• Skip to content
  • Skip to link menu
KDE 4.3 API Reference
  • KDE API Reference
  • kdelibs
  • Sitemap
  • Contact Us
 

KDECore

nsMBCSGroupProber.cpp

Go to the documentation of this file.
00001 /* -*- Mode: C; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
00002 /*  -*- C++ -*-
00003 *  Copyright (C) 1998 <developer@mozilla.org>
00004 *
00005 *
00006 *  Permission is hereby granted, free of charge, to any person obtaining
00007 *  a copy of this software and associated documentation files (the
00008 *  "Software"), to deal in the Software without restriction, including
00009 *  without limitation the rights to use, copy, modify, merge, publish,
00010 *  distribute, sublicense, and/or sell copies of the Software, and to
00011 *  permit persons to whom the Software is furnished to do so, subject to
00012 *  the following conditions:
00013 *
00014 *  The above copyright notice and this permission notice shall be included 
00015 *  in all copies or substantial portions of the Software.
00016 *
00017 *  THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
00018 *  EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
00019 *  MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
00020 *  NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
00021 *  LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
00022 *  OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
00023 *  WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
00024 */
00025 
00026 #include "nsMBCSGroupProber.h"
00027 
00028 #include <stdio.h>
00029 #include <stdlib.h>
00030 
00031 namespace kencodingprober {
00032 #ifdef DEBUG_PROBE
00033 static const char* const ProberName[] =
00034 {
00035   "Unicode",
00036   "SJIS",
00037   "EUCJP",
00038   "GB18030",
00039   "EUCKR",
00040   "Big5",
00041   "EUCTW",
00042 };
00043 
00044 #endif
00045 
00046 nsMBCSGroupProber::nsMBCSGroupProber()
00047 {
00048   mProbers[0] = new UnicodeGroupProber();
00049   mProbers[1] = new nsSJISProber();
00050   mProbers[2] = new nsEUCJPProber();
00051   mProbers[3] = new nsGB18030Prober();
00052   mProbers[4] = new nsEUCKRProber();
00053   mProbers[5] = new nsBig5Prober();
00054   mProbers[6] = new nsEUCTWProber();
00055   Reset();
00056 }
00057 
00058 nsMBCSGroupProber::~nsMBCSGroupProber()
00059 {
00060   for (unsigned int i = 0; i < NUM_OF_PROBERS; i++)
00061   {
00062     delete mProbers[i];
00063   }
00064 }
00065 
00066 const char* nsMBCSGroupProber::GetCharSetName()
00067 {
00068   if (mBestGuess == -1)
00069   {
00070     GetConfidence();
00071     if (mBestGuess == -1)
00072       mBestGuess = 0;
00073   }
00074   return mProbers[mBestGuess]->GetCharSetName();
00075 }
00076 
00077 void  nsMBCSGroupProber::Reset(void)
00078 {
00079   mActiveNum = 0;
00080   for (unsigned int i = 0; i < NUM_OF_PROBERS; i++)
00081   {
00082     if (mProbers[i])
00083     {
00084       mProbers[i]->Reset();
00085       mIsActive[i] = true;
00086       ++mActiveNum;
00087     }
00088     else
00089       mIsActive[i] = false;
00090   }
00091   mBestGuess = -1;
00092   mState = eDetecting;
00093 }
00094 
00095 nsProbingState nsMBCSGroupProber::HandleData(const char* aBuf, unsigned int aLen)
00096 {
00097   nsProbingState st;
00098   unsigned int i;
00099 
00100   //do filtering to reduce load to probers
00101   char *highbyteBuf;
00102   char *hptr;
00103   bool keepNext = true;   //assume previous is not ascii, it will do no harm except add some noise
00104   hptr = highbyteBuf = (char*)malloc(aLen);
00105   if (!hptr)
00106       return mState;
00107   for (i = 0; i < aLen; ++i)
00108   {
00109     if (aBuf[i] & 0x80)
00110     {
00111       *hptr++ = aBuf[i];
00112       keepNext = true;
00113     }
00114     else
00115     {
00116       //if previous is highbyte, keep this even it is a ASCII
00117       if (keepNext)
00118       {
00119           *hptr++ = aBuf[i];
00120           keepNext = false;
00121       }
00122     }
00123   }
00124 
00125   for (i = 0; i < NUM_OF_PROBERS; ++i)
00126   {
00127      if (!mIsActive[i])
00128        continue;
00129      st = mProbers[i]->HandleData(highbyteBuf, hptr - highbyteBuf);
00130      if (st == eFoundIt)
00131      {
00132        mBestGuess = i;
00133        mState = eFoundIt;
00134        break;
00135      }
00136      else if (st == eNotMe)
00137      {
00138        mIsActive[i] = false;
00139        mActiveNum--;
00140        if (mActiveNum <= 0)
00141        {
00142          mState = eNotMe;
00143          break;
00144        }
00145      }
00146   }
00147 
00148   free(highbyteBuf);
00149 
00150   return mState;
00151 }
00152 
00153 float nsMBCSGroupProber::GetConfidence(void)
00154 {
00155   unsigned int i;
00156   float bestConf = 0.0, cf;
00157 
00158   switch (mState)
00159   {
00160   case eFoundIt:
00161     return (float)0.99;
00162   case eNotMe:
00163     return (float)0.01;
00164   default:
00165     for (i = 0; i < NUM_OF_PROBERS; ++i)
00166     {
00167       if (!mIsActive[i])
00168         continue;
00169       cf = mProbers[i]->GetConfidence();
00170       if (bestConf < cf)
00171       {
00172         bestConf = cf;
00173         mBestGuess = i;
00174       }
00175     }
00176   }
00177   return bestConf;
00178 }
00179 
00180 #ifdef DEBUG_PROBE
00181 void nsMBCSGroupProber::DumpStatus()
00182 {
00183   unsigned int i;
00184   float cf;
00185   
00186   GetConfidence();
00187   for (i = 0; i < NUM_OF_PROBERS; i++)
00188   {
00189     if (!mIsActive[i])
00190       printf("  MBCS inactive: [%s] (confidence is too low).\r\n", ProberName[i]);
00191     else
00192     {
00193       cf = mProbers[i]->GetConfidence();
00194       printf("  MBCS %1.3f: [%s]\r\n", cf, ProberName[i]);
00195     }
00196   }
00197 }
00198 #endif
00199 }
00200 
00201 

KDECore

Skip menu "KDECore"
  • Main Page
  • Modules
  • Namespace List
  • Class Hierarchy
  • Alphabetical List
  • Class List
  • File List
  • Namespace Members
  • Class Members
  • Related Pages

kdelibs

Skip menu "kdelibs"
  • DNSSD
  • Interfaces
  •   KHexEdit
  •   KMediaPlayer
  •   KSpeech
  •   KTextEditor
  • Kate
  • kconf_update
  • KDE3Support
  •   KUnitTest
  • KDECore
  • KDED
  • KDEsu
  • KDEUI
  • KDocTools
  • KFile
  • KHTML
  • KImgIO
  • KInit
  • kio
  • KIOSlave
  • KJS
  •   KJS-API
  •   WTF
  • kjsembed
  • KNewStuff
  • KParts
  • KPty
  • Kross
  • KUtils
  • Nepomuk
  • Plasma
  • Solid
  • Sonnet
  • ThreadWeaver
Generated for kdelibs by doxygen 1.6.1
This website is maintained by Adriaan de Groot and Allen Winter.
KDE® and the K Desktop Environment® logo are registered trademarks of KDE e.V. | Legal