include/wx/encconv.h

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