X-Git-Url: https://git.saurik.com/apple/icu.git/blobdiff_plain/57a6839dcb3bba09e8228b822b290604668416fe..38fbf2fd31f5cd99b500914d6037b1d06b608645:/icuSources/i18n/unicode/regex.h diff --git a/icuSources/i18n/unicode/regex.h b/icuSources/i18n/unicode/regex.h index b2ef4ce3..8e3b1cb3 100644 --- a/icuSources/i18n/unicode/regex.h +++ b/icuSources/i18n/unicode/regex.h @@ -1,10 +1,12 @@ +// © 2016 and later: Unicode, Inc. and others. +// License & terms of use: http://www.unicode.org/copyright.html /* ********************************************************************** -* Copyright (C) 2002-2013, International Business Machines +* Copyright (C) 2002-2016, International Business Machines * Corporation and others. All Rights Reserved. ********************************************************************** * file name: regex.h -* encoding: US-ASCII +* encoding: UTF-8 * indentation:4 * * created on: 2002oct22 @@ -55,6 +57,9 @@ // Forward Declarations +struct UHashtable; + +#if U_SHOW_CPLUSPLUS_API U_NAMESPACE_BEGIN struct Regex8BitSet; @@ -80,7 +85,7 @@ class UVector64; * * @stable ICU 2.4 */ -class U_I18N_API RegexPattern: public UObject { +class U_I18N_API RegexPattern U_FINAL : public UObject { public: /** @@ -136,7 +141,7 @@ public: /** * Create an exact copy of this RegexPattern object. Since RegexPattern is not - * intended to be subclasses, clone() and the copy construction are + * intended to be subclassed, clone() and the copy construction are * equivalent operations. * @return the copy of this RegexPattern * @stable ICU 2.4 @@ -346,17 +351,17 @@ public: private: /** * Cause a compilation error if an application accidentally attempts to - * create a matcher with a (UChar *) string as input rather than + * create a matcher with a (char16_t *) string as input rather than * a UnicodeString. Avoids a dangling reference to a temporary string. *

- * To efficiently work with UChar *strings, wrap the data in a UnicodeString + * To efficiently work with char16_t *strings, wrap the data in a UnicodeString * using one of the aliasing constructors, such as - * UnicodeString(UBool isTerminated, const UChar *text, int32_t textLength); + * UnicodeString(UBool isTerminated, const char16_t *text, int32_t textLength); * or in a UText, using - * utext_openUChars(UText *ut, const UChar *text, int64_t textLength, UErrorCode *status); + * utext_openUChars(UText *ut, const char16_t *text, int64_t textLength, UErrorCode *status); * */ - RegexMatcher *matcher(const UChar *input, + RegexMatcher *matcher(const char16_t *input, UErrorCode &status) const; public: @@ -437,6 +442,41 @@ public: virtual UText *patternText(UErrorCode &status) const; + /** + * Get the group number corresponding to a named capture group. + * The returned number can be used with any function that access + * capture groups by number. + * + * The function returns an error status if the specified name does not + * appear in the pattern. + * + * @param groupName The capture group name. + * @param status A UErrorCode to receive any errors. + * + * @stable ICU 55 + */ + virtual int32_t groupNumberFromName(const UnicodeString &groupName, UErrorCode &status) const; + + + /** + * Get the group number corresponding to a named capture group. + * The returned number can be used with any function that access + * capture groups by number. + * + * The function returns an error status if the specified name does not + * appear in the pattern. + * + * @param groupName The capture group name, + * platform invariant characters only. + * @param nameLength The length of the name, or -1 if the name is + * nul-terminated. + * @param status A UErrorCode to receive any errors. + * + * @stable ICU 55 + */ + virtual int32_t groupNumberFromName(const char *groupName, int32_t nameLength, UErrorCode &status) const; + + /** * Split a string into fields. Somewhat like split() from Perl or Java. * Pattern matches identify delimiters that separate the input @@ -573,8 +613,6 @@ private: UVector32 *fGroupMap; // Map from capture group number to position of // the group's variables in the matcher stack frame. - int32_t fMaxCaptureDigits; - UnicodeSet **fStaticSets; // Ptr to static (shared) sets for predefined // regex character classes, e.g. Word. @@ -589,6 +627,8 @@ private: Regex8BitSet *fInitialChars8; UBool fNeedsAltInput; + UHashtable *fNamedCaptureMap; // Map from capture group names to numbers. + friend class RegexCompile; friend class RegexMatcher; friend class RegexCImpl; @@ -608,7 +648,7 @@ private: * @internal */ void dumpPattern() const; -#endif +#endif /* U_HIDE_INTERNAL_API */ }; @@ -622,7 +662,7 @@ private: * * @stable ICU 2.4 */ -class U_I18N_API RegexMatcher: public UObject { +class U_I18N_API RegexMatcher U_FINAL : public UObject { public: /** @@ -709,17 +749,17 @@ public: private: /** * Cause a compilation error if an application accidentally attempts to - * create a matcher with a (UChar *) string as input rather than + * create a matcher with a (char16_t *) string as input rather than * a UnicodeString. Avoids a dangling reference to a temporary string. *

- * To efficiently work with UChar *strings, wrap the data in a UnicodeString + * To efficiently work with char16_t *strings, wrap the data in a UnicodeString * using one of the aliasing constructors, such as - * UnicodeString(UBool isTerminated, const UChar *text, int32_t textLength); + * UnicodeString(UBool isTerminated, const char16_t *text, int32_t textLength); * or in a UText, using - * utext_openUChars(UText *ut, const UChar *text, int64_t textLength, UErrorCode *status); + * utext_openUChars(UText *ut, const char16_t *text, int64_t textLength, UErrorCode *status); * */ - RegexMatcher(const UnicodeString ®exp, const UChar *input, + RegexMatcher(const UnicodeString ®exp, const char16_t *input, uint32_t flags, UErrorCode &status); public: @@ -801,6 +841,21 @@ public: virtual UBool find(); + /** + * Find the next pattern match in the input string. + * The find begins searching the input at the location following the end of + * the previous match, or at the start of the string if there is no previous match. + * If a match is found, start(), end() and group() + * will provide more information regarding the match. + *

Note that if the input string is changed by the application, + * use find(startPos, status) instead of find(), because the saved starting + * position may not be valid with the altered input string.

+ * @param status A reference to a UErrorCode to receive any errors. + * @return TRUE if a match is found. + * @stable ICU 55 + */ + virtual UBool find(UErrorCode &status); + /** * Resets this RegexMatcher and then attempts to find the next substring of the * input string that matches the pattern, starting at the specified index. @@ -829,6 +884,11 @@ public: * Returns a string containing the text captured by the given group * during the previous match operation. Group(0) is the entire match. * + * A zero length string is returned both for capture groups that did not + * participate in the match and for actual zero length matches. + * To distinguish between these two cases use the function start(), + * which returns -1 for non-participating groups. + * * @param groupNum the capture group number * @param status A reference to a UErrorCode to receive any errors. * Possible errors are U_REGEX_INVALID_STATE if no match @@ -839,7 +899,6 @@ public: */ virtual UnicodeString group(int32_t groupNum, UErrorCode &status) const; - /** * Returns the number of capturing groups in this matcher's pattern. * @return the number of capture groups @@ -868,6 +927,11 @@ public: * Returns a shallow clone of the entire live input string with the UText current native index * set to the beginning of the requested group. * + * A group length of zero is returned both for capture groups that did not + * participate in the match and for actual zero length matches. + * To distinguish between these two cases use the function start(), + * which returns -1 for non-participating groups. + * * @param groupNum The capture group number. * @param dest The UText into which the input should be cloned, or NULL to create a new UText. * @param group_len A reference to receive the length of the desired capture group @@ -881,24 +945,6 @@ public: */ virtual UText *group(int32_t groupNum, UText *dest, int64_t &group_len, UErrorCode &status) const; - /** - * Returns a string containing the text captured by the given group - * during the previous match operation. Group(0) is the entire match. - * - * @param groupNum the capture group number - * @param dest A mutable UText in which the matching text is placed. - * If NULL, a new UText will be created (which may not be mutable). - * @param status A reference to a UErrorCode to receive any errors. - * Possible errors are U_REGEX_INVALID_STATE if no match - * has been attempted or the last match failed. - * @return A string containing the matched input text. If a pre-allocated UText - * was provided, it will always be used and returned. - * - * @internal ICU 4.4 technology preview - */ - virtual UText *group(int32_t groupNum, UText *dest, UErrorCode &status) const; - - /** * Returns the index in the input string of the start of the text matched * during the previous match operation. @@ -948,7 +994,6 @@ public: */ virtual int64_t start64(int32_t group, UErrorCode &status) const; - /** * Returns the index in the input string of the first character following the * text matched during the previous match operation. @@ -1018,7 +1063,6 @@ public: */ virtual int64_t end64(int32_t group, UErrorCode &status) const; - /** * Resets this matcher. The effect is to remove any memory of previous matches, * and to cause subsequent find() operations to begin at the beginning of @@ -1113,17 +1157,17 @@ public: private: /** * Cause a compilation error if an application accidentally attempts to - * reset a matcher with a (UChar *) string as input rather than + * reset a matcher with a (char16_t *) string as input rather than * a UnicodeString. Avoids a dangling reference to a temporary string. *

- * To efficiently work with UChar *strings, wrap the data in a UnicodeString + * To efficiently work with char16_t *strings, wrap the data in a UnicodeString * using one of the aliasing constructors, such as - * UnicodeString(UBool isTerminated, const UChar *text, int32_t textLength); + * UnicodeString(UBool isTerminated, const char16_t *text, int32_t textLength); * or in a UText, using - * utext_openUChars(UText *ut, const UChar *text, int64_t textLength, UErrorCode *status); + * utext_openUChars(UText *ut, const char16_t *text, int64_t textLength, UErrorCode *status); * */ - RegexMatcher &reset(const UChar *input); + RegexMatcher &reset(const char16_t *input); public: /** @@ -1744,11 +1788,13 @@ private: REStackFrame *resetStack(); inline REStackFrame *StateSave(REStackFrame *fp, int64_t savePatIdx, UErrorCode &status); void IncrementTime(UErrorCode &status); - UBool ReportFindProgress(int64_t matchIndex, UErrorCode &status); + + // Call user find callback function, if set. Return TRUE if operation should be interrupted. + inline UBool findProgressInterrupt(int64_t matchIndex, UErrorCode &status); int64_t appendGroup(int32_t groupNum, UText *dest, UErrorCode &status) const; - UBool findUsingChunk(); + UBool findUsingChunk(UErrorCode &status); void MatchChunkAt(int32_t startIdx, UBool toEnd, UErrorCode &status); UBool isChunkWordBoundary(int32_t pos); @@ -1836,5 +1882,7 @@ private: }; U_NAMESPACE_END +#endif // U_SHOW_CPLUSPLUS_API + #endif // UCONFIG_NO_REGULAR_EXPRESSIONS #endif