| 1 | // © 2016 and later: Unicode, Inc. and others. |
| 2 | // License & terms of use: http://www.unicode.org/copyright.html |
| 3 | /* |
| 4 | ******************************************************************************* |
| 5 | * Copyright (C) 2013-2014, International Business Machines |
| 6 | * Corporation and others. All Rights Reserved. |
| 7 | ******************************************************************************* |
| 8 | * collationruleparser.h |
| 9 | * |
| 10 | * created on: 2013apr10 |
| 11 | * created by: Markus W. Scherer |
| 12 | */ |
| 13 | |
| 14 | #ifndef __COLLATIONRULEPARSER_H__ |
| 15 | #define __COLLATIONRULEPARSER_H__ |
| 16 | |
| 17 | #include "unicode/utypes.h" |
| 18 | |
| 19 | #if !UCONFIG_NO_COLLATION |
| 20 | |
| 21 | #include "unicode/ucol.h" |
| 22 | #include "unicode/uniset.h" |
| 23 | #include "unicode/unistr.h" |
| 24 | |
| 25 | struct UParseError; |
| 26 | |
| 27 | U_NAMESPACE_BEGIN |
| 28 | |
| 29 | struct CollationData; |
| 30 | struct CollationTailoring; |
| 31 | |
| 32 | class Locale; |
| 33 | class Normalizer2; |
| 34 | |
| 35 | struct CollationSettings; |
| 36 | |
| 37 | class U_I18N_API CollationRuleParser : public UMemory { |
| 38 | public: |
| 39 | /** Special reset positions. */ |
| 40 | enum Position { |
| 41 | FIRST_TERTIARY_IGNORABLE, |
| 42 | LAST_TERTIARY_IGNORABLE, |
| 43 | FIRST_SECONDARY_IGNORABLE, |
| 44 | LAST_SECONDARY_IGNORABLE, |
| 45 | FIRST_PRIMARY_IGNORABLE, |
| 46 | LAST_PRIMARY_IGNORABLE, |
| 47 | FIRST_VARIABLE, |
| 48 | LAST_VARIABLE, |
| 49 | FIRST_REGULAR, |
| 50 | LAST_REGULAR, |
| 51 | FIRST_IMPLICIT, |
| 52 | LAST_IMPLICIT, |
| 53 | FIRST_TRAILING, |
| 54 | LAST_TRAILING |
| 55 | }; |
| 56 | |
| 57 | /** |
| 58 | * First character of contractions that encode special reset positions. |
| 59 | * U+FFFE cannot be tailored via rule syntax. |
| 60 | * |
| 61 | * The second contraction character is POS_BASE + Position. |
| 62 | */ |
| 63 | static const UChar POS_LEAD = 0xfffe; |
| 64 | /** |
| 65 | * Base for the second character of contractions that encode special reset positions. |
| 66 | * Braille characters U+28xx are printable and normalization-inert. |
| 67 | * @see POS_LEAD |
| 68 | */ |
| 69 | static const UChar POS_BASE = 0x2800; |
| 70 | |
| 71 | class U_I18N_API Sink : public UObject { |
| 72 | public: |
| 73 | virtual ~Sink(); |
| 74 | /** |
| 75 | * Adds a reset. |
| 76 | * strength=UCOL_IDENTICAL for &str. |
| 77 | * strength=UCOL_PRIMARY/UCOL_SECONDARY/UCOL_TERTIARY for &[before n]str where n=1/2/3. |
| 78 | */ |
| 79 | virtual void addReset(int32_t strength, const UnicodeString &str, |
| 80 | const char *&errorReason, UErrorCode &errorCode) = 0; |
| 81 | /** |
| 82 | * Adds a relation with strength and prefix | str / extension. |
| 83 | */ |
| 84 | virtual void addRelation(int32_t strength, const UnicodeString &prefix, |
| 85 | const UnicodeString &str, const UnicodeString &extension, |
| 86 | const char *&errorReason, UErrorCode &errorCode) = 0; |
| 87 | |
| 88 | virtual void suppressContractions(const UnicodeSet &set, const char *&errorReason, |
| 89 | UErrorCode &errorCode); |
| 90 | |
| 91 | virtual void optimize(const UnicodeSet &set, const char *&errorReason, |
| 92 | UErrorCode &errorCode); |
| 93 | }; |
| 94 | |
| 95 | class U_I18N_API Importer : public UObject { |
| 96 | public: |
| 97 | virtual ~Importer(); |
| 98 | virtual void getRules( |
| 99 | const char *localeID, const char *collationType, |
| 100 | UnicodeString &rules, |
| 101 | const char *&errorReason, UErrorCode &errorCode) = 0; |
| 102 | }; |
| 103 | |
| 104 | /** |
| 105 | * Constructor. |
| 106 | * The Sink must be set before parsing. |
| 107 | * The Importer can be set, otherwise [import locale] syntax is not supported. |
| 108 | */ |
| 109 | CollationRuleParser(const CollationData *base, UErrorCode &errorCode); |
| 110 | ~CollationRuleParser(); |
| 111 | |
| 112 | /** |
| 113 | * Sets the pointer to a Sink object. |
| 114 | * The pointer is aliased: Pointer copy without cloning or taking ownership. |
| 115 | */ |
| 116 | void setSink(Sink *sinkAlias) { |
| 117 | sink = sinkAlias; |
| 118 | } |
| 119 | |
| 120 | /** |
| 121 | * Sets the pointer to an Importer object. |
| 122 | * The pointer is aliased: Pointer copy without cloning or taking ownership. |
| 123 | */ |
| 124 | void setImporter(Importer *importerAlias) { |
| 125 | importer = importerAlias; |
| 126 | } |
| 127 | |
| 128 | void parse(const UnicodeString &ruleString, |
| 129 | CollationSettings &outSettings, |
| 130 | UParseError *outParseError, |
| 131 | UErrorCode &errorCode); |
| 132 | |
| 133 | const char *getErrorReason() const { return errorReason; } |
| 134 | |
| 135 | /** |
| 136 | * Gets a script or reorder code from its string representation. |
| 137 | * @return the script/reorder code, or |
| 138 | * -1 if not recognized |
| 139 | */ |
| 140 | static int32_t getReorderCode(const char *word); |
| 141 | |
| 142 | private: |
| 143 | /** UCOL_PRIMARY=0 .. UCOL_IDENTICAL=15 */ |
| 144 | static const int32_t STRENGTH_MASK = 0xf; |
| 145 | static const int32_t STARRED_FLAG = 0x10; |
| 146 | static const int32_t OFFSET_SHIFT = 8; |
| 147 | |
| 148 | void parse(const UnicodeString &ruleString, UErrorCode &errorCode); |
| 149 | void parseRuleChain(UErrorCode &errorCode); |
| 150 | int32_t parseResetAndPosition(UErrorCode &errorCode); |
| 151 | int32_t parseRelationOperator(UErrorCode &errorCode); |
| 152 | void parseRelationStrings(int32_t strength, int32_t i, UErrorCode &errorCode); |
| 153 | void parseStarredCharacters(int32_t strength, int32_t i, UErrorCode &errorCode); |
| 154 | int32_t parseTailoringString(int32_t i, UnicodeString &raw, UErrorCode &errorCode); |
| 155 | int32_t parseString(int32_t i, UnicodeString &raw, UErrorCode &errorCode); |
| 156 | |
| 157 | /** |
| 158 | * Sets str to a contraction of U+FFFE and (U+2800 + Position). |
| 159 | * @return rule index after the special reset position |
| 160 | */ |
| 161 | int32_t parseSpecialPosition(int32_t i, UnicodeString &str, UErrorCode &errorCode); |
| 162 | void parseSetting(UErrorCode &errorCode); |
| 163 | void parseReordering(const UnicodeString &raw, UErrorCode &errorCode); |
| 164 | static UColAttributeValue getOnOffValue(const UnicodeString &s); |
| 165 | |
| 166 | int32_t parseUnicodeSet(int32_t i, UnicodeSet &set, UErrorCode &errorCode); |
| 167 | int32_t readWords(int32_t i, UnicodeString &raw) const; |
| 168 | int32_t (int32_t i) const; |
| 169 | |
| 170 | void setParseError(const char *reason, UErrorCode &errorCode); |
| 171 | void setErrorContext(); |
| 172 | |
| 173 | /** |
| 174 | * ASCII [:P:] and [:S:]: |
| 175 | * [\u0021-\u002F \u003A-\u0040 \u005B-\u0060 \u007B-\u007E] |
| 176 | */ |
| 177 | static UBool isSyntaxChar(UChar32 c); |
| 178 | int32_t skipWhiteSpace(int32_t i) const; |
| 179 | |
| 180 | const Normalizer2 &nfd, &nfc; |
| 181 | |
| 182 | const UnicodeString *rules; |
| 183 | const CollationData *const baseData; |
| 184 | CollationSettings *settings; |
| 185 | UParseError *parseError; |
| 186 | const char *errorReason; |
| 187 | |
| 188 | Sink *sink; |
| 189 | Importer *importer; |
| 190 | |
| 191 | int32_t ruleIndex; |
| 192 | }; |
| 193 | |
| 194 | U_NAMESPACE_END |
| 195 | |
| 196 | #endif // !UCONFIG_NO_COLLATION |
| 197 | #endif // __COLLATIONRULEPARSER_H__ |
| 198 | |