File indexing completed on 2024-12-22 04:33:41

0001 /* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
0002 /* ***** BEGIN LICENSE BLOCK *****
0003  * Version: MPL 1.1/GPL 2.0/LGPL 2.1
0004  *
0005  * The contents of this file are subject to the Mozilla Public License Version
0006  * 1.1 (the "License"); you may not use this file except in compliance with
0007  * the License. You may obtain a copy of the License at
0008  * http://www.mozilla.org/MPL/
0009  *
0010  * Software distributed under the License is distributed on an "AS IS" basis,
0011  * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License
0012  * for the specific language governing rights and limitations under the
0013  * License.
0014  *
0015  * The Original Code is Mozilla Communicator client code.
0016  *
0017  * The Initial Developer of the Original Code is
0018  * Netscape Communications Corporation.
0019  * Portions created by the Initial Developer are Copyright (C) 1998
0020  * the Initial Developer. All Rights Reserved.
0021  *
0022  * Contributor(s):
0023  *
0024  * Alternatively, the contents of this file may be used under the terms of
0025  * either the GNU General Public License Version 2 or later (the "GPL"), or
0026  * the GNU Lesser General Public License Version 2.1 or later (the "LGPL"),
0027  * in which case the provisions of the GPL or the LGPL are applicable instead
0028  * of those above. If you wish to allow use of your version of this file only
0029  * under the terms of either the GPL or the LGPL, and not to allow others to
0030  * use your version of this file under the terms of the MPL, indicate your
0031  * decision by deleting the provisions above and replace them with the notice
0032  * and other provisions required by the GPL or the LGPL. If you do not delete
0033  * the provisions above, a recipient may use your version of this file under
0034  * the terms of any one of the MPL, the GPL or the LGPL.
0035  *
0036  * ***** END LICENSE BLOCK ***** */
0037 
0038 #ifndef __JPCNTX_H__
0039 #define __JPCNTX_H__
0040 
0041 #define NUM_OF_CATEGORY 6
0042 
0043 #include "nscore.h" 
0044 
0045 #define ENOUGH_REL_THRESHOLD  100
0046 #define MAX_REL_THRESHOLD     1000
0047 
0048 //hiragana frequency category table
0049 extern char jp2CharContext[83][83];
0050 
0051 class JapaneseContextAnalysis
0052 {
0053 public:
0054   JapaneseContextAnalysis() {Reset();}
0055   virtual ~JapaneseContextAnalysis() {}
0056 
0057   void HandleData(const char* aBuf, PRUint32 aLen);
0058 
0059   void HandleOneChar(const char* aStr, PRUint32 aCharLen)
0060   {
0061     PRInt32 order;
0062 
0063     //if we received enough data, stop here   
0064     if (mTotalRel > MAX_REL_THRESHOLD)   mDone = PR_TRUE;
0065     if (mDone)       return;
0066      
0067     //Only 2-bytes characters are of our interest
0068     order = (aCharLen == 2) ? GetOrder(aStr) : -1;
0069     if (order != -1 && mLastCharOrder != -1)
0070     {
0071       mTotalRel++;
0072       //count this sequence to its category counter
0073       mRelSample[int(jp2CharContext[mLastCharOrder][order])]++;
0074     }
0075     mLastCharOrder = order;
0076   };
0077 
0078   float GetConfidence();
0079   void      Reset(void);
0080   void      SetOpion(){}
0081   PRBool GotEnoughData() {return mTotalRel > ENOUGH_REL_THRESHOLD;};
0082 
0083 protected:
0084   virtual PRInt32 GetOrder(const char* str, PRUint32 *charLen) = 0;
0085   virtual PRInt32 GetOrder(const char* str) = 0;
0086 
0087   //category counters, each integer counts sequence in its category
0088   PRUint32 mRelSample[NUM_OF_CATEGORY];
0089 
0090   //total sequence received
0091   PRUint32 mTotalRel;
0092   
0093   //The order of previous char
0094   PRInt32  mLastCharOrder;
0095 
0096   //if last byte in current buffer is not the last byte of a character, we
0097   //need to know how many byte to skip in next buffer.
0098   PRUint32 mNeedToSkipCharNum;
0099 
0100   //If this flag is set to PR_TRUE, detection is done and conclusion has been made
0101   PRBool   mDone;
0102 };
0103 
0104 
0105 class SJISContextAnalysis : public JapaneseContextAnalysis
0106 {
0107   //SJISContextAnalysis(){}
0108 protected:
0109   PRInt32 GetOrder(const char* str, PRUint32 *charLen);
0110 
0111   PRInt32 GetOrder(const char* str)
0112   {
0113     //We only interested in Hiragana, so first byte is '\202'
0114     if (*str == '\202' && 
0115           (unsigned char)*(str+1) >= (unsigned char)0x9f && 
0116           (unsigned char)*(str+1) <= (unsigned char)0xf1)
0117       return (unsigned char)*(str+1) - (unsigned char)0x9f;
0118     return -1;
0119   };
0120 };
0121 
0122 class EUCJPContextAnalysis : public JapaneseContextAnalysis
0123 {
0124 protected:
0125   PRInt32 GetOrder(const char* str, PRUint32 *charLen);
0126   PRInt32 GetOrder(const char* str)
0127     //We only interested in Hiragana, so first byte is '\244'
0128   {
0129     if (*str == '\244' &&
0130           (unsigned char)*(str+1) >= (unsigned char)0xa1 &&
0131           (unsigned char)*(str+1) <= (unsigned char)0xf3)
0132       return (unsigned char)*(str+1) - (unsigned char)0xa1;
0133     return -1;
0134   };
0135 };
0136 
0137 #endif /* __JPCNTX_H__ */
0138