 | |
149 | /** |
150 | * Returns a Normalizer2 instance for Unicode NFKC_Casefold normalization. |
151 | * Same as getInstance(NULL, "nfkc_cf", UNORM2_COMPOSE, errorCode). |
152 | * Returns an unmodifiable singleton instance. Do not delete it. |
153 | * @param errorCode Standard ICU error code. Its input value must |
154 | * pass the U_SUCCESS() test, or else the function returns |
155 | * immediately. Check for U_FAILURE() on output or use with |
156 | * function chaining. (See User Guide for details.) |
157 | * @return the requested Normalizer2, if successful |
158 | * @stable ICU 49 |
159 | */ |
160 | static const Normalizer2 * |
161 | getNFKCCasefoldInstance(UErrorCode &errorCode); |
162 | |
163 | /** |
164 | * Returns a Normalizer2 instance which uses the specified data file |
165 | * (packageName/name similar to ucnv_openPackage() and ures_open()/ResourceBundle) |
166 | * and which composes or decomposes text according to the specified mode. |
167 | * Returns an unmodifiable singleton instance. Do not delete it. |
168 | * |
169 | * Use packageName=NULL for data files that are part of ICU's own data. |
170 | * Use name="nfc" and UNORM2_COMPOSE/UNORM2_DECOMPOSE for Unicode standard NFC/NFD. |
171 | * Use name="nfkc" and UNORM2_COMPOSE/UNORM2_DECOMPOSE for Unicode standard NFKC/NFKD. |
172 | * Use name="nfkc_cf" and UNORM2_COMPOSE for Unicode standard NFKC_CF=NFKC_Casefold. |
173 | * |
174 | * @param packageName NULL for ICU built-in data, otherwise application data package name |
175 | * @param name "nfc" or "nfkc" or "nfkc_cf" or name of custom data file |
176 | * @param mode normalization mode (compose or decompose etc.) |
177 | * @param errorCode Standard ICU error code. Its input value must |
178 | * pass the U_SUCCESS() test, or else the function returns |
179 | * immediately. Check for U_FAILURE() on output or use with |
180 | * function chaining. (See User Guide for details.) |
181 | * @return the requested Normalizer2, if successful |
182 | * @stable ICU 4.4 |
183 | */ |
184 | static const Normalizer2 * |
185 | getInstance(const char *packageName, |
186 | const char *name, |
187 | UNormalization2Mode mode, |
188 | UErrorCode &errorCode); |
189 | |
190 | /** |
191 | * Returns the normalized form of the source string. |
192 | * @param src source string |
193 | * @param errorCode Standard ICU error code. Its input value must |
194 | * pass the U_SUCCESS() test, or else the function returns |
195 | * immediately. Check for U_FAILURE() on output or use with |
196 | * function chaining. (See User Guide for details.) |
197 | * @return normalized src |
198 | * @stable ICU 4.4 |
199 | */ |
200 | UnicodeString |
201 | normalize(const UnicodeString &src, UErrorCode &errorCode) const { |
202 | UnicodeString result; |
203 | normalize(src, dest&: result, errorCode); |
204 | return result; |
205 | } |
206 | /** |
207 | * Writes the normalized form of the source string to the destination string |
208 | * (replacing its contents) and returns the destination string. |
209 | * The source and destination strings must be different objects. |
210 | * @param src source string |
211 | * @param dest destination string; its contents is replaced with normalized src |
212 | * @param errorCode Standard ICU error code. Its input value must |
213 | * pass the U_SUCCESS() test, or else the function returns |
214 | * immediately. Check for U_FAILURE() on output or use with |
215 | * function chaining. (See User Guide for details.) |
216 | * @return dest |
217 | * @stable ICU 4.4 |
218 | */ |
219 | virtual UnicodeString & |
220 | normalize(const UnicodeString &src, |
221 | UnicodeString &dest, |
222 | UErrorCode &errorCode) const = 0; |
223 | |
224 | /** |
225 | * Normalizes a UTF-8 string and optionally records how source substrings |
226 | * relate to changed and unchanged result substrings. |
227 | * |
228 | * Implemented completely for all built-in modes except for FCD. |
229 | * The base class implementation converts to & from UTF-16 and does not support edits. |
230 | * |
231 | * @param options Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT and U_EDITS_NO_RESET. |
232 | * @param src Source UTF-8 string. |
233 | * @param sink A ByteSink to which the normalized UTF-8 result string is written. |
234 | * sink.Flush() is called at the end. |
235 | * @param edits Records edits for index mapping, working with styled text, |
236 | * and getting only changes (if any). |
237 | * The Edits contents is undefined if any error occurs. |
238 | * This function calls edits->reset() first unless |
239 | * options includes U_EDITS_NO_RESET. edits can be nullptr. |
240 | * @param errorCode Standard ICU error code. Its input value must |
241 | * pass the U_SUCCESS() test, or else the function returns |
242 | * immediately. Check for U_FAILURE() on output or use with |
243 | * function chaining. (See User Guide for details.) |
244 | * @stable ICU 60 |
245 | */ |
246 | virtual void |
247 | normalizeUTF8(uint32_t options, StringPiece src, ByteSink &sink, |
248 | Edits *edits, UErrorCode &errorCode) const; |
249 | |
250 | /** |
251 | * Appends the normalized form of the second string to the first string |
252 | * (merging them at the boundary) and returns the first string. |
253 | * The result is normalized if the first string was normalized. |
254 | * The first and second strings must be different objects. |
255 | * @param first string, should be normalized |
256 | * @param second string, will be normalized |
257 | * @param errorCode Standard ICU error code. Its input value must |
258 | * pass the U_SUCCESS() test, or else the function returns |
259 | * immediately. Check for U_FAILURE() on output or use with |
260 | * function chaining. (See User Guide for details.) |
261 | * @return first |
262 | * @stable ICU 4.4 |
263 | */ |
264 | virtual UnicodeString & |
265 | normalizeSecondAndAppend(UnicodeString &first, |
266 | const UnicodeString &second, |
267 | UErrorCode &errorCode) const = 0; |
268 | /** |
269 | * Appends the second string to the first string |
270 | * (merging them at the boundary) and returns the first string. |
271 | * The result is normalized if both the strings were normalized. |
272 | * The first and second strings must be different objects. |
273 | * @param first string, should be normalized |
274 | * @param second string, should be normalized |
275 | * @param errorCode Standard ICU error code. Its input value must |
276 | * pass the U_SUCCESS() test, or else the function returns |
277 | * immediately. Check for U_FAILURE() on output or use with |
278 | * function chaining. (See User Guide for details.) |
279 | * @return first |
280 | * @stable ICU 4.4 |
281 | */ |
282 | virtual UnicodeString & |
283 | append(UnicodeString &first, |
284 | const UnicodeString &second, |
285 | UErrorCode &errorCode) const = 0; |
286 | |
287 | /** |
288 | * Gets the decomposition mapping of c. |
289 | * Roughly equivalent to normalizing the String form of c |
290 | * on a UNORM2_DECOMPOSE Normalizer2 instance, but much faster, and except that this function |
291 | * returns false and does not write a string |
292 | * if c does not have a decomposition mapping in this instance's data. |
293 | * This function is independent of the mode of the Normalizer2. |
294 | * @param c code point |
295 | * @param decomposition String object which will be set to c's |
296 | * decomposition mapping, if there is one. |
297 | * @return true if c has a decomposition, otherwise false |
298 | * @stable ICU 4.6 |
299 | */ |
300 | virtual UBool |
301 | getDecomposition(UChar32 c, UnicodeString &decomposition) const = 0; |
302 | |
303 | /** |
304 | * Gets the raw decomposition mapping of c. |
305 | * |
306 | * This is similar to the getDecomposition() method but returns the |
307 | * raw decomposition mapping as specified in UnicodeData.txt or |
308 | * (for custom data) in the mapping files processed by the gennorm2 tool. |
309 | * By contrast, getDecomposition() returns the processed, |
310 | * recursively-decomposed version of this mapping. |
311 | * |
312 | * When used on a standard NFKC Normalizer2 instance, |
313 | * getRawDecomposition() returns the Unicode Decomposition_Mapping (dm) property. |
314 | * |
315 | * When used on a standard NFC Normalizer2 instance, |
316 | * it returns the Decomposition_Mapping only if the Decomposition_Type (dt) is Canonical (Can); |
317 | * in this case, the result contains either one or two code points (=1..4 char16_ts). |
318 | * |
319 | * This function is independent of the mode of the Normalizer2. |
320 | * The default implementation returns false. |
321 | * @param c code point |
322 | * @param decomposition String object which will be set to c's |
323 | * raw decomposition mapping, if there is one. |
324 | * @return true if c has a decomposition, otherwise false |
325 | * @stable ICU 49 |
326 | */ |
327 | virtual UBool |
328 | getRawDecomposition(UChar32 c, UnicodeString &decomposition) const; |
329 | |
330 | /** |
331 | * Performs pairwise composition of a & b and returns the composite if there is one. |
332 | * |
333 | * Returns a composite code point c only if c has a two-way mapping to a+b. |
334 | * In standard Unicode normalization, this means that |
335 | * c has a canonical decomposition to a+b |
336 | * and c does not have the Full_Composition_Exclusion property. |
337 | * |
338 | * This function is independent of the mode of the Normalizer2. |
339 | * The default implementation returns a negative value. |
340 | * @param a A (normalization starter) code point. |
341 | * @param b Another code point. |
342 | * @return The non-negative composite code point if there is one; otherwise a negative value. |
343 | * @stable ICU 49 |
344 | */ |
345 | virtual UChar32 |
346 | composePair(UChar32 a, UChar32 b) const; |
347 | |
348 | /** |
349 | * Gets the combining class of c. |
350 | * The default implementation returns 0 |
351 | * but all standard implementations return the Unicode Canonical_Combining_Class value. |
352 | * @param c code point |
353 | * @return c's combining class |
354 | * @stable ICU 49 |
355 | */ |
356 | virtual uint8_t |
357 | getCombiningClass(UChar32 c) const; |
358 | |
359 | /** |
360 | * Tests if the string is normalized. |
361 | * Internally, in cases where the quickCheck() method would return "maybe" |
362 | * (which is only possible for the two COMPOSE modes) this method |
363 | * resolves to "yes" or "no" to provide a definitive result, |
364 | * at the cost of doing more work in those cases. |
365 | * @param s input string |
366 | * @param errorCode Standard ICU error code. Its input value must |
367 | * pass the U_SUCCESS() test, or else the function returns |
368 | * immediately. Check for U_FAILURE() on output or use with |
369 | * function chaining. (See User Guide for details.) |
370 | * @return true if s is normalized |
371 | * @stable ICU 4.4 |
372 | */ |
373 | virtual UBool |
374 | isNormalized(const UnicodeString &s, UErrorCode &errorCode) const = 0; |
375 | /** |
376 | * Tests if the UTF-8 string is normalized. |
377 | * Internally, in cases where the quickCheck() method would return "maybe" |
378 | * (which is only possible for the two COMPOSE modes) this method |
379 | * resolves to "yes" or "no" to provide a definitive result, |
380 | * at the cost of doing more work in those cases. |
381 | * |
382 | * This works for all normalization modes. |
383 | * It is optimized for UTF-8 for all built-in modes except for FCD. |
384 | * The base class implementation converts to UTF-16 and calls isNormalized(). |
385 | * |
386 | * @param s UTF-8 input string |
387 | * @param errorCode Standard ICU error code. Its input value must |
388 | * pass the U_SUCCESS() test, or else the function returns |
389 | * immediately. Check for U_FAILURE() on output or use with |
390 | * function chaining. (See User Guide for details.) |
391 | * @return true if s is normalized |
392 | * @stable ICU 60 |
393 | */ |
394 | virtual UBool |
395 | isNormalizedUTF8(StringPiece s, UErrorCode &errorCode) const; |
396 | |
397 | |
398 | /** |
399 | * Tests if the string is normalized. |
400 | * For the two COMPOSE modes, the result could be "maybe" in cases that |
401 | * would take a little more work to resolve definitively. |
402 | * Use spanQuickCheckYes() and normalizeSecondAndAppend() for a faster |
403 | * combination of quick check + normalization, to avoid |
404 | * re-checking the "yes" prefix. |
405 | * @param s input string |
406 | * @param errorCode Standard ICU error code. Its input value must |
407 | * pass the U_SUCCESS() test, or else the function returns |
408 | * immediately. Check for U_FAILURE() on output or use with |
409 | * function chaining. (See User Guide for details.) |
410 | * @return UNormalizationCheckResult |
411 | * @stable ICU 4.4 |
412 | */ |
413 | virtual UNormalizationCheckResult |
414 | quickCheck(const UnicodeString &s, UErrorCode &errorCode) const = 0; |
415 | |
416 | /** |
417 | * Returns the end of the normalized substring of the input string. |
418 | * In other words, with <code>end=spanQuickCheckYes(s, ec);</code> |
419 | * the substring <code>UnicodeString(s, 0, end)</code> |
420 | * will pass the quick check with a "yes" result. |
421 | * |
422 | * The returned end index is usually one or more characters before the |
423 | * "no" or "maybe" character: The end index is at a normalization boundary. |
424 | * (See the class documentation for more about normalization boundaries.) |
425 | * |
426 | * When the goal is a normalized string and most input strings are expected |
427 | * to be normalized already, then call this method, |
428 | * and if it returns a prefix shorter than the input string, |
429 | * copy that prefix and use normalizeSecondAndAppend() for the remainder. |
430 | * @param s input string |
431 | * @param errorCode Standard ICU error code. Its input value must |
432 | * pass the U_SUCCESS() test, or else the function returns |
433 | * immediately. Check for U_FAILURE() on output or use with |
434 | * function chaining. (See User Guide for details.) |
435 | * @return "yes" span end index |
436 | * @stable ICU 4.4 |
437 | */ |
438 | virtual int32_t |
439 | spanQuickCheckYes(const UnicodeString &s, UErrorCode &errorCode) const = 0; |
440 | |
441 | /** |
442 | * Tests if the character always has a normalization boundary before it, |
443 | * regardless of context. |
444 | * If true, then the character does not normalization-interact with |
445 | * preceding characters. |
446 | * In other words, a string containing this character can be normalized |
447 | * by processing portions before this character and starting from this |
448 | * character independently. |
449 | * This is used for iterative normalization. See the class documentation for details. |
450 | * @param c character to test |
451 | * @return true if c has a normalization boundary before it |
452 | * @stable ICU 4.4 |
453 | */ |
454 | virtual UBool hasBoundaryBefore(UChar32 c) const = 0; |
455 | |
456 | /** |
457 | * Tests if the character always has a normalization boundary after it, |
458 | * regardless of context. |
459 | * If true, then the character does not normalization-interact with |
460 | * following characters. |
461 | * In other words, a string containing this character can be normalized |
462 | * by processing portions up to this character and after this |
463 | * character independently. |
464 | * This is used for iterative normalization. See the class documentation for details. |
465 | * Note that this operation may be significantly slower than hasBoundaryBefore(). |
466 | * @param c character to test |
467 | * @return true if c has a normalization boundary after it |
468 | * @stable ICU 4.4 |
469 | */ |
470 | virtual UBool hasBoundaryAfter(UChar32 c) const = 0; |
471 | |
472 | /** |
473 | * Tests if the character is normalization-inert. |
474 | * If true, then the character does not change, nor normalization-interact with |
475 | * preceding or following characters. |
476 | * In other words, a string containing this character can be normalized |
477 | * by processing portions before this character and after this |
478 | * character independently. |
479 | * This is used for iterative normalization. See the class documentation for details. |
480 | * Note that this operation may be significantly slower than hasBoundaryBefore(). |
481 | * @param c character to test |
482 | * @return true if c is normalization-inert |
483 | * @stable ICU 4.4 |
484 | */ |
485 | virtual UBool isInert(UChar32 c) const = 0; |
486 | }; |
487 | |
488 | /** |
489 | * Normalization filtered by a UnicodeSet. |
490 | * Normalizes portions of the text contained in the filter set and leaves |
491 | * portions not contained in the filter set unchanged. |
492 | * Filtering is done via UnicodeSet::span(..., USET_SPAN_SIMPLE). |
493 | * Not-in-the-filter text is treated as "is normalized" and "quick check yes". |
494 | * This class implements all of (and only) the Normalizer2 API. |
495 | * An instance of this class is unmodifiable/immutable but is constructed and |
496 | * must be destructed by the owner. |
497 | * @stable ICU 4.4 |
498 | */ |
499 | class U_COMMON_API FilteredNormalizer2 : public Normalizer2 { |
500 | public: |
501 | /** |
502 | * Constructs a filtered normalizer wrapping any Normalizer2 instance |
503 | * and a filter set. |
504 | * Both are aliased and must not be modified or deleted while this object |
505 | * is used. |
506 | * The filter set should be frozen; otherwise the performance will suffer greatly. |
507 | * @param n2 wrapped Normalizer2 instance |
508 | * @param filterSet UnicodeSet which determines the characters to be normalized |
509 | * @stable ICU 4.4 |
510 | */ |
511 | FilteredNormalizer2(const Normalizer2 &n2, const UnicodeSet &filterSet) : |
512 | norm2(n2), set(filterSet) {} |
513 | |
514 | /** |
515 | * Destructor. |
516 | * @stable ICU 4.4 |
517 | */ |
518 | ~FilteredNormalizer2(); |
519 | |
520 | /** |
521 | * Writes the normalized form of the source string to the destination string |
522 | * (replacing its contents) and returns the destination string. |
523 | * The source and destination strings must be different objects. |
524 | * @param src source string |
525 | * @param dest destination string; its contents is replaced with normalized src |
526 | * @param errorCode Standard ICU error code. Its input value must |
527 | * pass the U_SUCCESS() test, or else the function returns |
528 | * immediately. Check for U_FAILURE() on output or use with |
529 | * function chaining. (See User Guide for details.) |
530 | * @return dest |
531 | * @stable ICU 4.4 |
532 | */ |
533 | virtual UnicodeString & |
534 | normalize(const UnicodeString &src, |
535 | UnicodeString &dest, |
536 | UErrorCode &errorCode) const U_OVERRIDE; |
537 | |
538 | /** |
539 | * Normalizes a UTF-8 string and optionally records how source substrings |
540 | * relate to changed and unchanged result substrings. |
541 | * |
542 | * Implemented completely for most built-in modes except for FCD. |
543 | * The base class implementation converts to & from UTF-16 and does not support edits. |
544 | * |
545 | * @param options Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT and U_EDITS_NO_RESET. |
546 | * @param src Source UTF-8 string. |
547 | * @param sink A ByteSink to which the normalized UTF-8 result string is written. |
548 | * sink.Flush() is called at the end. |
549 | * @param edits Records edits for index mapping, working with styled text, |
550 | * and getting only changes (if any). |
551 | * The Edits contents is undefined if any error occurs. |
552 | * This function calls edits->reset() first unless |
553 | * options includes U_EDITS_NO_RESET. edits can be nullptr. |
554 | * @param errorCode Standard ICU error code. Its input value must |
555 | * pass the U_SUCCESS() test, or else the function returns |
556 | * immediately. Check for U_FAILURE() on output or use with |
557 | * function chaining. (See User Guide for details.) |
558 | * @stable ICU 60 |
559 | */ |
560 | virtual void |
561 | normalizeUTF8(uint32_t options, StringPiece src, ByteSink &sink, |
562 | Edits *edits, UErrorCode &errorCode) const U_OVERRIDE; |
563 | |
564 | /** |
565 | * Appends the normalized form of the second string to the first string |
566 | * (merging them at the boundary) and returns the first string. |
567 | * The result is normalized if the first string was normalized. |
568 | * The first and second strings must be different objects. |
569 | * @param first string, should be normalized |
570 | * @param second string, will be normalized |
571 | * @param errorCode Standard ICU error code. Its input value must |
572 | * pass the U_SUCCESS() test, or else the function returns |
573 | * immediately. Check for U_FAILURE() on output or use with |
574 | * function chaining. (See User Guide for details.) |
575 | * @return first |
576 | * @stable ICU 4.4 |
577 | */ |
578 | virtual UnicodeString & |
579 | normalizeSecondAndAppend(UnicodeString &first, |
580 | const UnicodeString &second, |
581 | UErrorCode &errorCode) const U_OVERRIDE; |
582 | /** |
583 | * Appends the second string to the first string |
584 | * (merging them at the boundary) and returns the first string. |
585 | * The result is normalized if both the strings were normalized. |
586 | * The first and second strings must be different objects. |
587 | * @param first string, should be normalized |
588 | * @param second string, should be normalized |
589 | * @param errorCode Standard ICU error code. Its input value must |
590 | * pass the U_SUCCESS() test, or else the function returns |
591 | * immediately. Check for U_FAILURE() on output or use with |
592 | * function chaining. (See User Guide for details.) |
593 | * @return first |
594 | * @stable ICU 4.4 |
595 | */ |
596 | virtual UnicodeString & |
597 | append(UnicodeString &first, |
598 | const UnicodeString &second, |
599 | UErrorCode &errorCode) const U_OVERRIDE; |
600 | |
601 | /** |
602 | * Gets the decomposition mapping of c. |
603 | * For details see the base class documentation. |
604 | * |
605 | * This function is independent of the mode of the Normalizer2. |
606 | * @param c code point |
607 | * @param decomposition String object which will be set to c's |
608 | * decomposition mapping, if there is one. |
609 | * @return true if c has a decomposition, otherwise false |
610 | * @stable ICU 4.6 |
611 | */ |
612 | virtual UBool |
613 | getDecomposition(UChar32 c, UnicodeString &decomposition) const U_OVERRIDE; |
614 | |
615 | /** |
616 | * Gets the raw decomposition mapping of c. |
617 | * For details see the base class documentation. |
618 | * |
619 | * This function is independent of the mode of the Normalizer2. |
620 | * @param c code point |
621 | * @param decomposition String object which will be set to c's |
622 | * raw decomposition mapping, if there is one. |
623 | * @return true if c has a decomposition, otherwise false |
624 | * @stable ICU 49 |
625 | */ |
626 | virtual UBool |
627 | getRawDecomposition(UChar32 c, UnicodeString &decomposition) const U_OVERRIDE; |
628 | |
629 | /** |
630 | * Performs pairwise composition of a & b and returns the composite if there is one. |
631 | * For details see the base class documentation. |
632 | * |
633 | * This function is independent of the mode of the Normalizer2. |
634 | * @param a A (normalization starter) code point. |
635 | * @param b Another code point. |
636 | * @return The non-negative composite code point if there is one; otherwise a negative value. |
637 | * @stable ICU 49 |
638 | */ |
639 | virtual UChar32 |
640 | composePair(UChar32 a, UChar32 b) const U_OVERRIDE; |
641 | |
642 | /** |
643 | * Gets the combining class of c. |
644 | * The default implementation returns 0 |
645 | * but all standard implementations return the Unicode Canonical_Combining_Class value. |
646 | * @param c code point |
647 | * @return c's combining class |
648 | * @stable ICU 49 |
649 | */ |
650 | virtual uint8_t |
651 | getCombiningClass(UChar32 c) const U_OVERRIDE; |
652 | |
653 | /** |
654 | * Tests if the string is normalized. |
655 | * For details see the Normalizer2 base class documentation. |
656 | * @param s input string |
657 | * @param errorCode Standard ICU error code. Its input value must |
658 | * pass the U_SUCCESS() test, or else the function returns |
659 | * immediately. Check for U_FAILURE() on output or use with |
660 | * function chaining. (See User Guide for details.) |
661 | * @return true if s is normalized |
662 | * @stable ICU 4.4 |
663 | */ |
664 | virtual UBool |
665 | isNormalized(const UnicodeString &s, UErrorCode &errorCode) const U_OVERRIDE; |
666 | /** |
667 | * Tests if the UTF-8 string is normalized. |
668 | * Internally, in cases where the quickCheck() method would return "maybe" |
669 | * (which is only possible for the two COMPOSE modes) this method |
670 | * resolves to "yes" or "no" to provide a definitive result, |
671 | * at the cost of doing more work in those cases. |
672 | * |
673 | * This works for all normalization modes. |
674 | * It is optimized for UTF-8 for all built-in modes except for FCD. |
675 | * The base class implementation converts to UTF-16 and calls isNormalized(). |
676 | * |
677 | * @param s UTF-8 input string |
678 | * @param errorCode Standard ICU error code. Its input value must |
679 | * pass the U_SUCCESS() test, or else the function returns |
680 | * immediately. Check for U_FAILURE() on output or use with |
681 | * function chaining. (See User Guide for details.) |
682 | * @return true if s is normalized |
683 | * @stable ICU 60 |
684 | */ |
685 | virtual UBool |
686 | isNormalizedUTF8(StringPiece s, UErrorCode &errorCode) const U_OVERRIDE; |
687 | /** |
688 | * Tests if the string is normalized. |
689 | * For details see the Normalizer2 base class documentation. |
690 | * @param s input string |
691 | * @param errorCode Standard ICU error code. Its input value must |
692 | * pass the U_SUCCESS() test, or else the function returns |
693 | * immediately. Check for U_FAILURE() on output or use with |
694 | * function chaining. (See User Guide for details.) |
695 | * @return UNormalizationCheckResult |
696 | * @stable ICU 4.4 |
697 | */ |
698 | virtual UNormalizationCheckResult |
699 | quickCheck(const UnicodeString &s, UErrorCode &errorCode) const U_OVERRIDE; |
700 | /** |
701 | * Returns the end of the normalized substring of the input string. |
702 | * For details see the Normalizer2 base class documentation. |
703 | * @param s input string |
704 | * @param errorCode Standard ICU error code. Its input value must |
705 | * pass the U_SUCCESS() test, or else the function returns |
706 | * immediately. Check for U_FAILURE() on output or use with |
707 | * function chaining. (See User Guide for details.) |
708 | * @return "yes" span end index |
709 | * @stable ICU 4.4 |
710 | */ |
711 | virtual int32_t |
712 | spanQuickCheckYes(const UnicodeString &s, UErrorCode &errorCode) const U_OVERRIDE; |
713 | |
714 | /** |
715 | * Tests if the character always has a normalization boundary before it, |
716 | * regardless of context. |
717 | * For details see the Normalizer2 base class documentation. |
718 | * @param c character to test |
719 | * @return true if c has a normalization boundary before it |
720 | * @stable ICU 4.4 |
721 | */ |
722 | virtual UBool hasBoundaryBefore(UChar32 c) const U_OVERRIDE; |
723 | |
724 | /** |
725 | * Tests if the character always has a normalization boundary after it, |
726 | * regardless of context. |
727 | * For details see the Normalizer2 base class documentation. |
728 | * @param c character to test |
729 | * @return true if c has a normalization boundary after it |
730 | * @stable ICU 4.4 |
731 | */ |
732 | virtual UBool hasBoundaryAfter(UChar32 c) const U_OVERRIDE; |
733 | |
734 | /** |
735 | * Tests if the character is normalization-inert. |
736 | * For details see the Normalizer2 base class documentation. |
737 | * @param c character to test |
738 | * @return true if c is normalization-inert |
739 | * @stable ICU 4.4 |
740 | */ |
741 | virtual UBool isInert(UChar32 c) const U_OVERRIDE; |
742 | private: |
743 | UnicodeString & |
744 | normalize(const UnicodeString &src, |
745 | UnicodeString &dest, |
746 | USetSpanCondition spanCondition, |
747 | UErrorCode &errorCode) const; |
748 | |
749 | void |
750 | normalizeUTF8(uint32_t options, const char *src, int32_t length, |
751 | ByteSink &sink, Edits *edits, |
752 | USetSpanCondition spanCondition, |
753 | UErrorCode &errorCode) const; |
754 | |
755 | UnicodeString & |
756 | normalizeSecondAndAppend(UnicodeString &first, |
757 | const UnicodeString &second, |
758 | UBool doNormalize, |
759 | UErrorCode &errorCode) const; |
760 | |
761 | const Normalizer2 &norm2; |
762 | const UnicodeSet &set; |
763 | }; |
764 | |
765 | U_NAMESPACE_END |
766 | |
767 | #endif // !UCONFIG_NO_NORMALIZATION |
768 | |
769 | #endif /* U_SHOW_CPLUSPLUS_API */ |
770 | |
771 | #endif // __NORMALIZER2_H__ |
772 | |