Wed, 31 Dec 2014 06:09:35 +0100
Cloned upstream origin tor-browser at tor-browser-31.3.0esr-4.5-1-build1
revision ID fc1c9ff7c1b2defdbc039f12214767608f46423f for hacking purpose.
michael@0 | 1 | /* |
michael@0 | 2 | ******************************************************************************* |
michael@0 | 3 | * |
michael@0 | 4 | * Copyright (C) 2009-2013, International Business Machines |
michael@0 | 5 | * Corporation and others. All Rights Reserved. |
michael@0 | 6 | * |
michael@0 | 7 | ******************************************************************************* |
michael@0 | 8 | * file name: normalizer2.h |
michael@0 | 9 | * encoding: US-ASCII |
michael@0 | 10 | * tab size: 8 (not used) |
michael@0 | 11 | * indentation:4 |
michael@0 | 12 | * |
michael@0 | 13 | * created on: 2009nov22 |
michael@0 | 14 | * created by: Markus W. Scherer |
michael@0 | 15 | */ |
michael@0 | 16 | |
michael@0 | 17 | #ifndef __NORMALIZER2_H__ |
michael@0 | 18 | #define __NORMALIZER2_H__ |
michael@0 | 19 | |
michael@0 | 20 | /** |
michael@0 | 21 | * \file |
michael@0 | 22 | * \brief C++ API: New API for Unicode Normalization. |
michael@0 | 23 | */ |
michael@0 | 24 | |
michael@0 | 25 | #include "unicode/utypes.h" |
michael@0 | 26 | |
michael@0 | 27 | #if !UCONFIG_NO_NORMALIZATION |
michael@0 | 28 | |
michael@0 | 29 | #include "unicode/uniset.h" |
michael@0 | 30 | #include "unicode/unistr.h" |
michael@0 | 31 | #include "unicode/unorm2.h" |
michael@0 | 32 | |
michael@0 | 33 | U_NAMESPACE_BEGIN |
michael@0 | 34 | |
michael@0 | 35 | /** |
michael@0 | 36 | * Unicode normalization functionality for standard Unicode normalization or |
michael@0 | 37 | * for using custom mapping tables. |
michael@0 | 38 | * All instances of this class are unmodifiable/immutable. |
michael@0 | 39 | * Instances returned by getInstance() are singletons that must not be deleted by the caller. |
michael@0 | 40 | * The Normalizer2 class is not intended for public subclassing. |
michael@0 | 41 | * |
michael@0 | 42 | * The primary functions are to produce a normalized string and to detect whether |
michael@0 | 43 | * a string is already normalized. |
michael@0 | 44 | * The most commonly used normalization forms are those defined in |
michael@0 | 45 | * http://www.unicode.org/unicode/reports/tr15/ |
michael@0 | 46 | * However, this API supports additional normalization forms for specialized purposes. |
michael@0 | 47 | * For example, NFKC_Casefold is provided via getInstance("nfkc_cf", COMPOSE) |
michael@0 | 48 | * and can be used in implementations of UTS #46. |
michael@0 | 49 | * |
michael@0 | 50 | * Not only are the standard compose and decompose modes supplied, |
michael@0 | 51 | * but additional modes are provided as documented in the Mode enum. |
michael@0 | 52 | * |
michael@0 | 53 | * Some of the functions in this class identify normalization boundaries. |
michael@0 | 54 | * At a normalization boundary, the portions of the string |
michael@0 | 55 | * before it and starting from it do not interact and can be handled independently. |
michael@0 | 56 | * |
michael@0 | 57 | * The spanQuickCheckYes() stops at a normalization boundary. |
michael@0 | 58 | * When the goal is a normalized string, then the text before the boundary |
michael@0 | 59 | * can be copied, and the remainder can be processed with normalizeSecondAndAppend(). |
michael@0 | 60 | * |
michael@0 | 61 | * The hasBoundaryBefore(), hasBoundaryAfter() and isInert() functions test whether |
michael@0 | 62 | * a character is guaranteed to be at a normalization boundary, |
michael@0 | 63 | * regardless of context. |
michael@0 | 64 | * This is used for moving from one normalization boundary to the next |
michael@0 | 65 | * or preceding boundary, and for performing iterative normalization. |
michael@0 | 66 | * |
michael@0 | 67 | * Iterative normalization is useful when only a small portion of a |
michael@0 | 68 | * longer string needs to be processed. |
michael@0 | 69 | * For example, in ICU, iterative normalization is used by the NormalizationTransliterator |
michael@0 | 70 | * (to avoid replacing already-normalized text) and ucol_nextSortKeyPart() |
michael@0 | 71 | * (to process only the substring for which sort key bytes are computed). |
michael@0 | 72 | * |
michael@0 | 73 | * The set of normalization boundaries returned by these functions may not be |
michael@0 | 74 | * complete: There may be more boundaries that could be returned. |
michael@0 | 75 | * Different functions may return different boundaries. |
michael@0 | 76 | * @stable ICU 4.4 |
michael@0 | 77 | */ |
michael@0 | 78 | class U_COMMON_API Normalizer2 : public UObject { |
michael@0 | 79 | public: |
michael@0 | 80 | /** |
michael@0 | 81 | * Destructor. |
michael@0 | 82 | * @stable ICU 4.4 |
michael@0 | 83 | */ |
michael@0 | 84 | ~Normalizer2(); |
michael@0 | 85 | |
michael@0 | 86 | /** |
michael@0 | 87 | * Returns a Normalizer2 instance for Unicode NFC normalization. |
michael@0 | 88 | * Same as getInstance(NULL, "nfc", UNORM2_COMPOSE, errorCode). |
michael@0 | 89 | * Returns an unmodifiable singleton instance. Do not delete it. |
michael@0 | 90 | * @param errorCode Standard ICU error code. Its input value must |
michael@0 | 91 | * pass the U_SUCCESS() test, or else the function returns |
michael@0 | 92 | * immediately. Check for U_FAILURE() on output or use with |
michael@0 | 93 | * function chaining. (See User Guide for details.) |
michael@0 | 94 | * @return the requested Normalizer2, if successful |
michael@0 | 95 | * @stable ICU 49 |
michael@0 | 96 | */ |
michael@0 | 97 | static const Normalizer2 * |
michael@0 | 98 | getNFCInstance(UErrorCode &errorCode); |
michael@0 | 99 | |
michael@0 | 100 | /** |
michael@0 | 101 | * Returns a Normalizer2 instance for Unicode NFD normalization. |
michael@0 | 102 | * Same as getInstance(NULL, "nfc", UNORM2_DECOMPOSE, errorCode). |
michael@0 | 103 | * Returns an unmodifiable singleton instance. Do not delete it. |
michael@0 | 104 | * @param errorCode Standard ICU error code. Its input value must |
michael@0 | 105 | * pass the U_SUCCESS() test, or else the function returns |
michael@0 | 106 | * immediately. Check for U_FAILURE() on output or use with |
michael@0 | 107 | * function chaining. (See User Guide for details.) |
michael@0 | 108 | * @return the requested Normalizer2, if successful |
michael@0 | 109 | * @stable ICU 49 |
michael@0 | 110 | */ |
michael@0 | 111 | static const Normalizer2 * |
michael@0 | 112 | getNFDInstance(UErrorCode &errorCode); |
michael@0 | 113 | |
michael@0 | 114 | /** |
michael@0 | 115 | * Returns a Normalizer2 instance for Unicode NFKC normalization. |
michael@0 | 116 | * Same as getInstance(NULL, "nfkc", UNORM2_COMPOSE, errorCode). |
michael@0 | 117 | * Returns an unmodifiable singleton instance. Do not delete it. |
michael@0 | 118 | * @param errorCode Standard ICU error code. Its input value must |
michael@0 | 119 | * pass the U_SUCCESS() test, or else the function returns |
michael@0 | 120 | * immediately. Check for U_FAILURE() on output or use with |
michael@0 | 121 | * function chaining. (See User Guide for details.) |
michael@0 | 122 | * @return the requested Normalizer2, if successful |
michael@0 | 123 | * @stable ICU 49 |
michael@0 | 124 | */ |
michael@0 | 125 | static const Normalizer2 * |
michael@0 | 126 | getNFKCInstance(UErrorCode &errorCode); |
michael@0 | 127 | |
michael@0 | 128 | /** |
michael@0 | 129 | * Returns a Normalizer2 instance for Unicode NFKD normalization. |
michael@0 | 130 | * Same as getInstance(NULL, "nfkc", UNORM2_DECOMPOSE, errorCode). |
michael@0 | 131 | * Returns an unmodifiable singleton instance. Do not delete it. |
michael@0 | 132 | * @param errorCode Standard ICU error code. Its input value must |
michael@0 | 133 | * pass the U_SUCCESS() test, or else the function returns |
michael@0 | 134 | * immediately. Check for U_FAILURE() on output or use with |
michael@0 | 135 | * function chaining. (See User Guide for details.) |
michael@0 | 136 | * @return the requested Normalizer2, if successful |
michael@0 | 137 | * @stable ICU 49 |
michael@0 | 138 | */ |
michael@0 | 139 | static const Normalizer2 * |
michael@0 | 140 | getNFKDInstance(UErrorCode &errorCode); |
michael@0 | 141 | |
michael@0 | 142 | /** |
michael@0 | 143 | * Returns a Normalizer2 instance for Unicode NFKC_Casefold normalization. |
michael@0 | 144 | * Same as getInstance(NULL, "nfkc_cf", UNORM2_COMPOSE, errorCode). |
michael@0 | 145 | * Returns an unmodifiable singleton instance. Do not delete it. |
michael@0 | 146 | * @param errorCode Standard ICU error code. Its input value must |
michael@0 | 147 | * pass the U_SUCCESS() test, or else the function returns |
michael@0 | 148 | * immediately. Check for U_FAILURE() on output or use with |
michael@0 | 149 | * function chaining. (See User Guide for details.) |
michael@0 | 150 | * @return the requested Normalizer2, if successful |
michael@0 | 151 | * @stable ICU 49 |
michael@0 | 152 | */ |
michael@0 | 153 | static const Normalizer2 * |
michael@0 | 154 | getNFKCCasefoldInstance(UErrorCode &errorCode); |
michael@0 | 155 | |
michael@0 | 156 | /** |
michael@0 | 157 | * Returns a Normalizer2 instance which uses the specified data file |
michael@0 | 158 | * (packageName/name similar to ucnv_openPackage() and ures_open()/ResourceBundle) |
michael@0 | 159 | * and which composes or decomposes text according to the specified mode. |
michael@0 | 160 | * Returns an unmodifiable singleton instance. Do not delete it. |
michael@0 | 161 | * |
michael@0 | 162 | * Use packageName=NULL for data files that are part of ICU's own data. |
michael@0 | 163 | * Use name="nfc" and UNORM2_COMPOSE/UNORM2_DECOMPOSE for Unicode standard NFC/NFD. |
michael@0 | 164 | * Use name="nfkc" and UNORM2_COMPOSE/UNORM2_DECOMPOSE for Unicode standard NFKC/NFKD. |
michael@0 | 165 | * Use name="nfkc_cf" and UNORM2_COMPOSE for Unicode standard NFKC_CF=NFKC_Casefold. |
michael@0 | 166 | * |
michael@0 | 167 | * @param packageName NULL for ICU built-in data, otherwise application data package name |
michael@0 | 168 | * @param name "nfc" or "nfkc" or "nfkc_cf" or name of custom data file |
michael@0 | 169 | * @param mode normalization mode (compose or decompose etc.) |
michael@0 | 170 | * @param errorCode Standard ICU error code. Its input value must |
michael@0 | 171 | * pass the U_SUCCESS() test, or else the function returns |
michael@0 | 172 | * immediately. Check for U_FAILURE() on output or use with |
michael@0 | 173 | * function chaining. (See User Guide for details.) |
michael@0 | 174 | * @return the requested Normalizer2, if successful |
michael@0 | 175 | * @stable ICU 4.4 |
michael@0 | 176 | */ |
michael@0 | 177 | static const Normalizer2 * |
michael@0 | 178 | getInstance(const char *packageName, |
michael@0 | 179 | const char *name, |
michael@0 | 180 | UNormalization2Mode mode, |
michael@0 | 181 | UErrorCode &errorCode); |
michael@0 | 182 | |
michael@0 | 183 | /** |
michael@0 | 184 | * Returns the normalized form of the source string. |
michael@0 | 185 | * @param src source string |
michael@0 | 186 | * @param errorCode Standard ICU error code. Its input value must |
michael@0 | 187 | * pass the U_SUCCESS() test, or else the function returns |
michael@0 | 188 | * immediately. Check for U_FAILURE() on output or use with |
michael@0 | 189 | * function chaining. (See User Guide for details.) |
michael@0 | 190 | * @return normalized src |
michael@0 | 191 | * @stable ICU 4.4 |
michael@0 | 192 | */ |
michael@0 | 193 | UnicodeString |
michael@0 | 194 | normalize(const UnicodeString &src, UErrorCode &errorCode) const { |
michael@0 | 195 | UnicodeString result; |
michael@0 | 196 | normalize(src, result, errorCode); |
michael@0 | 197 | return result; |
michael@0 | 198 | } |
michael@0 | 199 | /** |
michael@0 | 200 | * Writes the normalized form of the source string to the destination string |
michael@0 | 201 | * (replacing its contents) and returns the destination string. |
michael@0 | 202 | * The source and destination strings must be different objects. |
michael@0 | 203 | * @param src source string |
michael@0 | 204 | * @param dest destination string; its contents is replaced with normalized src |
michael@0 | 205 | * @param errorCode Standard ICU error code. Its input value must |
michael@0 | 206 | * pass the U_SUCCESS() test, or else the function returns |
michael@0 | 207 | * immediately. Check for U_FAILURE() on output or use with |
michael@0 | 208 | * function chaining. (See User Guide for details.) |
michael@0 | 209 | * @return dest |
michael@0 | 210 | * @stable ICU 4.4 |
michael@0 | 211 | */ |
michael@0 | 212 | virtual UnicodeString & |
michael@0 | 213 | normalize(const UnicodeString &src, |
michael@0 | 214 | UnicodeString &dest, |
michael@0 | 215 | UErrorCode &errorCode) const = 0; |
michael@0 | 216 | /** |
michael@0 | 217 | * Appends the normalized form of the second string to the first string |
michael@0 | 218 | * (merging them at the boundary) and returns the first string. |
michael@0 | 219 | * The result is normalized if the first string was normalized. |
michael@0 | 220 | * The first and second strings must be different objects. |
michael@0 | 221 | * @param first string, should be normalized |
michael@0 | 222 | * @param second string, will be normalized |
michael@0 | 223 | * @param errorCode Standard ICU error code. Its input value must |
michael@0 | 224 | * pass the U_SUCCESS() test, or else the function returns |
michael@0 | 225 | * immediately. Check for U_FAILURE() on output or use with |
michael@0 | 226 | * function chaining. (See User Guide for details.) |
michael@0 | 227 | * @return first |
michael@0 | 228 | * @stable ICU 4.4 |
michael@0 | 229 | */ |
michael@0 | 230 | virtual UnicodeString & |
michael@0 | 231 | normalizeSecondAndAppend(UnicodeString &first, |
michael@0 | 232 | const UnicodeString &second, |
michael@0 | 233 | UErrorCode &errorCode) const = 0; |
michael@0 | 234 | /** |
michael@0 | 235 | * Appends the second string to the first string |
michael@0 | 236 | * (merging them at the boundary) and returns the first string. |
michael@0 | 237 | * The result is normalized if both the strings were normalized. |
michael@0 | 238 | * The first and second strings must be different objects. |
michael@0 | 239 | * @param first string, should be normalized |
michael@0 | 240 | * @param second string, should be normalized |
michael@0 | 241 | * @param errorCode Standard ICU error code. Its input value must |
michael@0 | 242 | * pass the U_SUCCESS() test, or else the function returns |
michael@0 | 243 | * immediately. Check for U_FAILURE() on output or use with |
michael@0 | 244 | * function chaining. (See User Guide for details.) |
michael@0 | 245 | * @return first |
michael@0 | 246 | * @stable ICU 4.4 |
michael@0 | 247 | */ |
michael@0 | 248 | virtual UnicodeString & |
michael@0 | 249 | append(UnicodeString &first, |
michael@0 | 250 | const UnicodeString &second, |
michael@0 | 251 | UErrorCode &errorCode) const = 0; |
michael@0 | 252 | |
michael@0 | 253 | /** |
michael@0 | 254 | * Gets the decomposition mapping of c. |
michael@0 | 255 | * Roughly equivalent to normalizing the String form of c |
michael@0 | 256 | * on a UNORM2_DECOMPOSE Normalizer2 instance, but much faster, and except that this function |
michael@0 | 257 | * returns FALSE and does not write a string |
michael@0 | 258 | * if c does not have a decomposition mapping in this instance's data. |
michael@0 | 259 | * This function is independent of the mode of the Normalizer2. |
michael@0 | 260 | * @param c code point |
michael@0 | 261 | * @param decomposition String object which will be set to c's |
michael@0 | 262 | * decomposition mapping, if there is one. |
michael@0 | 263 | * @return TRUE if c has a decomposition, otherwise FALSE |
michael@0 | 264 | * @stable ICU 4.6 |
michael@0 | 265 | */ |
michael@0 | 266 | virtual UBool |
michael@0 | 267 | getDecomposition(UChar32 c, UnicodeString &decomposition) const = 0; |
michael@0 | 268 | |
michael@0 | 269 | /** |
michael@0 | 270 | * Gets the raw decomposition mapping of c. |
michael@0 | 271 | * |
michael@0 | 272 | * This is similar to the getDecomposition() method but returns the |
michael@0 | 273 | * raw decomposition mapping as specified in UnicodeData.txt or |
michael@0 | 274 | * (for custom data) in the mapping files processed by the gennorm2 tool. |
michael@0 | 275 | * By contrast, getDecomposition() returns the processed, |
michael@0 | 276 | * recursively-decomposed version of this mapping. |
michael@0 | 277 | * |
michael@0 | 278 | * When used on a standard NFKC Normalizer2 instance, |
michael@0 | 279 | * getRawDecomposition() returns the Unicode Decomposition_Mapping (dm) property. |
michael@0 | 280 | * |
michael@0 | 281 | * When used on a standard NFC Normalizer2 instance, |
michael@0 | 282 | * it returns the Decomposition_Mapping only if the Decomposition_Type (dt) is Canonical (Can); |
michael@0 | 283 | * in this case, the result contains either one or two code points (=1..4 UChars). |
michael@0 | 284 | * |
michael@0 | 285 | * This function is independent of the mode of the Normalizer2. |
michael@0 | 286 | * The default implementation returns FALSE. |
michael@0 | 287 | * @param c code point |
michael@0 | 288 | * @param decomposition String object which will be set to c's |
michael@0 | 289 | * raw decomposition mapping, if there is one. |
michael@0 | 290 | * @return TRUE if c has a decomposition, otherwise FALSE |
michael@0 | 291 | * @stable ICU 49 |
michael@0 | 292 | */ |
michael@0 | 293 | virtual UBool |
michael@0 | 294 | getRawDecomposition(UChar32 c, UnicodeString &decomposition) const; |
michael@0 | 295 | |
michael@0 | 296 | /** |
michael@0 | 297 | * Performs pairwise composition of a & b and returns the composite if there is one. |
michael@0 | 298 | * |
michael@0 | 299 | * Returns a composite code point c only if c has a two-way mapping to a+b. |
michael@0 | 300 | * In standard Unicode normalization, this means that |
michael@0 | 301 | * c has a canonical decomposition to a+b |
michael@0 | 302 | * and c does not have the Full_Composition_Exclusion property. |
michael@0 | 303 | * |
michael@0 | 304 | * This function is independent of the mode of the Normalizer2. |
michael@0 | 305 | * The default implementation returns a negative value. |
michael@0 | 306 | * @param a A (normalization starter) code point. |
michael@0 | 307 | * @param b Another code point. |
michael@0 | 308 | * @return The non-negative composite code point if there is one; otherwise a negative value. |
michael@0 | 309 | * @stable ICU 49 |
michael@0 | 310 | */ |
michael@0 | 311 | virtual UChar32 |
michael@0 | 312 | composePair(UChar32 a, UChar32 b) const; |
michael@0 | 313 | |
michael@0 | 314 | /** |
michael@0 | 315 | * Gets the combining class of c. |
michael@0 | 316 | * The default implementation returns 0 |
michael@0 | 317 | * but all standard implementations return the Unicode Canonical_Combining_Class value. |
michael@0 | 318 | * @param c code point |
michael@0 | 319 | * @return c's combining class |
michael@0 | 320 | * @stable ICU 49 |
michael@0 | 321 | */ |
michael@0 | 322 | virtual uint8_t |
michael@0 | 323 | getCombiningClass(UChar32 c) const; |
michael@0 | 324 | |
michael@0 | 325 | /** |
michael@0 | 326 | * Tests if the string is normalized. |
michael@0 | 327 | * Internally, in cases where the quickCheck() method would return "maybe" |
michael@0 | 328 | * (which is only possible for the two COMPOSE modes) this method |
michael@0 | 329 | * resolves to "yes" or "no" to provide a definitive result, |
michael@0 | 330 | * at the cost of doing more work in those cases. |
michael@0 | 331 | * @param s input string |
michael@0 | 332 | * @param errorCode Standard ICU error code. Its input value must |
michael@0 | 333 | * pass the U_SUCCESS() test, or else the function returns |
michael@0 | 334 | * immediately. Check for U_FAILURE() on output or use with |
michael@0 | 335 | * function chaining. (See User Guide for details.) |
michael@0 | 336 | * @return TRUE if s is normalized |
michael@0 | 337 | * @stable ICU 4.4 |
michael@0 | 338 | */ |
michael@0 | 339 | virtual UBool |
michael@0 | 340 | isNormalized(const UnicodeString &s, UErrorCode &errorCode) const = 0; |
michael@0 | 341 | |
michael@0 | 342 | /** |
michael@0 | 343 | * Tests if the string is normalized. |
michael@0 | 344 | * For the two COMPOSE modes, the result could be "maybe" in cases that |
michael@0 | 345 | * would take a little more work to resolve definitively. |
michael@0 | 346 | * Use spanQuickCheckYes() and normalizeSecondAndAppend() for a faster |
michael@0 | 347 | * combination of quick check + normalization, to avoid |
michael@0 | 348 | * re-checking the "yes" prefix. |
michael@0 | 349 | * @param s input string |
michael@0 | 350 | * @param errorCode Standard ICU error code. Its input value must |
michael@0 | 351 | * pass the U_SUCCESS() test, or else the function returns |
michael@0 | 352 | * immediately. Check for U_FAILURE() on output or use with |
michael@0 | 353 | * function chaining. (See User Guide for details.) |
michael@0 | 354 | * @return UNormalizationCheckResult |
michael@0 | 355 | * @stable ICU 4.4 |
michael@0 | 356 | */ |
michael@0 | 357 | virtual UNormalizationCheckResult |
michael@0 | 358 | quickCheck(const UnicodeString &s, UErrorCode &errorCode) const = 0; |
michael@0 | 359 | |
michael@0 | 360 | /** |
michael@0 | 361 | * Returns the end of the normalized substring of the input string. |
michael@0 | 362 | * In other words, with <code>end=spanQuickCheckYes(s, ec);</code> |
michael@0 | 363 | * the substring <code>UnicodeString(s, 0, end)</code> |
michael@0 | 364 | * will pass the quick check with a "yes" result. |
michael@0 | 365 | * |
michael@0 | 366 | * The returned end index is usually one or more characters before the |
michael@0 | 367 | * "no" or "maybe" character: The end index is at a normalization boundary. |
michael@0 | 368 | * (See the class documentation for more about normalization boundaries.) |
michael@0 | 369 | * |
michael@0 | 370 | * When the goal is a normalized string and most input strings are expected |
michael@0 | 371 | * to be normalized already, then call this method, |
michael@0 | 372 | * and if it returns a prefix shorter than the input string, |
michael@0 | 373 | * copy that prefix and use normalizeSecondAndAppend() for the remainder. |
michael@0 | 374 | * @param s input string |
michael@0 | 375 | * @param errorCode Standard ICU error code. Its input value must |
michael@0 | 376 | * pass the U_SUCCESS() test, or else the function returns |
michael@0 | 377 | * immediately. Check for U_FAILURE() on output or use with |
michael@0 | 378 | * function chaining. (See User Guide for details.) |
michael@0 | 379 | * @return "yes" span end index |
michael@0 | 380 | * @stable ICU 4.4 |
michael@0 | 381 | */ |
michael@0 | 382 | virtual int32_t |
michael@0 | 383 | spanQuickCheckYes(const UnicodeString &s, UErrorCode &errorCode) const = 0; |
michael@0 | 384 | |
michael@0 | 385 | /** |
michael@0 | 386 | * Tests if the character always has a normalization boundary before it, |
michael@0 | 387 | * regardless of context. |
michael@0 | 388 | * If true, then the character does not normalization-interact with |
michael@0 | 389 | * preceding characters. |
michael@0 | 390 | * In other words, a string containing this character can be normalized |
michael@0 | 391 | * by processing portions before this character and starting from this |
michael@0 | 392 | * character independently. |
michael@0 | 393 | * This is used for iterative normalization. See the class documentation for details. |
michael@0 | 394 | * @param c character to test |
michael@0 | 395 | * @return TRUE if c has a normalization boundary before it |
michael@0 | 396 | * @stable ICU 4.4 |
michael@0 | 397 | */ |
michael@0 | 398 | virtual UBool hasBoundaryBefore(UChar32 c) const = 0; |
michael@0 | 399 | |
michael@0 | 400 | /** |
michael@0 | 401 | * Tests if the character always has a normalization boundary after it, |
michael@0 | 402 | * regardless of context. |
michael@0 | 403 | * If true, then the character does not normalization-interact with |
michael@0 | 404 | * following characters. |
michael@0 | 405 | * In other words, a string containing this character can be normalized |
michael@0 | 406 | * by processing portions up to this character and after this |
michael@0 | 407 | * character independently. |
michael@0 | 408 | * This is used for iterative normalization. See the class documentation for details. |
michael@0 | 409 | * Note that this operation may be significantly slower than hasBoundaryBefore(). |
michael@0 | 410 | * @param c character to test |
michael@0 | 411 | * @return TRUE if c has a normalization boundary after it |
michael@0 | 412 | * @stable ICU 4.4 |
michael@0 | 413 | */ |
michael@0 | 414 | virtual UBool hasBoundaryAfter(UChar32 c) const = 0; |
michael@0 | 415 | |
michael@0 | 416 | /** |
michael@0 | 417 | * Tests if the character is normalization-inert. |
michael@0 | 418 | * If true, then the character does not change, nor normalization-interact with |
michael@0 | 419 | * preceding or following characters. |
michael@0 | 420 | * In other words, a string containing this character can be normalized |
michael@0 | 421 | * by processing portions before this character and after this |
michael@0 | 422 | * character independently. |
michael@0 | 423 | * This is used for iterative normalization. See the class documentation for details. |
michael@0 | 424 | * Note that this operation may be significantly slower than hasBoundaryBefore(). |
michael@0 | 425 | * @param c character to test |
michael@0 | 426 | * @return TRUE if c is normalization-inert |
michael@0 | 427 | * @stable ICU 4.4 |
michael@0 | 428 | */ |
michael@0 | 429 | virtual UBool isInert(UChar32 c) const = 0; |
michael@0 | 430 | }; |
michael@0 | 431 | |
michael@0 | 432 | /** |
michael@0 | 433 | * Normalization filtered by a UnicodeSet. |
michael@0 | 434 | * Normalizes portions of the text contained in the filter set and leaves |
michael@0 | 435 | * portions not contained in the filter set unchanged. |
michael@0 | 436 | * Filtering is done via UnicodeSet::span(..., USET_SPAN_SIMPLE). |
michael@0 | 437 | * Not-in-the-filter text is treated as "is normalized" and "quick check yes". |
michael@0 | 438 | * This class implements all of (and only) the Normalizer2 API. |
michael@0 | 439 | * An instance of this class is unmodifiable/immutable but is constructed and |
michael@0 | 440 | * must be destructed by the owner. |
michael@0 | 441 | * @stable ICU 4.4 |
michael@0 | 442 | */ |
michael@0 | 443 | class U_COMMON_API FilteredNormalizer2 : public Normalizer2 { |
michael@0 | 444 | public: |
michael@0 | 445 | /** |
michael@0 | 446 | * Constructs a filtered normalizer wrapping any Normalizer2 instance |
michael@0 | 447 | * and a filter set. |
michael@0 | 448 | * Both are aliased and must not be modified or deleted while this object |
michael@0 | 449 | * is used. |
michael@0 | 450 | * The filter set should be frozen; otherwise the performance will suffer greatly. |
michael@0 | 451 | * @param n2 wrapped Normalizer2 instance |
michael@0 | 452 | * @param filterSet UnicodeSet which determines the characters to be normalized |
michael@0 | 453 | * @stable ICU 4.4 |
michael@0 | 454 | */ |
michael@0 | 455 | FilteredNormalizer2(const Normalizer2 &n2, const UnicodeSet &filterSet) : |
michael@0 | 456 | norm2(n2), set(filterSet) {} |
michael@0 | 457 | |
michael@0 | 458 | /** |
michael@0 | 459 | * Destructor. |
michael@0 | 460 | * @stable ICU 4.4 |
michael@0 | 461 | */ |
michael@0 | 462 | ~FilteredNormalizer2(); |
michael@0 | 463 | |
michael@0 | 464 | /** |
michael@0 | 465 | * Writes the normalized form of the source string to the destination string |
michael@0 | 466 | * (replacing its contents) and returns the destination string. |
michael@0 | 467 | * The source and destination strings must be different objects. |
michael@0 | 468 | * @param src source string |
michael@0 | 469 | * @param dest destination string; its contents is replaced with normalized src |
michael@0 | 470 | * @param errorCode Standard ICU error code. Its input value must |
michael@0 | 471 | * pass the U_SUCCESS() test, or else the function returns |
michael@0 | 472 | * immediately. Check for U_FAILURE() on output or use with |
michael@0 | 473 | * function chaining. (See User Guide for details.) |
michael@0 | 474 | * @return dest |
michael@0 | 475 | * @stable ICU 4.4 |
michael@0 | 476 | */ |
michael@0 | 477 | virtual UnicodeString & |
michael@0 | 478 | normalize(const UnicodeString &src, |
michael@0 | 479 | UnicodeString &dest, |
michael@0 | 480 | UErrorCode &errorCode) const; |
michael@0 | 481 | /** |
michael@0 | 482 | * Appends the normalized form of the second string to the first string |
michael@0 | 483 | * (merging them at the boundary) and returns the first string. |
michael@0 | 484 | * The result is normalized if the first string was normalized. |
michael@0 | 485 | * The first and second strings must be different objects. |
michael@0 | 486 | * @param first string, should be normalized |
michael@0 | 487 | * @param second string, will be normalized |
michael@0 | 488 | * @param errorCode Standard ICU error code. Its input value must |
michael@0 | 489 | * pass the U_SUCCESS() test, or else the function returns |
michael@0 | 490 | * immediately. Check for U_FAILURE() on output or use with |
michael@0 | 491 | * function chaining. (See User Guide for details.) |
michael@0 | 492 | * @return first |
michael@0 | 493 | * @stable ICU 4.4 |
michael@0 | 494 | */ |
michael@0 | 495 | virtual UnicodeString & |
michael@0 | 496 | normalizeSecondAndAppend(UnicodeString &first, |
michael@0 | 497 | const UnicodeString &second, |
michael@0 | 498 | UErrorCode &errorCode) const; |
michael@0 | 499 | /** |
michael@0 | 500 | * Appends the second string to the first string |
michael@0 | 501 | * (merging them at the boundary) and returns the first string. |
michael@0 | 502 | * The result is normalized if both the strings were normalized. |
michael@0 | 503 | * The first and second strings must be different objects. |
michael@0 | 504 | * @param first string, should be normalized |
michael@0 | 505 | * @param second string, should be normalized |
michael@0 | 506 | * @param errorCode Standard ICU error code. Its input value must |
michael@0 | 507 | * pass the U_SUCCESS() test, or else the function returns |
michael@0 | 508 | * immediately. Check for U_FAILURE() on output or use with |
michael@0 | 509 | * function chaining. (See User Guide for details.) |
michael@0 | 510 | * @return first |
michael@0 | 511 | * @stable ICU 4.4 |
michael@0 | 512 | */ |
michael@0 | 513 | virtual UnicodeString & |
michael@0 | 514 | append(UnicodeString &first, |
michael@0 | 515 | const UnicodeString &second, |
michael@0 | 516 | UErrorCode &errorCode) const; |
michael@0 | 517 | |
michael@0 | 518 | /** |
michael@0 | 519 | * Gets the decomposition mapping of c. |
michael@0 | 520 | * For details see the base class documentation. |
michael@0 | 521 | * |
michael@0 | 522 | * This function is independent of the mode of the Normalizer2. |
michael@0 | 523 | * @param c code point |
michael@0 | 524 | * @param decomposition String object which will be set to c's |
michael@0 | 525 | * decomposition mapping, if there is one. |
michael@0 | 526 | * @return TRUE if c has a decomposition, otherwise FALSE |
michael@0 | 527 | * @stable ICU 4.6 |
michael@0 | 528 | */ |
michael@0 | 529 | virtual UBool |
michael@0 | 530 | getDecomposition(UChar32 c, UnicodeString &decomposition) const; |
michael@0 | 531 | |
michael@0 | 532 | /** |
michael@0 | 533 | * Gets the raw decomposition mapping of c. |
michael@0 | 534 | * For details see the base class documentation. |
michael@0 | 535 | * |
michael@0 | 536 | * This function is independent of the mode of the Normalizer2. |
michael@0 | 537 | * @param c code point |
michael@0 | 538 | * @param decomposition String object which will be set to c's |
michael@0 | 539 | * raw decomposition mapping, if there is one. |
michael@0 | 540 | * @return TRUE if c has a decomposition, otherwise FALSE |
michael@0 | 541 | * @stable ICU 49 |
michael@0 | 542 | */ |
michael@0 | 543 | virtual UBool |
michael@0 | 544 | getRawDecomposition(UChar32 c, UnicodeString &decomposition) const; |
michael@0 | 545 | |
michael@0 | 546 | /** |
michael@0 | 547 | * Performs pairwise composition of a & b and returns the composite if there is one. |
michael@0 | 548 | * For details see the base class documentation. |
michael@0 | 549 | * |
michael@0 | 550 | * This function is independent of the mode of the Normalizer2. |
michael@0 | 551 | * @param a A (normalization starter) code point. |
michael@0 | 552 | * @param b Another code point. |
michael@0 | 553 | * @return The non-negative composite code point if there is one; otherwise a negative value. |
michael@0 | 554 | * @stable ICU 49 |
michael@0 | 555 | */ |
michael@0 | 556 | virtual UChar32 |
michael@0 | 557 | composePair(UChar32 a, UChar32 b) const; |
michael@0 | 558 | |
michael@0 | 559 | /** |
michael@0 | 560 | * Gets the combining class of c. |
michael@0 | 561 | * The default implementation returns 0 |
michael@0 | 562 | * but all standard implementations return the Unicode Canonical_Combining_Class value. |
michael@0 | 563 | * @param c code point |
michael@0 | 564 | * @return c's combining class |
michael@0 | 565 | * @stable ICU 49 |
michael@0 | 566 | */ |
michael@0 | 567 | virtual uint8_t |
michael@0 | 568 | getCombiningClass(UChar32 c) const; |
michael@0 | 569 | |
michael@0 | 570 | /** |
michael@0 | 571 | * Tests if the string is normalized. |
michael@0 | 572 | * For details see the Normalizer2 base class documentation. |
michael@0 | 573 | * @param s input string |
michael@0 | 574 | * @param errorCode Standard ICU error code. Its input value must |
michael@0 | 575 | * pass the U_SUCCESS() test, or else the function returns |
michael@0 | 576 | * immediately. Check for U_FAILURE() on output or use with |
michael@0 | 577 | * function chaining. (See User Guide for details.) |
michael@0 | 578 | * @return TRUE if s is normalized |
michael@0 | 579 | * @stable ICU 4.4 |
michael@0 | 580 | */ |
michael@0 | 581 | virtual UBool |
michael@0 | 582 | isNormalized(const UnicodeString &s, UErrorCode &errorCode) const; |
michael@0 | 583 | /** |
michael@0 | 584 | * Tests if the string is normalized. |
michael@0 | 585 | * For details see the Normalizer2 base class documentation. |
michael@0 | 586 | * @param s input string |
michael@0 | 587 | * @param errorCode Standard ICU error code. Its input value must |
michael@0 | 588 | * pass the U_SUCCESS() test, or else the function returns |
michael@0 | 589 | * immediately. Check for U_FAILURE() on output or use with |
michael@0 | 590 | * function chaining. (See User Guide for details.) |
michael@0 | 591 | * @return UNormalizationCheckResult |
michael@0 | 592 | * @stable ICU 4.4 |
michael@0 | 593 | */ |
michael@0 | 594 | virtual UNormalizationCheckResult |
michael@0 | 595 | quickCheck(const UnicodeString &s, UErrorCode &errorCode) const; |
michael@0 | 596 | /** |
michael@0 | 597 | * Returns the end of the normalized substring of the input string. |
michael@0 | 598 | * For details see the Normalizer2 base class documentation. |
michael@0 | 599 | * @param s input string |
michael@0 | 600 | * @param errorCode Standard ICU error code. Its input value must |
michael@0 | 601 | * pass the U_SUCCESS() test, or else the function returns |
michael@0 | 602 | * immediately. Check for U_FAILURE() on output or use with |
michael@0 | 603 | * function chaining. (See User Guide for details.) |
michael@0 | 604 | * @return "yes" span end index |
michael@0 | 605 | * @stable ICU 4.4 |
michael@0 | 606 | */ |
michael@0 | 607 | virtual int32_t |
michael@0 | 608 | spanQuickCheckYes(const UnicodeString &s, UErrorCode &errorCode) const; |
michael@0 | 609 | |
michael@0 | 610 | /** |
michael@0 | 611 | * Tests if the character always has a normalization boundary before it, |
michael@0 | 612 | * regardless of context. |
michael@0 | 613 | * For details see the Normalizer2 base class documentation. |
michael@0 | 614 | * @param c character to test |
michael@0 | 615 | * @return TRUE if c has a normalization boundary before it |
michael@0 | 616 | * @stable ICU 4.4 |
michael@0 | 617 | */ |
michael@0 | 618 | virtual UBool hasBoundaryBefore(UChar32 c) const; |
michael@0 | 619 | |
michael@0 | 620 | /** |
michael@0 | 621 | * Tests if the character always has a normalization boundary after it, |
michael@0 | 622 | * regardless of context. |
michael@0 | 623 | * For details see the Normalizer2 base class documentation. |
michael@0 | 624 | * @param c character to test |
michael@0 | 625 | * @return TRUE if c has a normalization boundary after it |
michael@0 | 626 | * @stable ICU 4.4 |
michael@0 | 627 | */ |
michael@0 | 628 | virtual UBool hasBoundaryAfter(UChar32 c) const; |
michael@0 | 629 | |
michael@0 | 630 | /** |
michael@0 | 631 | * Tests if the character is normalization-inert. |
michael@0 | 632 | * For details see the Normalizer2 base class documentation. |
michael@0 | 633 | * @param c character to test |
michael@0 | 634 | * @return TRUE if c is normalization-inert |
michael@0 | 635 | * @stable ICU 4.4 |
michael@0 | 636 | */ |
michael@0 | 637 | virtual UBool isInert(UChar32 c) const; |
michael@0 | 638 | private: |
michael@0 | 639 | UnicodeString & |
michael@0 | 640 | normalize(const UnicodeString &src, |
michael@0 | 641 | UnicodeString &dest, |
michael@0 | 642 | USetSpanCondition spanCondition, |
michael@0 | 643 | UErrorCode &errorCode) const; |
michael@0 | 644 | |
michael@0 | 645 | UnicodeString & |
michael@0 | 646 | normalizeSecondAndAppend(UnicodeString &first, |
michael@0 | 647 | const UnicodeString &second, |
michael@0 | 648 | UBool doNormalize, |
michael@0 | 649 | UErrorCode &errorCode) const; |
michael@0 | 650 | |
michael@0 | 651 | const Normalizer2 &norm2; |
michael@0 | 652 | const UnicodeSet &set; |
michael@0 | 653 | }; |
michael@0 | 654 | |
michael@0 | 655 | U_NAMESPACE_END |
michael@0 | 656 | |
michael@0 | 657 | #endif // !UCONFIG_NO_NORMALIZATION |
michael@0 | 658 | #endif // __NORMALIZER2_H__ |