diff options
Diffstat (limited to 'thirdparty/icu4c/common/unicode/uniset.h')
| -rw-r--r-- | thirdparty/icu4c/common/unicode/uniset.h | 45 |
1 files changed, 29 insertions, 16 deletions
diff --git a/thirdparty/icu4c/common/unicode/uniset.h b/thirdparty/icu4c/common/unicode/uniset.h index 33e35c4def..84774d9f36 100644 --- a/thirdparty/icu4c/common/unicode/uniset.h +++ b/thirdparty/icu4c/common/unicode/uniset.h @@ -282,7 +282,7 @@ class RuleCharacterIterator; * @author Alan Liu * @stable ICU 2.0 */ -class U_COMMON_API UnicodeSet U_FINAL : public UnicodeFilter { +class U_COMMON_API UnicodeSet final : public UnicodeFilter { private: /** * Enough for sets with few ranges. @@ -297,8 +297,8 @@ private: int32_t len = 1; // length of list used; 1 <= len <= capacity uint8_t fFlags = 0; // Bit flag (see constants above) - BMPSet *bmpSet = nullptr; // The set is frozen iff either bmpSet or stringSpan is not NULL. - UChar32* buffer = nullptr; // internal buffer, may be NULL + BMPSet *bmpSet = nullptr; // The set is frozen iff either bmpSet or stringSpan is not nullptr. + UChar32* buffer = nullptr; // internal buffer, may be nullptr int32_t bufferCapacity = 0; // capacity of buffer /** @@ -430,9 +430,11 @@ public: * description for the syntax of the pattern language. * @param pattern a string specifying what characters are in the set * @param options bitmask for options to apply to the pattern. - * Valid options are USET_IGNORE_SPACE and USET_CASE_INSENSITIVE. + * Valid options are USET_IGNORE_SPACE and + * at most one of USET_CASE_INSENSITIVE, USET_ADD_CASE_MAPPINGS, USET_SIMPLE_CASE_INSENSITIVE. + * These case options are mutually exclusive. * @param symbols a symbol table mapping variable names to values - * and stand-in characters to UnicodeSets; may be NULL + * and stand-in characters to UnicodeSets; may be nullptr * @param status returns <code>U_ILLEGAL_ARGUMENT_ERROR</code> if the pattern * contains a syntax error. * @internal @@ -450,9 +452,11 @@ public: * @param pos on input, the position in pattern at which to start parsing. * On output, the position after the last character parsed. * @param options bitmask for options to apply to the pattern. - * Valid options are USET_IGNORE_SPACE and USET_CASE_INSENSITIVE. + * Valid options are USET_IGNORE_SPACE and + * at most one of USET_CASE_INSENSITIVE, USET_ADD_CASE_MAPPINGS, USET_SIMPLE_CASE_INSENSITIVE. + * These case options are mutually exclusive. * @param symbols a symbol table mapping variable names to values - * and stand-in characters to UnicodeSets; may be NULL + * and stand-in characters to UnicodeSets; may be nullptr * @param status input-output error code * @stable ICU 2.8 */ @@ -645,9 +649,11 @@ public: * A frozen set will not be modified. * @param pattern a string specifying what characters are in the set * @param options bitmask for options to apply to the pattern. - * Valid options are USET_IGNORE_SPACE and USET_CASE_INSENSITIVE. + * Valid options are USET_IGNORE_SPACE and + * at most one of USET_CASE_INSENSITIVE, USET_ADD_CASE_MAPPINGS, USET_SIMPLE_CASE_INSENSITIVE. + * These case options are mutually exclusive. * @param symbols a symbol table mapping variable names to - * values and stand-ins to UnicodeSets; may be NULL + * values and stand-ins to UnicodeSets; may be nullptr * @param status returns <code>U_ILLEGAL_ARGUMENT_ERROR</code> if the pattern * contains a syntax error. *<em> Empties the set passed before applying the pattern.</em> @@ -683,9 +689,11 @@ public: * pattern.length() if the closing ']' is the last character of * the pattern string. * @param options bitmask for options to apply to the pattern. - * Valid options are USET_IGNORE_SPACE and USET_CASE_INSENSITIVE. + * Valid options are USET_IGNORE_SPACE and + * at most one of USET_CASE_INSENSITIVE, USET_ADD_CASE_MAPPINGS, USET_SIMPLE_CASE_INSENSITIVE. + * These case options are mutually exclusive. * @param symbols a symbol table mapping variable names to - * values and stand-ins to UnicodeSets; may be NULL + * values and stand-ins to UnicodeSets; may be nullptr * @param status returns <code>U_ILLEGAL_ARGUMENT_ERROR</code> if the pattern * contains a syntax error. * @return a reference to this @@ -1390,7 +1398,7 @@ public: /** * Close this set over the given attribute. For the attribute - * USET_CASE, the result is to modify this set so that: + * USET_CASE_INSENSITIVE, the result is to modify this set so that: * * 1. For each character or string 'a' in this set, all strings or * characters 'b' such that foldCase(a) == foldCase(b) are added @@ -1408,8 +1416,10 @@ public: * A frozen set will not be modified. * * @param attribute bitmask for attributes to close over. - * Currently only the USET_CASE bit is supported. Any undefined bits - * are ignored. + * Valid options: + * At most one of USET_CASE_INSENSITIVE, USET_ADD_CASE_MAPPINGS, USET_SIMPLE_CASE_INSENSITIVE. + * These case options are mutually exclusive. + * Unrelated options bits are ignored. * @return a reference to this set. * @stable ICU 4.2 */ @@ -1488,7 +1498,7 @@ public: * bits followed by least significant 16 bits. * * @param dest pointer to buffer of destCapacity 16-bit integers. - * May be NULL only if destCapacity is zero. + * May be nullptr only if destCapacity is zero. * @param destCapacity size of dest, or zero. Must not be negative. * @param ec error code. Will be set to U_INDEX_OUTOFBOUNDS_ERROR * if n+2*m > 0x7FFF. Will be set to U_BUFFER_OVERFLOW_ERROR if @@ -1579,6 +1589,9 @@ private: int32_t depth, UErrorCode& ec); + void closeOverCaseInsensitive(bool simple); + void closeOverAddCaseMappings(); + //---------------------------------------------------------------- // Implementation: Utility methods //---------------------------------------------------------------- @@ -1718,7 +1731,7 @@ inline bool UnicodeSet::operator!=(const UnicodeSet& o) const { } inline UBool UnicodeSet::isFrozen() const { - return (UBool)(bmpSet!=NULL || stringSpan!=NULL); + return (UBool)(bmpSet!=nullptr || stringSpan!=nullptr); } inline UBool UnicodeSet::containsSome(UChar32 start, UChar32 end) const { |