sax/inc/xml2utf.hxx

   1 /*************************************************************************
   2  *
   3  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
   4  *
   5  * Copyright 2000, 2010 Oracle and/or its affiliates.
   6  *
   7  * OpenOffice.org - a multi-platform office productivity suite
   8  *
   9  * This file is part of OpenOffice.org.
  10  *
  11  * OpenOffice.org is free software: you can redistribute it and/or modify
  12  * it under the terms of the GNU Lesser General Public License version 3
  13  * only, as published by the Free Software Foundation.
  14  *
  15  * OpenOffice.org is distributed in the hope that it will be useful,
  16  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  17  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  18  * GNU Lesser General Public License version 3 for more details
  19  * (a copy is included in the LICENSE file that accompanied this code).
  20  *
  21  * You should have received a copy of the GNU Lesser General Public License
  22  * version 3 along with OpenOffice.org.  If not, see
  23  * <http://www.openoffice.org/license.html>
  24  * for a copy of the LGPLv3 License.
  25  *
  26  ************************************************************************/
  27
  28 // TODO: Woher?
  29 #define Max( a, b )     (((a)>(b)) ? (a) : (b) )
  30 #define Min( a, b )     (((a)<(b)) ? (a) : (b) )
  31
  32 /*
  33 *
  34 * Text2UnicodeConverter
  35 *
  36 **/
  37 namespace sax_expatwrap {
  38
  39 class Text2UnicodeConverter
  40 {
  41
  42 public:
  43     Text2UnicodeConverter( const ::rtl::OString & sEncoding );
  44     ~Text2UnicodeConverter();
  45
  46     ::com::sun::star::uno::Sequence < sal_Unicode > convert( const ::com::sun::star::uno::Sequence<sal_Int8> & );
  47     sal_Bool canContinue() {  return m_bCanContinue; }
  48
  49 private:
  50     void init( rtl_TextEncoding encoding );
  51
  52     rtl_TextToUnicodeConverter  m_convText2Unicode;
  53     rtl_TextToUnicodeContext    m_contextText2Unicode;
  54     sal_Bool                    m_bCanContinue;
  55     sal_Bool                    m_bInitialized;
  56     rtl_TextEncoding            m_rtlEncoding;
  57     ::com::sun::star::uno::Sequence<sal_Int8> m_seqSource;
  58 };
  59
  60 /*----------------------------------------
  61 *
  62 * Unicode2TextConverter
  63 *
  64 **-----------------------------------------*/
  65 class Unicode2TextConverter
  66 {
  67 public:
  68     Unicode2TextConverter( rtl_TextEncoding encoding );
  69     ~Unicode2TextConverter();
  70
  71     inline ::com::sun::star::uno::Sequence<sal_Int8> convert( const ::rtl::OUString &s )
  72         {
  73             return convert( s.getStr() , s.getLength() );
  74         }
  75     ::com::sun::star::uno::Sequence<sal_Int8> convert( const sal_Unicode * , sal_Int32 nLength );
  76     sal_Bool canContinue() {  return m_bCanContinue; }
  77
  78 private:
  79     void init( rtl_TextEncoding encoding );
  80
  81     rtl_UnicodeToTextConverter  m_convUnicode2Text;
  82     rtl_UnicodeToTextContext    m_contextUnicode2Text;
  83     sal_Bool                    m_bCanContinue;
  84     sal_Bool                    m_bInitialized;
  85     rtl_TextEncoding            m_rtlEncoding;
  86     ::com::sun::star::uno::Sequence<sal_Unicode>        m_seqSource;
  87 };
  88
  89
  90
  91 /*----------------------------------------
  92 *
  93 * XMLFile2UTFConverter
  94 *
  95 **-----------------------------------------*/
  96 class XMLFile2UTFConverter
  97 {
  98 public:
  99     XMLFile2UTFConverter( ):
 100         m_bStarted( sal_False ),
 101         m_pText2Unicode( 0 ),
 102         m_pUnicode2Text( 0 )
 103         {}
 104
 105     ~XMLFile2UTFConverter();
 106
 107     void setInputStream( ::com::sun::star::uno::Reference< ::com::sun::star::io::XInputStream > &r ) { m_in = r; }
 108     void setEncoding( const ::rtl::OString &s ) { m_sEncoding = s; }
 109
 110
 111
 112     // @param nMaxToRead The number of chars, that should be read. Note that this is no exact number. There
 113     //                   may be returned less or more bytes than ordered.
 114     sal_Int32 readAndConvert( ::com::sun::star::uno::Sequence<sal_Int8> &seq , sal_Int32 nMaxToRead )
 115         throw ( ::com::sun::star::io::IOException,
 116                 ::com::sun::star::io::NotConnectedException ,
 117                 ::com::sun::star::io::BufferSizeExceededException ,
 118                 ::com::sun::star::uno::RuntimeException );
 119
 120 private:
 121
 122     // Called only on first Sequence of bytes. Tries to figure out file format and encoding information.
 123     // @return TRUE, when encoding information could be retrieved
 124     // @return FALSE, when no encoding information was found in file
 125     sal_Bool scanForEncoding( ::com::sun::star::uno::Sequence<sal_Int8> &seq );
 126
 127     // Called only on first Sequence of bytes. Tries to figure out
 128     // if enough data is available to scan encoding
 129     // @return TRUE, when encoding is retrievable
 130     // @return FALSE, when more data is needed
 131     sal_Bool isEncodingRecognizable( const ::com::sun::star::uno::Sequence< sal_Int8 > & seq );
 132
 133     // When encoding attribute is within the text (in the first line), it is removed.
 134     void removeEncoding( ::com::sun::star::uno::Sequence<sal_Int8> &seq );
 135
 136     // Initializes decoding depending on m_sEncoding setting
 137     void initializeDecoding();
 138 private:
 139     ::com::sun::star::uno::Reference< ::com::sun::star::io::XInputStream >  m_in;
 140
 141     sal_Bool m_bStarted;
 142     ::rtl::OString m_sEncoding;
 143
 144     Text2UnicodeConverter *m_pText2Unicode;
 145     Unicode2TextConverter *m_pUnicode2Text;
 146 };
 147 }