include/wx/encconv.h

   1 /////////////////////////////////////////////////////////////////////////////
   2 // Name:        encconv.h
   3 // Purpose:     wxEncodingConverter class for converting between different
   4 //              font encodings
   5 // Author:      Vaclav Slavik
   6 // Copyright:   (c) 1999 Vaclav Slavik
   7 // Licence:     wxWindows Licence
   8 /////////////////////////////////////////////////////////////////////////////
   9
  10 #ifndef __ENCCONV_H__
  11 #define __ENCCONV_H__
  12
  13 #ifdef __GNUG__
  14 #pragma interface "encconv.h"
  15 #endif
  16
  17 #include "wx/defs.h"
  18 #include "wx/font.h"
  19 #include "wx/dynarray.h"
  20
  21
  22
  23 enum {
  24     wxCONVERT_STRICT,
  25     wxCONVERT_SUBSTITUTE
  26 };
  27
  28
  29 enum {
  30     wxPLATFORM_CURRENT = -1,
  31
  32     wxPLATFORM_UNIX = 0,
  33     wxPLATFORM_WINDOWS,
  34     wxPLATFORM_OS2,
  35     wxPLATFORM_MAC,
  36 };
  37
  38 WX_DEFINE_ARRAY(wxFontEncoding, wxFontEncodingArray);
  39
  40
  41 //--------------------------------------------------------------------------------
  42 // wxEncodingConverter
  43 //                  This class is capable of converting strings between any two
  44 //                  8bit encodings/charsets. It can also convert from/to Unicode
  45 //--------------------------------------------------------------------------------
  46
  47 class WXDLLEXPORT wxEncodingConverter : public wxObject
  48 {
  49     public:
  50
  51             wxEncodingConverter();
  52             ~wxEncodingConverter() { if (m_Table) delete[] m_Table; }
  53
  54             // Initialize convertion. Both output or input encoding may
  55             // be wxFONTENCODING_UNICODE, but only if wxUSE_ENCODING is set to 1.
  56             //
  57             // All subsequent calls to Convert() will interpret it's argument
  58             // as a string in input_enc encoding and will output string in
  59             // output_enc encoding.
  60             //
  61             // You must call this method before calling Convert. You may call
  62             // it more than once in order to switch to another conversion
  63             //
  64             // Method affects behaviour of Convert() in case input character
  65             // cannot be converted because it does not exist in output encoding:
  66             //     wxCONVERT_STRICT --
  67             //              follow behaviour of GNU Recode - just copy unconvertable
  68             //              characters to output and don't change them (it's integer
  69             //              value will stay the same)
  70             //     wxCONVERT_SUBSTITUTE --
  71             //              try some (lossy) substitutions - e.g. replace
  72             //              unconvertable latin capitals with acute by ordinary
  73             //              capitals, replace en-dash or em-dash by '-' etc.
  74             //     both modes gurantee that output string will have same length
  75             //     as input string
  76             //
  77             // Returns FALSE if given conversion is impossible, TRUE otherwise
  78             // (conversion may be impossible either if you try to convert
  79             // to Unicode with non-Unicode build of wxWindows or if input
  80             // or output encoding is not supported.)
  81             bool Init(wxFontEncoding input_enc, wxFontEncoding output_enc, int method = wxCONVERT_STRICT);
  82
  83             // Convert input string according to settings passed to Init.
  84             // Note that you must call Init before using Convert!
  85             wxString Convert(const wxString& input);
  86             void Convert(const wxChar* input, wxChar* output);
  87             void Convert(wxChar* str) { Convert(str, str); }
  88
  89             // Return equivalent(s) for given font that are used
  90             // under given platform. wxPLATFORM_CURRENT means the plaform
  91             // this binary was compiled for
  92             //
  93             // Examples:
  94             //     current platform          enc    returned value
  95             // -----------------------------------------------------
  96             //     unix                   CP1250         {ISO8859_2}
  97             //     unix                ISO8859_2                  {}
  98             //     windows             ISO8859_2            {CP1250}
  99             //
 100             // Equivalence is defined in terms of convertibility:
 101             // 2 encodings are equivalent if you can convert text between
 102             // then without loosing information (it may - and will - happen
 103             // that you loose special chars like quotation marks or em-dashes
 104             // but you shouldn't loose any diacritics and language-specific
 105             // characters when converting between equivalent encodings).
 106             //
 107             // Convert() method is not limited to converting between
 108             // equivalent encodings, it can convert between arbitrary
 109             // two encodings!
 110             //
 111             // Remember that this function does _NOT_ check for presence of
 112             // fonts in system. It only tells you what are most suitable
 113             // encodings. (It usually returns only one encoding)
 114             //
 115             // Note that argument enc itself may be present in returned array!
 116             // (so that you can -- as a side effect -- detect whether the
 117             // encoding is native for this platform or not)
 118             static wxFontEncodingArray GetPlatformEquivalents(wxFontEncoding enc, int platform = wxPLATFORM_CURRENT);
 119
 120             // Similar to GetPlatformEquivalent, but this one will return ALL
 121             // equivalent encodings, regardless the platform, including itself.
 122             static wxFontEncodingArray GetAllEquivalents(wxFontEncoding enc);
 123
 124     private:
 125
 126             wxChar *m_Table;
 127             bool m_UnicodeInput;
 128             bool m_JustCopy;
 129
 130 };
 131
 132
 133 #endif  // __ENCCONV_H__
 134
 135
 136
 137
 138
 139
 140
 141
 142
 143
 144
 145
 146
 147
 148
 149
 150