diff options
Diffstat (limited to 'thirdparty/icu4c/common/unicode/uset.h')
| -rw-r--r-- | thirdparty/icu4c/common/unicode/uset.h | 58 |
1 files changed, 45 insertions, 13 deletions
diff --git a/thirdparty/icu4c/common/unicode/uset.h b/thirdparty/icu4c/common/unicode/uset.h index 5dd890e148..ee4e0036d2 100644 --- a/thirdparty/icu4c/common/unicode/uset.h +++ b/thirdparty/icu4c/common/unicode/uset.h @@ -53,6 +53,12 @@ typedef struct USet USet; /** * Bitmask values to be passed to uset_openPatternOptions() or * uset_applyPattern() taking an option parameter. + * + * Use at most one of USET_CASE_INSENSITIVE, USET_ADD_CASE_MAPPINGS, USET_SIMPLE_CASE_INSENSITIVE. + * These case options are mutually exclusive. + * + * Undefined options bits are ignored, and reserved for future use. + * * @stable ICU 2.4 */ enum { @@ -60,13 +66,13 @@ enum { * Ignore white space within patterns unless quoted or escaped. * @stable ICU 2.4 */ - USET_IGNORE_SPACE = 1, + USET_IGNORE_SPACE = 1, /** * Enable case insensitive matching. E.g., "[ab]" with this flag * will match 'a', 'A', 'b', and 'B'. "[^ab]" with this flag will * match all except 'a', 'A', 'b', and 'B'. This performs a full - * closure over case mappings, e.g. U+017F for s. + * closure over case mappings, e.g. 'ſ' (U+017F long s) for 's'. * * The resulting set is a superset of the input for the code points but * not for the strings. @@ -88,17 +94,36 @@ enum { * * @stable ICU 2.4 */ - USET_CASE_INSENSITIVE = 2, + USET_CASE_INSENSITIVE = 2, /** - * Enable case insensitive matching. E.g., "[ab]" with this flag - * will match 'a', 'A', 'b', and 'B'. "[^ab]" with this flag will - * match all except 'a', 'A', 'b', and 'B'. This adds the lower-, - * title-, and uppercase mappings as well as the case folding + * Adds all case mappings for each element in the set. + * This adds the full lower-, title-, and uppercase mappings as well as the full case folding * of each existing element in the set. + * + * Unlike the “case insensitive” options, this does not perform a closure. + * For example, it does not add 'ſ' (U+017F long s) for 's', + * 'K' (U+212A Kelvin sign) for 'k', or replace set strings by their case-folded versions. + * * @stable ICU 3.2 */ - USET_ADD_CASE_MAPPINGS = 4 + USET_ADD_CASE_MAPPINGS = 4, + +#ifndef U_HIDE_DRAFT_API + /** + * Enable case insensitive matching. + * Same as USET_CASE_INSENSITIVE but using only Simple_Case_Folding (scf) mappings, + * which map each code point to one code point, + * not full Case_Folding (cf) mappings, which map some code points to multiple code points. + * + * This is designed for case-insensitive matches, for example in certain + * regular expression implementations where only Simple_Case_Folding mappings are used, + * such as in ECMAScript (JavaScript) regular expressions. + * + * @draft ICU 73 + */ + USET_SIMPLE_CASE_INSENSITIVE = 6 +#endif // U_HIDE_DRAFT_API }; /** @@ -299,7 +324,9 @@ uset_openPattern(const UChar* pattern, int32_t patternLength, * @param patternLength the length of the pattern, or -1 if null * terminated * @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 ec the error code * @stable ICU 2.4 */ @@ -414,7 +441,10 @@ uset_set(USet* set, * The character at pattern[0] must be a '['. * @param patternLength The length of the UChar string. -1 if NUL terminated. * @param options A 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 status Returns an error if the pattern cannot be parsed. * @return Upon successful parse, the value is either * the index of the character after the closing ']' @@ -804,7 +834,7 @@ uset_clear(USet* set); /** * 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 @@ -824,8 +854,10 @@ uset_clear(USet* set); * @param set the set * * @param attributes 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. * @stable ICU 4.2 */ U_CAPI void U_EXPORT2 |