1/* 2 ********************************************************************** 3 * Copyright (C) 2005-2012, International Business Machines 4 * Corporation and others. All Rights Reserved. 5 ********************************************************************** 6 */ 7 8#ifndef __CSRMBCS_H 9#define __CSRMBCS_H 10 11#include "unicode/utypes.h" 12 13#if !UCONFIG_NO_CONVERSION 14 15#include "csrecog.h" 16 17U_NAMESPACE_BEGIN 18 19// "Character" iterated character class. 20// Recognizers for specific mbcs encodings make their "characters" available 21// by providing a nextChar() function that fills in an instance of IteratedChar 22// with the next char from the input. 23// The returned characters are not converted to Unicode, but remain as the raw 24// bytes (concatenated into an int) from the codepage data. 25// 26// For Asian charsets, use the raw input rather than the input that has been 27// stripped of markup. Detection only considers multi-byte chars, effectively 28// stripping markup anyway, and double byte chars do occur in markup too. 29// 30class IteratedChar : public UMemory 31{ 32public: 33 uint32_t charValue; // 1-4 bytes from the raw input data 34 int32_t index; 35 int32_t nextIndex; 36 UBool error; 37 UBool done; 38 39public: 40 IteratedChar(); 41 //void reset(); 42 int32_t nextByte(InputText* det); 43}; 44 45#if U_PLATFORM_IS_DARWIN_BASED 46#define MAX_KEY_STRING_WITH_NULL 16 47#endif 48 49class CharsetRecog_mbcs : public CharsetRecognizer { 50 51protected: 52 /** 53 * Test the match of this charset with the input text data 54 * which is obtained via the CharsetDetector object. 55 * 56 * @param det The CharsetDetector, which contains the input text 57 * to be checked for being in this charset. 58 * @return Two values packed into one int (Damn java, anyhow) 59 * <br/> 60 * bits 0-7: the match confidence, ranging from 0-100 61 * <br/> 62 * bits 8-15: The match reason, an enum-like value. 63 */ 64#if U_PLATFORM_IS_DARWIN_BASED 65 int32_t match_mbcs(InputText* det, const uint16_t commonChars[], int32_t commonCharsLen, const uint8_t (*keyStrings)[MAX_KEY_STRING_WITH_NULL] ) const; 66#else 67 int32_t match_mbcs(InputText* det, const uint16_t commonChars[], int32_t commonCharsLen) const; 68#endif 69 70public: 71 72 virtual ~CharsetRecog_mbcs(); 73 74 /** 75 * Get the IANA name of this charset. 76 * @return the charset name. 77 */ 78 79 const char *getName() const = 0; 80 const char *getLanguage() const = 0; 81 UBool match(InputText* input, CharsetMatch *results) const = 0; 82 83 /** 84 * Get the next character (however many bytes it is) from the input data 85 * Subclasses for specific charset encodings must implement this function 86 * to get characters according to the rules of their encoding scheme. 87 * 88 * This function is not a method of class IteratedChar only because 89 * that would require a lot of extra derived classes, which is awkward. 90 * @param it The IteratedChar "struct" into which the returned char is placed. 91 * @param det The charset detector, which is needed to get at the input byte data 92 * being iterated over. 93 * @return True if a character was returned, false at end of input. 94 */ 95 virtual UBool nextChar(IteratedChar *it, InputText *textIn) const = 0; 96 97}; 98 99 100/** 101 * Shift-JIS charset recognizer. 102 * 103 */ 104class CharsetRecog_sjis : public CharsetRecog_mbcs { 105public: 106 virtual ~CharsetRecog_sjis(); 107 108 UBool nextChar(IteratedChar *it, InputText *det) const; 109 110 UBool match(InputText* input, CharsetMatch *results) const; 111 112 const char *getName() const; 113 const char *getLanguage() const; 114 115}; 116 117 118/** 119 * EUC charset recognizers. One abstract class that provides the common function 120 * for getting the next character according to the EUC encoding scheme, 121 * and nested derived classes for EUC_KR, EUC_JP, EUC_CN. 122 * 123 */ 124class CharsetRecog_euc : public CharsetRecog_mbcs 125{ 126public: 127 virtual ~CharsetRecog_euc(); 128 129 const char *getName() const = 0; 130 const char *getLanguage() const = 0; 131 132 UBool match(InputText* input, CharsetMatch *results) const = 0; 133 /* 134 * (non-Javadoc) 135 * Get the next character value for EUC based encodings. 136 * Character "value" is simply the raw bytes that make up the character 137 * packed into an int. 138 */ 139 UBool nextChar(IteratedChar *it, InputText *det) const; 140}; 141 142/** 143 * The charset recognize for EUC-JP. A singleton instance of this class 144 * is created and kept by the public CharsetDetector class 145 */ 146class CharsetRecog_euc_jp : public CharsetRecog_euc 147{ 148public: 149 virtual ~CharsetRecog_euc_jp(); 150 151 const char *getName() const; 152 const char *getLanguage() const; 153 154 UBool match(InputText* input, CharsetMatch *results) const; 155}; 156 157/** 158 * The charset recognize for EUC-KR. A singleton instance of this class 159 * is created and kept by the public CharsetDetector class 160 */ 161class CharsetRecog_euc_kr : public CharsetRecog_euc 162{ 163public: 164 virtual ~CharsetRecog_euc_kr(); 165 166 const char *getName() const; 167 const char *getLanguage() const; 168 169 UBool match(InputText* input, CharsetMatch *results) const; 170}; 171 172/** 173 * 174 * Big5 charset recognizer. 175 * 176 */ 177class CharsetRecog_big5 : public CharsetRecog_mbcs 178{ 179public: 180 virtual ~CharsetRecog_big5(); 181 182 UBool nextChar(IteratedChar* it, InputText* det) const; 183 184 const char *getName() const; 185 const char *getLanguage() const; 186 187 UBool match(InputText* input, CharsetMatch *results) const; 188}; 189 190 191/** 192 * 193 * GB-18030 recognizer. Uses simplified Chinese statistics. 194 * 195 */ 196class CharsetRecog_gb_18030 : public CharsetRecog_mbcs 197{ 198public: 199 virtual ~CharsetRecog_gb_18030(); 200 201 UBool nextChar(IteratedChar* it, InputText* det) const; 202 203 const char *getName() const; 204 const char *getLanguage() const; 205 206 UBool match(InputText* input, CharsetMatch *results) const; 207}; 208 209U_NAMESPACE_END 210 211#endif 212#endif /* __CSRMBCS_H */ 213