summaryrefslogtreecommitdiffstats
path: root/extensions/universalchardet/src/base/CharDistribution.h
blob: bd5143b8e27101bc3c76487d7ba4b3e3d5209f87 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
/* This Source Code Form is subject to the terms of the Mozilla Public
 * License, v. 2.0. If a copy of the MPL was not distributed with this
 * file, You can obtain one at http://mozilla.org/MPL/2.0/. */

#ifndef CharDistribution_h__
#define CharDistribution_h__

#include "nscore.h"

#define ENOUGH_DATA_THRESHOLD 1024
 
class CharDistributionAnalysis
{
public:
  CharDistributionAnalysis() {Reset();}

  //feed a block of data and do distribution analysis
  void HandleData(const char* aBuf, uint32_t aLen) {}
  
  //Feed a character with known length
  void HandleOneChar(const char* aStr, uint32_t aCharLen)
  {
    int32_t order;

    //we only care about 2-bytes character in our distribution analysis
    order = (aCharLen == 2) ? GetOrder(aStr) : -1;

    if (order >= 0)
    {
      mTotalChars++;
      //order is valid
      if ((uint32_t)order < mTableSize)
      {
        if (512 > mCharToFreqOrder[order])
          mFreqChars++;
      }
    }
  }

  //return confidence base on existing data
  float GetConfidence(void);

  //Reset analyser, clear any state 
  void      Reset()
  {
    mDone = false;
    mTotalChars = 0;
    mFreqChars = 0;
    mDataThreshold = 0;
  }

  //It is not necessary to receive all data to draw conclusion. For charset detection,
  // certain amount of data is enough
  bool GotEnoughData() {return mTotalChars > ENOUGH_DATA_THRESHOLD;}

protected:
  //we do not handle character base on its original encoding string, but 
  //convert this encoding string to a number, here called order.
  //This allow multiple encoding of a language to share one frequency table 
  virtual int32_t GetOrder(const char* str) {return -1;}
  
  //If this flag is set to true, detection is done and conclusion has been made
  bool     mDone;

  //The number of characters whose frequency order is less than 512
  uint32_t mFreqChars;

  //Total character encounted.
  uint32_t mTotalChars;

  //Number of hi-byte characters needed to trigger detection
  uint32_t mDataThreshold;

  //Mapping table to get frequency order from char order (get from GetOrder())
  const int16_t  *mCharToFreqOrder;

  //Size of above table
  uint32_t mTableSize;

  //This is a constant value varies from language to language, it is used in 
  //calculating confidence. See my paper for further detail.
  float    mTypicalDistributionRatio;
};


class EUCTWDistributionAnalysis: public CharDistributionAnalysis
{
public:
  EUCTWDistributionAnalysis();
protected:

  //for euc-TW encoding, we are interested 
  //  first  byte range: 0xc4 -- 0xfe
  //  second byte range: 0xa1 -- 0xfe
  //no validation needed here. State machine has done that
  int32_t GetOrder(const char* str) 
  { if ((unsigned char)*str >= (unsigned char)0xc4)  
      return 94*((unsigned char)str[0]-(unsigned char)0xc4) + (unsigned char)str[1] - (unsigned char)0xa1;
    else
      return -1;
  }
};


class EUCKRDistributionAnalysis : public CharDistributionAnalysis
{
public:
  EUCKRDistributionAnalysis();
protected:
  //for euc-KR encoding, we are interested 
  //  first  byte range: 0xb0 -- 0xfe
  //  second byte range: 0xa1 -- 0xfe
  //no validation needed here. State machine has done that
  int32_t GetOrder(const char* str) 
  { if ((unsigned char)*str >= (unsigned char)0xb0)  
      return 94*((unsigned char)str[0]-(unsigned char)0xb0) + (unsigned char)str[1] - (unsigned char)0xa1;
    else
      return -1;
  }
};

class GB2312DistributionAnalysis : public CharDistributionAnalysis
{
public:
  GB2312DistributionAnalysis();
protected:
  //for GB2312 encoding, we are interested 
  //  first  byte range: 0xb0 -- 0xfe
  //  second byte range: 0xa1 -- 0xfe
  //no validation needed here. State machine has done that
  int32_t GetOrder(const char* str) 
  { if ((unsigned char)*str >= (unsigned char)0xb0 && (unsigned char)str[1] >= (unsigned char)0xa1)  
      return 94*((unsigned char)str[0]-(unsigned char)0xb0) + (unsigned char)str[1] - (unsigned char)0xa1;
    else
      return -1;
  }
};


class Big5DistributionAnalysis : public CharDistributionAnalysis
{
public:
  Big5DistributionAnalysis();
protected:
  //for big5 encoding, we are interested 
  //  first  byte range: 0xa4 -- 0xfe
  //  second byte range: 0x40 -- 0x7e , 0xa1 -- 0xfe
  //no validation needed here. State machine has done that
  int32_t GetOrder(const char* str) 
  { if ((unsigned char)*str >= (unsigned char)0xa4)  
      if ((unsigned char)str[1] >= (unsigned char)0xa1)
        return 157*((unsigned char)str[0]-(unsigned char)0xa4) + (unsigned char)str[1] - (unsigned char)0xa1 +63;
      else
        return 157*((unsigned char)str[0]-(unsigned char)0xa4) + (unsigned char)str[1] - (unsigned char)0x40;
    else
      return -1;
  }
};

class SJISDistributionAnalysis : public CharDistributionAnalysis
{
public:
  SJISDistributionAnalysis();
protected:
  //for sjis encoding, we are interested 
  //  first  byte range: 0x81 -- 0x9f , 0xe0 -- 0xfe
  //  second byte range: 0x40 -- 0x7e,  0x81 -- oxfe
  //no validation needed here. State machine has done that
  int32_t GetOrder(const char* str) 
  { 
    int32_t order;
    if ((unsigned char)*str >= (unsigned char)0x81 && (unsigned char)*str <= (unsigned char)0x9f)  
      order = 188 * ((unsigned char)str[0]-(unsigned char)0x81);
    else if ((unsigned char)*str >= (unsigned char)0xe0 && (unsigned char)*str <= (unsigned char)0xef)  
      order = 188 * ((unsigned char)str[0]-(unsigned char)0xe0 + 31);
    else
      return -1;
    order += (unsigned char)*(str+1) - 0x40;
    if ((unsigned char)str[1] > (unsigned char)0x7f)
      order--;
    return order;
  }
};

class EUCJPDistributionAnalysis : public CharDistributionAnalysis
{
public:
  EUCJPDistributionAnalysis();
protected:
  //for euc-JP encoding, we are interested 
  //  first  byte range: 0xa0 -- 0xfe
  //  second byte range: 0xa1 -- 0xfe
  //no validation needed here. State machine has done that
  int32_t GetOrder(const char* str) 
  { if ((unsigned char)*str >= (unsigned char)0xa0)  
      return 94*((unsigned char)str[0]-(unsigned char)0xa1) + (unsigned char)str[1] - (unsigned char)0xa1;
    else
      return -1;
  }
};

#endif //CharDistribution_h__