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 #if defined(__GNUG__) && !defined(NO_GCC_PRAGMA)
  14 #pragma interface "encconv.h"
  15 #endif
  16
  17 #include "wx/defs.h"
  18
  19 #include "wx/object.h"
  20 #include "wx/fontenc.h"
  21 #include "wx/dynarray.h"
  22
  23 // ----------------------------------------------------------------------------
  24 // constants
  25 // ----------------------------------------------------------------------------
  26
  27 enum
  28 {
  29     wxCONVERT_STRICT,
  30     wxCONVERT_SUBSTITUTE
  31 };
  32
  33
  34 enum
  35 {
  36     wxPLATFORM_CURRENT = -1,
  37
  38     wxPLATFORM_UNIX = 0,
  39     wxPLATFORM_WINDOWS,
  40     wxPLATFORM_OS2,
  41     wxPLATFORM_MAC
  42 };
  43
  44 // ----------------------------------------------------------------------------
  45 // types
  46 // ----------------------------------------------------------------------------
  47
  48 WX_DEFINE_ARRAY_INT(wxFontEncoding, wxFontEncodingArray);
  49
  50 //--------------------------------------------------------------------------------
  51 // wxEncodingConverter
  52 //                  This class is capable of converting strings between any two
  53 //                  8bit encodings/charsets. It can also convert from/to Unicode
  54 //--------------------------------------------------------------------------------
  55
  56 class WXDLLIMPEXP_BASE wxEncodingConverter : public wxObject
  57 {
  58     public:
  59
  60             wxEncodingConverter();
  61             ~wxEncodingConverter() { if (m_Table) delete[] m_Table; }
  62
  63             // Initialize conversion. Both output or input encoding may
  64             // be wxFONTENCODING_UNICODE, but only if wxUSE_WCHAR_T is set to 1.
  65             //
  66             // All subsequent calls to Convert() will interpret it's argument
  67             // as a string in input_enc encoding and will output string in
  68             // output_enc encoding.
  69             //
  70             // You must call this method before calling Convert. You may call
  71             // it more than once in order to switch to another conversion
  72             //
  73             // Method affects behaviour of Convert() in case input character
  74             // cannot be converted because it does not exist in output encoding:
  75             //     wxCONVERT_STRICT --
  76             //              follow behaviour of GNU Recode - just copy unconvertable
  77             //              characters to output and don't change them (it's integer
  78             //              value will stay the same)
  79             //     wxCONVERT_SUBSTITUTE --
  80             //              try some (lossy) substitutions - e.g. replace
  81             //              unconvertable latin capitals with acute by ordinary
  82             //              capitals, replace en-dash or em-dash by '-' etc.
  83             //     both modes gurantee that output string will have same length
  84             //     as input string
  85             //
  86             // Returns false if given conversion is impossible, true otherwise
  87             // (conversion may be impossible either if you try to convert
  88             // to Unicode with non-Unicode build of wxWidgets or if input
  89             // or output encoding is not supported.)
  90             bool Init(wxFontEncoding input_enc, wxFontEncoding output_enc, int method = wxCONVERT_STRICT);
  91
  92             // Convert input string according to settings passed to Init.
  93             // Note that you must call Init before using Convert!
  94             bool Convert(const char* input, char* output) const;
  95             bool Convert(char* str) const { return Convert(str, str); }
  96             wxString Convert(const wxString& input) const;
  97
  98 #if wxUSE_WCHAR_T
  99             bool Convert(const char* input, wchar_t* output) const;
 100             bool Convert(const wchar_t* input, char* output) const;
 101             bool Convert(const wchar_t* input, wchar_t* output) const;
 102             bool Convert(wchar_t* str) const { return Convert(str, str); }
 103 #endif
 104             // Return equivalent(s) for given font that are used
 105             // under given platform. wxPLATFORM_CURRENT means the plaform
 106             // this binary was compiled for
 107             //
 108             // Examples:
 109             //     current platform          enc    returned value
 110             // -----------------------------------------------------
 111             //     unix                   CP1250         {ISO8859_2}
 112             //     unix                ISO8859_2                  {}
 113             //     windows             ISO8859_2            {CP1250}
 114             //
 115             // Equivalence is defined in terms of convertibility:
 116             // 2 encodings are equivalent if you can convert text between
 117             // then without loosing information (it may - and will - happen
 118             // that you loose special chars like quotation marks or em-dashes
 119             // but you shouldn't loose any diacritics and language-specific
 120             // characters when converting between equivalent encodings).
 121             //
 122             // Convert() method is not limited to converting between
 123             // equivalent encodings, it can convert between arbitrary
 124             // two encodings!
 125             //
 126             // Remember that this function does _NOT_ check for presence of
 127             // fonts in system. It only tells you what are most suitable
 128             // encodings. (It usually returns only one encoding)
 129             //
 130             // Note that argument enc itself may be present in returned array!
 131             // (so that you can -- as a side effect -- detect whether the
 132             // encoding is native for this platform or not)
 133             static wxFontEncodingArray GetPlatformEquivalents(wxFontEncoding enc, int platform = wxPLATFORM_CURRENT);
 134
 135             // Similar to GetPlatformEquivalent, but this one will return ALL
 136             // equivalent encodings, regardless the platform, including itself.
 137             static wxFontEncodingArray GetAllEquivalents(wxFontEncoding enc);
 138
 139             // Return true if [any text in] one multibyte encoding can be
 140             // converted to another one losslessly.
 141             //
 142             // Do not call this with wxFONTENCODING_UNICODE, it doesn't make
 143             // sense (always works in one sense and always depends on the text
 144             // to convert in the other)
 145             static bool CanConvert(wxFontEncoding encIn, wxFontEncoding encOut)
 146             {
 147                 return GetAllEquivalents(encIn).Index(encOut) != wxNOT_FOUND;
 148             }
 149
 150     private:
 151
 152 #if wxUSE_WCHAR_T
 153             wchar_t *m_Table;
 154 #else
 155             char *m_Table;
 156 #endif
 157             bool m_UnicodeInput, m_UnicodeOutput;
 158             bool m_JustCopy;
 159
 160     DECLARE_NO_COPY_CLASS(wxEncodingConverter)
 161 };
 162
 163 #endif  // _WX_ENCCONV_H_