1 // © 2016 and later: Unicode, Inc. and others. 2 // License & terms of use: http://www.unicode.org/copyright.html 4 ******************************************************************************* 6 * Copyright (C) 2009-2013, International Business Machines 7 * Corporation and others. All Rights Reserved. 9 ******************************************************************************* 10 * file name: normalizer2.h 12 * tab size: 8 (not used) 15 * created on: 2009nov22 16 * created by: Markus W. Scherer 19 #ifndef __NORMALIZER2_H__ 20 #define __NORMALIZER2_H__ 29 #if U_SHOW_CPLUSPLUS_API 31 #if !UCONFIG_NO_NORMALIZATION 223 normalize(src, result, errorCode);
532 norm2(n2), set(filterSet) {}
770 normalizeUTF8(uint32_t options,
constchar *src, int32_t length,
787 #endif// !UCONFIG_NO_NORMALIZATION 789 #endif/* U_SHOW_CPLUSPLUS_API */ 791 #endif// __NORMALIZER2_H__ A ByteSink can be filled with bytes.
Records lengths of string edits but not replacement text.
Normalization filtered by a UnicodeSet.
virtual UnicodeString & normalize(const UnicodeString &src, UnicodeString &dest, UErrorCode &errorCode) const override
Writes the normalized form of the source string to the destination string (replacing its contents) an...
virtual UBool isNormalizedUTF8(StringPiece s, UErrorCode &errorCode) const override
Tests if the UTF-8 string is normalized.
~FilteredNormalizer2()
Destructor.
virtual UnicodeString & normalizeSecondAndAppend(UnicodeString &first, const UnicodeString &second, UErrorCode &errorCode) const override
Appends the normalized form of the second string to the first string (merging them at the boundary) a...
virtual UNormalizationCheckResult quickCheck(const UnicodeString &s, UErrorCode &errorCode) const override
Tests if the string is normalized.
virtual UBool isInert(UChar32 c) const override
Tests if the character is normalization-inert.
virtual UChar32 composePair(UChar32 a, UChar32 b) const override
Performs pairwise composition of a & b and returns the composite if there is one.
virtual UBool getDecomposition(UChar32 c, UnicodeString &decomposition) const override
Gets the decomposition mapping of c.
virtual UBool isNormalized(const UnicodeString &s, UErrorCode &errorCode) const override
Tests if the string is normalized.
virtual UBool getRawDecomposition(UChar32 c, UnicodeString &decomposition) const override
Gets the raw decomposition mapping of c.
virtual uint8_t getCombiningClass(UChar32 c) const override
Gets the combining class of c.
virtual void normalizeUTF8(uint32_t options, StringPiece src, ByteSink &sink, Edits *edits, UErrorCode &errorCode) const override
Normalizes a UTF-8 string and optionally records how source substrings relate to changed and unchange...
virtual UBool hasBoundaryBefore(UChar32 c) const override
Tests if the character always has a normalization boundary before it, regardless of context.
virtual UnicodeString & append(UnicodeString &first, const UnicodeString &second, UErrorCode &errorCode) const override
Appends the second string to the first string (merging them at the boundary) and returns the first st...
FilteredNormalizer2(const Normalizer2 &n2, const UnicodeSet &filterSet)
Constructs a filtered normalizer wrapping any Normalizer2 instance and a filter set.
virtual UBool hasBoundaryAfter(UChar32 c) const override
Tests if the character always has a normalization boundary after it, regardless of context.
virtual int32_t spanQuickCheckYes(const UnicodeString &s, UErrorCode &errorCode) const override
Returns the end of the normalized substring of the input string.
Unicode normalization functionality for standard Unicode normalization or for using custom mapping ta...
virtual UBool getDecomposition(UChar32 c, UnicodeString &decomposition) const =0
Gets the decomposition mapping of c.
virtual int32_t spanQuickCheckYes(const UnicodeString &s, UErrorCode &errorCode) const =0
Returns the end of the normalized substring of the input string.
virtual UBool isNormalized(const UnicodeString &s, UErrorCode &errorCode) const =0
Tests if the string is normalized.
virtual UBool hasBoundaryBefore(UChar32 c) const =0
Tests if the character always has a normalization boundary before it, regardless of context.
static const Normalizer2 * getNFKCInstance(UErrorCode &errorCode)
Returns a Normalizer2 instance for Unicode NFKC normalization.
~Normalizer2()
Destructor.
virtual UChar32 composePair(UChar32 a, UChar32 b) const
Performs pairwise composition of a & b and returns the composite if there is one.
static const Normalizer2 * getNFCInstance(UErrorCode &errorCode)
Returns a Normalizer2 instance for Unicode NFC normalization.
virtual UnicodeString & normalize(const UnicodeString &src, UnicodeString &dest, UErrorCode &errorCode) const =0
Writes the normalized form of the source string to the destination string (replacing its contents) an...
static const Normalizer2 * getInstance(const char *packageName, const char *name, UNormalization2Mode mode, UErrorCode &errorCode)
Returns a Normalizer2 instance which uses the specified data file (packageName/name similar to ucnv_o...
virtual UnicodeString & append(UnicodeString &first, const UnicodeString &second, UErrorCode &errorCode) const =0
Appends the second string to the first string (merging them at the boundary) and returns the first st...
virtual UnicodeString & normalizeSecondAndAppend(UnicodeString &first, const UnicodeString &second, UErrorCode &errorCode) const =0
Appends the normalized form of the second string to the first string (merging them at the boundary) a...
UnicodeString normalize(const UnicodeString &src, UErrorCode &errorCode) const
Returns the normalized form of the source string.
virtual UNormalizationCheckResult quickCheck(const UnicodeString &s, UErrorCode &errorCode) const =0
Tests if the string is normalized.
virtual UBool hasBoundaryAfter(UChar32 c) const =0
Tests if the character always has a normalization boundary after it, regardless of context.
virtual uint8_t getCombiningClass(UChar32 c) const
Gets the combining class of c.
static const Normalizer2 * getNFKCSimpleCasefoldInstance(UErrorCode &errorCode)
Returns a Normalizer2 instance for a variant of Unicode toNFKC_Casefold() normalization which is equi...
virtual UBool isInert(UChar32 c) const =0
Tests if the character is normalization-inert.
static const Normalizer2 * getNFKCCasefoldInstance(UErrorCode &errorCode)
Returns a Normalizer2 instance for Unicode toNFKC_Casefold() normalization which is equivalent to app...
virtual void normalizeUTF8(uint32_t options, StringPiece src, ByteSink &sink, Edits *edits, UErrorCode &errorCode) const
Normalizes a UTF-8 string and optionally records how source substrings relate to changed and unchange...
virtual UBool isNormalizedUTF8(StringPiece s, UErrorCode &errorCode) const
Tests if the UTF-8 string is normalized.
virtual UBool getRawDecomposition(UChar32 c, UnicodeString &decomposition) const
Gets the raw decomposition mapping of c.
static const Normalizer2 * getNFDInstance(UErrorCode &errorCode)
Returns a Normalizer2 instance for Unicode NFD normalization.
static const Normalizer2 * getNFKDInstance(UErrorCode &errorCode)
Returns a Normalizer2 instance for Unicode NFKD normalization.
A string-like object that points to a sized piece of memory.
UObject is the common ICU "boilerplate" class.
A mutable set of Unicode characters and multicharacter strings.
UnicodeString is a string class that stores Unicode characters directly and provides similar function...
C++ API: StringPiece: Read-only byte string wrapper class.
int32_t UChar32
Define UChar32 as a type for single Unicode code points.
int8_t UBool
The ICU boolean type, a signed-byte integer.
C API: New API for Unicode Normalization.
UNormalizationCheckResult
Result values for normalization quick check functions.
UNormalization2Mode
Constants for normalization modes.
USetSpanCondition
Argument values for whether span() and similar functions continue while the current character is cont...
Basic definitions for ICU, for both C and C++ APIs.
UErrorCode
Standard ICU4C error code type, a substitute for exceptions.
#define U_COMMON_API
Set to export library symbols from inside the common library, and to import them from outside.
