StarEncoding.hxx
Go to the documentation of this file.
00001 /* -*- Mode: C++; c-default-style: "k&r"; indent-tabs-mode: nil; tab-width: 2; c-basic-offset: 2 -*- */
00002 
00003 /* libstaroffice
00004 * Version: MPL 2.0 / LGPLv2+
00005 *
00006 * The contents of this file are subject to the Mozilla Public License Version
00007 * 2.0 (the "License"); you may not use this file except in compliance with
00008 * the License or as specified alternatively below. You may obtain a copy of
00009 * the License at http://www.mozilla.org/MPL/
00010 *
00011 * Software distributed under the License is distributed on an "AS IS" basis,
00012 * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License
00013 * for the specific language governing rights and limitations under the
00014 * License.
00015 *
00016 * Major Contributor(s):
00017 * Copyright (C) 2002 William Lachance (wrlach@gmail.com)
00018 * Copyright (C) 2002,2004 Marc Maurer (uwog@uwog.net)
00019 * Copyright (C) 2004-2006 Fridrich Strba (fridrich.strba@bluewin.ch)
00020 * Copyright (C) 2006, 2007 Andrew Ziem
00021 * Copyright (C) 2011, 2012 Alonso Laurent (alonso@loria.fr)
00022 *
00023 *
00024 * All Rights Reserved.
00025 *
00026 * For minor contributions see the git repository.
00027 *
00028 * Alternatively, the contents of this file may be used under the terms of
00029 * the GNU Lesser General Public License Version 2 or later (the "LGPLv2+"),
00030 * in which case the provisions of the LGPLv2+ are applicable
00031 * instead of those above.
00032 */
00033 
00034 /*
00035  * StarEncoding to read/parse some basic encoding in StarOffice documents
00036  *
00037  */
00038 #ifndef STAR_ENCODING
00039 #  define STAR_ENCODING
00040 
00041 #include <vector>
00042 
00043 #include "libstaroffice_internal.hxx"
00044 
00050 class StarEncoding
00051 {
00052 public:
00054   enum Encoding {
00055     E_DONTKNOW=0,E_MS_1252,E_APPLE_ROMAN,E_IBM_437,E_IBM_850,E_IBM_860,E_IBM_861,E_IBM_863,
00056     E_IBM_865,/*SYSTEM=9,*/E_SYMBOL=10,E_ASCII_US,E_ISO_8859_1,E_ISO_8859_2,E_ISO_8859_3,E_ISO_8859_4,
00057     E_ISO_8859_5,E_ISO_8859_6,E_ISO_8859_7,E_ISO_8859_8,E_ISO_8859_9,E_ISO_8859_14,E_ISO_8859_15,E_IBM_737,
00058     E_IBM_775,E_IBM_852,E_IBM_855,E_IBM_857,E_IBM_862,E_IBM_864,E_IBM_866,E_IBM_869,
00059     E_MS_874,E_MS_1250,E_MS_1251,E_MS_1253,E_MS_1254,E_MS_1255,E_MS_1256,E_MS_1257,
00060     // 40
00061     E_MS_1258,/*E_APPLE_ARABIC,*/E_APPLE_CENTEURO=42,E_APPLE_CROATIAN,E_APPLE_CYRILLIC,/*E_APPLE_DEVANAGARI,E_APPLE_FARSI,*/E_APPLE_GREEK=47,
00062     /*E_APPLE_GUJARATI,E_APPLE_GURMUKHI,E_APPLE_HEBREW,*/E_APPLE_ICELAND=51,E_APPLE_ROMANIAN,/*E_APPLE_THAI,*/E_APPLE_TURKISH=54,E_APPLE_UKRAINIAN,
00063     E_APPLE_CHINSIMP,E_APPLE_CHINTRAD,E_APPLE_JAPANESE,E_APPLE_KOREAN,E_MS_932,E_MS_936,E_MS_949,E_MS_950,
00064     E_SHIFT_JIS,E_GB_2312,E_GBT_12345,E_GBK,E_BIG5,E_EUC_JP,E_EUC_CN,/*E_EUC_TW,*/
00065     /*E_ISO_2022_JP,E_ISO_2022_CN,*/E_KOI8_R=74,E_UTF7,E_UTF8,E_ISO_8859_10,E_ISO_8859_13,E_EUC_KR,
00066     // 80
00067     /*E_ISO_2022_KR,*/E_JIS_X_0201=81,E_JIS_X_0208,E_JIS_X_0212,E_MS_1361,/*E_GB_18030,*/E_BIG5_HKSCS=86,E_TIS_620,
00068     E_KOI8_U,E_ISCII_DEVANAGARI,
00069 
00070     // E_USER_START=0x8000,E_USER_END=0xEFFF,
00071 
00072     E_UCS4=0xFFFE,E_UCS2=0xFFFF
00073   };
00074 
00076   StarEncoding();
00078   virtual ~StarEncoding();
00079 
00081   static Encoding getEncodingForId(int id);
00083   static bool convert(std::vector<uint8_t> const &src, Encoding encoding, std::vector<uint32_t> &dest, std::vector<size_t> &srcPositions);
00084 
00085 protected:
00091   static bool read(std::vector<uint8_t> const &src, size_t &pos, Encoding encoding, std::vector<uint32_t> &dest);
00092 };
00093 #endif
00094 // vim: set filetype=cpp tabstop=2 shiftwidth=2 cindent autoindent smartindent noexpandtab: