icuSources/common/unicode/normalizer2.h

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