1 // © 2016 and later: Unicode, Inc. and others. 2 // License & terms of use: http://www.unicode.org/copyright.html 4 ******************************************************************************* 6 * Copyright (C) 2009-2015, International Business Machines 7 * Corporation and others. All Rights Reserved. 9 ******************************************************************************* 12 * tab size: 8 (not used) 15 * created on: 2009dec15 16 * created by: Markus W. Scherer 37 #if U_SHOW_CPLUSPLUS_API 39 #endif// U_SHOW_CPLUSPLUS_API 125 #if !UCONFIG_NO_NORMALIZATION 271 #if U_SHOW_CPLUSPLUS_API 308 constUChar *src, int32_t length,
309 UChar *dest, int32_t capacity,
331 UChar *first, int32_t firstLength, int32_t firstCapacity,
332 constUChar *second, int32_t secondLength,
354 UChar *first, int32_t firstLength, int32_t firstCapacity,
355 constUChar *second, int32_t secondLength,
464 constUChar *s, int32_t length,
486 constUChar *s, int32_t length,
515 constUChar *s, int32_t length,
621 constUChar *s2, int32_t length2,
625 #endif/* !UCONFIG_NO_NORMALIZATION */ 626 #endif/* __UNORM2_H__ */ "Smart pointer" class, closes a UNormalizer2 via unorm2_close().
C++ API: "Smart pointers" for use with and in ICU4C C++ code.
#define U_DEFINE_LOCAL_OPEN_POINTER(LocalPointerClassName, Type, closeFunction)
"Smart pointer" definition macro, deletes objects via the closeFunction.
C API: Bit set option bit constants for various string and character processing functions.
int32_t UChar32
Define UChar32 as a type for single Unicode code points.
int8_t UBool
The ICU boolean type, a signed-byte integer.
#define U_CAPI
This is used to declare a function as a public ICU C API.
char16_t UChar
The base type for UTF-16 code units and pointers.
UNormalizationCheckResult
Result values for normalization quick check functions.
@ UNORM_YES
The input string is in the normalization form.
@ UNORM_MAYBE
The input string may or may not be in the normalization form.
@ UNORM_NO
The input string is not in the normalization form.
U_CAPI int32_t unorm2_getRawDecomposition(const UNormalizer2 *norm2, UChar32 c, UChar *decomposition, int32_t capacity, UErrorCode *pErrorCode)
Gets the raw decomposition mapping of c.
U_CAPI UNormalizer2 * unorm2_openFiltered(const UNormalizer2 *norm2, const USet *filterSet, UErrorCode *pErrorCode)
Constructs a filtered normalizer wrapping any UNormalizer2 instance and a filter set.
U_CAPI uint8_t unorm2_getCombiningClass(const UNormalizer2 *norm2, UChar32 c)
Gets the combining class of c.
U_CAPI UBool unorm2_isNormalized(const UNormalizer2 *norm2, const UChar *s, int32_t length, UErrorCode *pErrorCode)
Tests if the string is normalized.
U_CAPI int32_t unorm2_normalize(const UNormalizer2 *norm2, const UChar *src, int32_t length, UChar *dest, int32_t capacity, UErrorCode *pErrorCode)
Writes the normalized form of the source string to the destination string (replacing its contents) an...
U_CAPI int32_t unorm2_append(const UNormalizer2 *norm2, UChar *first, int32_t firstLength, int32_t firstCapacity, const UChar *second, int32_t secondLength, UErrorCode *pErrorCode)
Appends the second string to the first string (merging them at the boundary) and returns the length o...
U_CAPI UBool unorm2_hasBoundaryAfter(const UNormalizer2 *norm2, UChar32 c)
Tests if the character always has a normalization boundary after it, regardless of context.
U_CAPI void unorm2_close(UNormalizer2 *norm2)
Closes a UNormalizer2 instance from unorm2_openFiltered().
UNormalization2Mode
Constants for normalization modes.
@ UNORM2_COMPOSE
Decomposition followed by composition.
@ UNORM2_FCD
"Fast C or D" form.
@ UNORM2_COMPOSE_CONTIGUOUS
Compose only contiguously.
@ UNORM2_DECOMPOSE
Map, and reorder canonically.
U_CAPI const UNormalizer2 * unorm2_getNFKCInstance(UErrorCode *pErrorCode)
Returns a UNormalizer2 instance for Unicode NFKC normalization.
U_CAPI int32_t unorm2_spanQuickCheckYes(const UNormalizer2 *norm2, const UChar *s, int32_t length, UErrorCode *pErrorCode)
Returns the end of the normalized substring of the input string.
U_CAPI UChar32 unorm2_composePair(const UNormalizer2 *norm2, UChar32 a, UChar32 b)
Performs pairwise composition of a & b and returns the composite if there is one.
U_CAPI const UNormalizer2 * unorm2_getInstance(const char *packageName, const char *name, UNormalization2Mode mode, UErrorCode *pErrorCode)
Returns a UNormalizer2 instance which uses the specified data file (packageName/name similar to ucnv_...
U_CAPI const UNormalizer2 * unorm2_getNFKCSimpleCasefoldInstance(UErrorCode *pErrorCode)
Returns a UNormalizer2 instance for a variant of Unicode toNFKC_Casefold() normalization which is equ...
U_CAPI const UNormalizer2 * unorm2_getNFKCCasefoldInstance(UErrorCode *pErrorCode)
Returns a UNormalizer2 instance for Unicode toNFKC_Casefold() normalization which is equivalent to ap...
U_CAPI UBool unorm2_hasBoundaryBefore(const UNormalizer2 *norm2, UChar32 c)
Tests if the character always has a normalization boundary before it, regardless of context.
U_CAPI UBool unorm2_isInert(const UNormalizer2 *norm2, UChar32 c)
Tests if the character is normalization-inert.
U_CAPI const UNormalizer2 * unorm2_getNFDInstance(UErrorCode *pErrorCode)
Returns a UNormalizer2 instance for Unicode NFD normalization.
U_CAPI int32_t unorm2_getDecomposition(const UNormalizer2 *norm2, UChar32 c, UChar *decomposition, int32_t capacity, UErrorCode *pErrorCode)
Gets the decomposition mapping of c.
U_CAPI int32_t unorm2_normalizeSecondAndAppend(const UNormalizer2 *norm2, UChar *first, int32_t firstLength, int32_t firstCapacity, const UChar *second, int32_t secondLength, UErrorCode *pErrorCode)
Appends the normalized form of the second string to the first string (merging them at the boundary) a...
U_CAPI UNormalizationCheckResult unorm2_quickCheck(const UNormalizer2 *norm2, const UChar *s, int32_t length, UErrorCode *pErrorCode)
Tests if the string is normalized.
U_CAPI const UNormalizer2 * unorm2_getNFKDInstance(UErrorCode *pErrorCode)
Returns a UNormalizer2 instance for Unicode NFKD normalization.
U_CAPI int32_t unorm_compare(const UChar *s1, int32_t length1, const UChar *s2, int32_t length2, uint32_t options, UErrorCode *pErrorCode)
Compares two strings for canonical equivalence.
U_CAPI const UNormalizer2 * unorm2_getNFCInstance(UErrorCode *pErrorCode)
Returns a UNormalizer2 instance for Unicode NFC normalization.
struct UNormalizer2 UNormalizer2
C typedef for struct UNormalizer2.
struct USet USet
USet is the C API type corresponding to C++ class UnicodeSet.
Basic definitions for ICU, for both C and C++ APIs.
UErrorCode
Standard ICU4C error code type, a substitute for exceptions.
