1 // © 2016 and later: Unicode, Inc. and others. 2 // License & terms of use: http://www.unicode.org/copyright.html 3 /* 4 ********************************************************************** 5 * Copyright (C) 1999-2007, International Business Machines 6 * Corporation and others. All Rights Reserved. 7 ********************************************************************** 8 * Date Name Description 9 * 11/17/99 aliu Creation. 10 ********************************************************************** 11 */ 12 #ifndef RBT_H 13 #define RBT_H 14 15 #include "unicode/utypes.h" 16 17 #if !UCONFIG_NO_TRANSLITERATION 18 19 #include "unicode/translit.h" 20 #include "unicode/utypes.h" 21 #include "unicode/parseerr.h" 22 #include "unicode/udata.h" 23 24 #define U_ICUDATA_TRANSLIT U_ICUDATA_NAME U_TREE_SEPARATOR_STRING "translit" 25 26 U_NAMESPACE_BEGIN 27 28 class TransliterationRuleData; 29 30 /** 31 * <code>RuleBasedTransliterator</code> is a transliterator 32 * built from a set of rules as defined for 33 * Transliterator::createFromRules(). 34 * See the C++ class Transliterator documentation for the rule syntax. 35 * 36 * @author Alan Liu 37 * @internal Use transliterator factory methods instead since this class will be removed in that release. 38 */ 39 class RuleBasedTransliterator : public Transliterator { 40 private: 41 /** 42 * The data object is immutable, so we can freely share it with 43 * other instances of RBT, as long as we do NOT own this object. 44 * TODO: data is no longer immutable. See bugs #1866, 2155 45 */ 46 TransliterationRuleData* fData; 47 48 /** 49 * If true, we own the data object and must delete it. 50 */ 51 UBool isDataOwned; 52 53 public: 54 55 /** 56 * Constructs a new transliterator from the given rules. 57 * @param rules rules, separated by ';' 58 * @param direction either FORWARD or REVERSE. 59 * @exception IllegalArgumentException if rules are malformed. 60 * @internal Use transliterator factory methods instead since this class will be removed in that release. 61 */ 62 RuleBasedTransliterator(const UnicodeString& id, 63 const UnicodeString& rules, 64 UTransDirection direction, 65 UnicodeFilter* adoptedFilter, 66 UParseError& parseError, 67 UErrorCode& status); 68 69 /** 70 * Constructs a new transliterator from the given rules. 71 * @param rules rules, separated by ';' 72 * @param direction either FORWARD or REVERSE. 73 * @exception IllegalArgumentException if rules are malformed. 74 * @internal Use transliterator factory methods instead since this class will be removed in that release. 75 */ 76 /*RuleBasedTransliterator(const UnicodeString& id, 77 const UnicodeString& rules, 78 UTransDirection direction, 79 UnicodeFilter* adoptedFilter, 80 UErrorCode& status);*/ 81 82 /** 83 * Convenience constructor with no filter. 84 * @internal Use transliterator factory methods instead since this class will be removed in that release. 85 */ 86 /*RuleBasedTransliterator(const UnicodeString& id, 87 const UnicodeString& rules, 88 UTransDirection direction, 89 UErrorCode& status);*/ 90 91 /** 92 * Convenience constructor with no filter and FORWARD direction. 93 * @internal Use transliterator factory methods instead since this class will be removed in that release. 94 */ 95 /*RuleBasedTransliterator(const UnicodeString& id, 96 const UnicodeString& rules, 97 UErrorCode& status);*/ 98 99 /** 100 * Convenience constructor with FORWARD direction. 101 * @internal Use transliterator factory methods instead since this class will be removed in that release. 102 */ 103 /*RuleBasedTransliterator(const UnicodeString& id, 104 const UnicodeString& rules, 105 UnicodeFilter* adoptedFilter, 106 UErrorCode& status);*/ 107 private: 108 109 friend class TransliteratorRegistry; // to access TransliterationRuleData convenience ctor 110 /** 111 * Convenience constructor. 112 * @param id the id for the transliterator. 113 * @param theData the rule data for the transliterator. 114 * @param adoptedFilter the filter for the transliterator 115 */ 116 RuleBasedTransliterator(const UnicodeString& id, 117 const TransliterationRuleData* theData, 118 UnicodeFilter* adoptedFilter = 0); 119 120 121 friend class Transliterator; // to access following ct 122 123 /** 124 * Internal constructor. 125 * @param id the id for the transliterator. 126 * @param theData the rule data for the transliterator. 127 * @param isDataAdopted determine who will own the 'data' object. True, the caller should not delete 'data'. 128 */ 129 RuleBasedTransliterator(const UnicodeString& id, 130 TransliterationRuleData* data, 131 UBool isDataAdopted); 132 133 public: 134 135 /** 136 * Copy constructor. 137 * @internal Use transliterator factory methods instead since this class will be removed in that release. 138 */ 139 RuleBasedTransliterator(const RuleBasedTransliterator&); 140 141 virtual ~RuleBasedTransliterator(); 142 143 /** 144 * Implement Transliterator API. 145 * @internal Use transliterator factory methods instead since this class will be removed in that release. 146 */ 147 virtual RuleBasedTransliterator* clone() const; 148 149 protected: 150 /** 151 * Implements {@link Transliterator#handleTransliterate}. 152 * @internal Use transliterator factory methods instead since this class will be removed in that release. 153 */ 154 virtual void handleTransliterate(Replaceable& text, UTransPosition& offsets, 155 UBool isIncremental) const; 156 157 public: 158 /** 159 * Return a representation of this transliterator as source rules. 160 * These rules will produce an equivalent transliterator if used 161 * to construct a new transliterator. 162 * @param result the string to receive the rules. Previous 163 * contents will be deleted. 164 * @param escapeUnprintable if true then convert unprintable 165 * character to their hex escape representations, \uxxxx or 166 * \Uxxxxxxxx. Unprintable characters are those other than 167 * U+000A, U+0020..U+007E. 168 * @internal Use transliterator factory methods instead since this class will be removed in that release. 169 */ 170 virtual UnicodeString& toRules(UnicodeString& result, 171 UBool escapeUnprintable) const; 172 173 protected: 174 /** 175 * Implement Transliterator framework 176 */ 177 virtual void handleGetSourceSet(UnicodeSet& result) const; 178 179 public: 180 /** 181 * Override Transliterator framework 182 */ 183 virtual UnicodeSet& getTargetSet(UnicodeSet& result) const; 184 185 /** 186 * Return the class ID for this class. This is useful only for 187 * comparing to a return value from getDynamicClassID(). For example: 188 * <pre> 189 * . Base* polymorphic_pointer = createPolymorphicObject(); 190 * . if (polymorphic_pointer->getDynamicClassID() == 191 * . Derived::getStaticClassID()) ... 192 * </pre> 193 * @return The class ID for all objects of this class. 194 * @internal Use transliterator factory methods instead since this class will be removed in that release. 195 */ 196 U_I18N_API static UClassID U_EXPORT2 getStaticClassID(void); 197 198 /** 199 * Returns a unique class ID <b>polymorphically</b>. This method 200 * is to implement a simple version of RTTI, since not all C++ 201 * compilers support genuine RTTI. Polymorphic operator==() and 202 * clone() methods call this method. 203 * 204 * @return The class ID for this object. All objects of a given 205 * class have the same class ID. Objects of other classes have 206 * different class IDs. 207 */ 208 virtual UClassID getDynamicClassID(void) const; 209 210 private: 211 212 void _construct(const UnicodeString& rules, 213 UTransDirection direction, 214 UParseError& parseError, 215 UErrorCode& status); 216 }; 217 218 219 U_NAMESPACE_END 220 221 #endif /* #if !UCONFIG_NO_TRANSLITERATION */ 222 223 #endif 224