tanl: tanl: text/Utf8Utils.h Source File

00001 /*
00002 **      Tanl
00003 **      text/Utf8Utils.h: utilities for handling UTF-8
00004 ** ----------------------------------------------------------------------
00005 **      Copyright (c) 2003  Giuseppe Attardi (attardi@di.unipi.it).
00006 ** ----------------------------------------------------------------------
00007 **
00008 **  This file is part of Tanl.
00009 **
00010 **  Tanl is free software; you can redistribute it and/or modify it
00011 **  under the terms of the GNU General Public License, version 3,
00012 **  as published by the Free Software Foundation.
00013 **
00014 **  Tanl is distributed in the hope that it will be useful,
00015 **  but WITHOUT ANY WARRANTY; without even the implied warranty of
00016 **  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
00017 **  GNU General Public License for more details.
00018 **
00019 **  You should have received a copy of the GNU General Public License
00020 **  along with this program.  If not, see <http://www.gnu.org/licenses/>.
00021 **  ----------------------------------------------------------------------
00022 */
00023 
00024 #ifndef Tanl_Text_Utf8Utils_h
00025 #define Tanl_Text_Utf8Utils_h
00026 
00027 // Standard
00028 #include <sys/types.h>
00029 #include <string.h>
00030 
00031 #if defined(_WIN32)
00032 # include <tchar.h>
00033 #endif
00034 
00035 // local
00036 #include "text.h"
00037 #include "Char.h"
00038 
00039 namespace Tanl {
00040 namespace Text {
00041 
00045 namespace Unicode
00046 {
00047   typedef unsigned char byte;
00048 
00049   void incUtfPtr(const char*& ptr);
00050   void decUtfPtr(const char*& ptr, const char* begin);
00051   size_t utfDiff(const char* end, const char* begin);
00052   bool isAscii(const char* begin, const char* end);
00053 
00054   UCS2 fetchChar(const char*& begin, const char* end);
00055   inline UCS2 toChar(const char* begin, const char* end)
00056   {
00057     const char* it = begin;
00058     return fetchChar(it, end);
00059   }
00060   int byteLength(const char* it);
00064   int utfLength(UCS2 ch);
00065   
00069   int utfLength(const UCS2* begin, const UCS2* end);
00073   int ucLength(const UCS2* ch);
00074   inline int uc2Length(const UCS2* ch) { return ucLength(ch); }
00078   int uc4Length(const UCS4* ch);
00089   int ToUtf8(byte*& it, byte* end, UCS2 ucc);
00090 
00091   int ToUtf8(char*& it, Char c, char* end = 0) {
00092     return ToUtf8((byte*&)it, (byte*)end, (UCS2)c);
00093   }
00100   int ToUtf8(byte* it, const size_t max, const UCS2* ucbegin, const size_t len);
00101 
00111   UCS2* ToUc(UCS2* buffer, const char* source, int length);
00122   UCS2* ToUc(UCS2* buffer, const UCS4* source, int length = -1);
00133   UCS2* ToUc(UCS2* buffer, const wchar_t* source, int length = -1);
00137   UCS2* utf8ToUc(UCS2* buffer, const char* source, int length = -1);
00138 
00139   inline size_t stringLength(const char* str) { return strlen(str); }
00140   inline size_t stringLength(const UCS2* str) { return ucLength(str); }
00141   inline size_t stringLength(const UCS4* str) { return uc4Length(str); }
00142   
00143   template <typename ToCharType, typename FromCharType>
00144   inline ToCharType* convertTo(ToCharType* buffer, const FromCharType* source, int length = -1)
00145   {
00146     if (length == -1)
00147       length = stringLength(source);
00148     ToCharType* tit = buffer;
00149     register const FromCharType* fit = source;
00150     register const FromCharType* end = source + length;
00151     for (; fit < end; ++tit, ++fit)
00152       *tit = *fit;
00153     return buffer;
00154   }
00155 
00156 } // namespace Unicode
00157 } // namespace text
00158 } // namespace Tanl
00159 
00160 
00161 #endif //Tanl_Text_Utf8Utils_h