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
46class CharsetRecog_mbcs : public CharsetRecognizer {
47
48protected:
49    /**
50     * Test the match of this charset with the input text data
51     *      which is obtained via the CharsetDetector object.
52     *
53     * @param det  The CharsetDetector, which contains the input text
54     *             to be checked for being in this charset.
55     * @return     Two values packed into one int  (Damn java, anyhow)
56     *             <br/>
57     *             bits 0-7:  the match confidence, ranging from 0-100
58     *             <br/>
59     *             bits 8-15: The match reason, an enum-like value.
60     */
61    int32_t match_mbcs(InputText* det, const uint16_t commonChars[], int32_t commonCharsLen) const;
62
63public:
64
65    virtual ~CharsetRecog_mbcs();
66
67    /**
68     * Get the IANA name of this charset.
69     * @return the charset name.
70     */
71
72    const char *getName() const = 0;
73    const char *getLanguage() const = 0;
74    UBool match(InputText* input, CharsetMatch *results) const = 0;
75
76    /**
77     * Get the next character (however many bytes it is) from the input data
78     *    Subclasses for specific charset encodings must implement this function
79     *    to get characters according to the rules of their encoding scheme.
80     *
81     *  This function is not a method of class IteratedChar only because
82     *   that would require a lot of extra derived classes, which is awkward.
83     * @param it  The IteratedChar "struct" into which the returned char is placed.
84     * @param det The charset detector, which is needed to get at the input byte data
85     *            being iterated over.
86     * @return    True if a character was returned, false at end of input.
87     */
88    virtual UBool nextChar(IteratedChar *it, InputText *textIn) const = 0;
89
90};
91
92
93/**
94 *   Shift-JIS charset recognizer.
95 *
96 */
97class CharsetRecog_sjis : public CharsetRecog_mbcs {
98public:
99    virtual ~CharsetRecog_sjis();
100
101    UBool nextChar(IteratedChar *it, InputText *det) const;
102
103    UBool match(InputText* input, CharsetMatch *results) const;
104
105    const char *getName() const;
106    const char *getLanguage() const;
107
108};
109
110
111/**
112 *   EUC charset recognizers.  One abstract class that provides the common function
113 *             for getting the next character according to the EUC encoding scheme,
114 *             and nested derived classes for EUC_KR, EUC_JP, EUC_CN.
115 *
116 */
117class CharsetRecog_euc : public CharsetRecog_mbcs
118{
119public:
120    virtual ~CharsetRecog_euc();
121
122    const char *getName() const = 0;
123    const char *getLanguage() const = 0;
124
125    UBool match(InputText* input, CharsetMatch *results) const = 0;
126    /*
127     *  (non-Javadoc)
128     *  Get the next character value for EUC based encodings.
129     *  Character "value" is simply the raw bytes that make up the character
130     *     packed into an int.
131     */
132    UBool nextChar(IteratedChar *it, InputText *det) const;
133};
134
135/**
136 * The charset recognize for EUC-JP.  A singleton instance of this class
137 *    is created and kept by the public CharsetDetector class
138 */
139class CharsetRecog_euc_jp : public CharsetRecog_euc
140{
141public:
142    virtual ~CharsetRecog_euc_jp();
143
144    const char *getName() const;
145    const char *getLanguage() const;
146
147    UBool match(InputText* input, CharsetMatch *results) const;
148};
149
150/**
151 * The charset recognize for EUC-KR.  A singleton instance of this class
152 *    is created and kept by the public CharsetDetector class
153 */
154class CharsetRecog_euc_kr : public CharsetRecog_euc
155{
156public:
157    virtual ~CharsetRecog_euc_kr();
158
159    const char *getName() const;
160    const char *getLanguage() const;
161
162    UBool match(InputText* input, CharsetMatch *results) const;
163};
164
165/**
166 *
167 *   Big5 charset recognizer.
168 *
169 */
170class CharsetRecog_big5 : public CharsetRecog_mbcs
171{
172public:
173    virtual ~CharsetRecog_big5();
174
175    UBool nextChar(IteratedChar* it, InputText* det) const;
176
177    const char *getName() const;
178    const char *getLanguage() const;
179
180    UBool match(InputText* input, CharsetMatch *results) const;
181};
182
183
184/**
185 *
186 *   GB-18030 recognizer. Uses simplified Chinese statistics.
187 *
188 */
189class CharsetRecog_gb_18030 : public CharsetRecog_mbcs
190{
191public:
192    virtual ~CharsetRecog_gb_18030();
193
194    UBool nextChar(IteratedChar* it, InputText* det) const;
195
196    const char *getName() const;
197    const char *getLanguage() const;
198
199    UBool match(InputText* input, CharsetMatch *results) const;
200};
201
202U_NAMESPACE_END
203
204#endif
205#endif /* __CSRMBCS_H */
206