interface/wx/encconv.h

   1 /////////////////////////////////////////////////////////////////////////////
   2 // Name:        encconv.h
   3 // Purpose:     interface of wxEncodingConverter
   4 // Author:      wxWidgets team
   5 // RCS-ID:      $Id$
   6 // Licence:     wxWindows license
   7 /////////////////////////////////////////////////////////////////////////////
   8
   9 /**
  10     @class wxEncodingConverter
  11
  12     This class is capable of converting strings between two 8-bit encodings/charsets.
  13     It can also convert from/to Unicode (but only if you compiled wxWidgets
  14     with @c wxUSE_WCHAR_T set to 1).
  15
  16     Only a limited subset of encodings is supported by wxEncodingConverter:
  17     @c wxFONTENCODING_ISO8859_1..15, @c wxFONTENCODING_CP1250..1257 and
  18     @c wxFONTENCODING_KOI8.
  19
  20     @note
  21
  22     Please use wxMBConv classes instead if possible. wxCSConv has much better
  23     support for various encodings than wxEncodingConverter.
  24     wxEncodingConverter is useful only if you rely on wxCONVERT_SUBSTITUTE mode
  25     of operation (see wxEncodingConverter::Init()).
  26
  27     @library{wxbase}
  28     @category{misc}
  29
  30     @see wxFontMapper, wxMBConv, @ref overview_nonenglish
  31 */
  32 class wxEncodingConverter : public wxObject
  33 {
  34 public:
  35     /**
  36         Constructor.
  37     */
  38     wxEncodingConverter();
  39
  40     /**
  41         Return @true if (any text in) multibyte encoding @a encIn can be converted to
  42         another one (@a encOut) losslessly.
  43
  44         Do not call this method with @c wxFONTENCODING_UNICODE as either parameter,
  45         it doesn't make sense (always works in one sense and always depends
  46         on the text to convert in the other).
  47     */
  48     static bool CanConvert(wxFontEncoding encIn,
  49                            wxFontEncoding encOut);
  50
  51     /**
  52         @name Conversion functions
  53
  54         @{
  55     */
  56     /**
  57         Convert input string according to settings passed to Init() and writes
  58         the result to output.
  59
  60         All the Convert() function overloads return @true if the conversion was
  61         lossless and @false if at least one of the characters couldn't be converted
  62         was and replaced with '?' in the output.
  63
  64         Note that if @c wxCONVERT_SUBSTITUTE was passed to Init(), substitution is
  65         considered a lossless operation.
  66
  67         @note You must call Init() before using this method!
  68
  69         @note wchar_t versions of the method are not available if wxWidgets was
  70               compiled with @c wxUSE_WCHAR_T set to 0.
  71     */
  72     bool Convert(const char* input, char* output) const;
  73     bool Convert(const wchar_t* input, wchar_t* output) const;
  74     bool Convert(const char* input, wchar_t* output) const;
  75     bool Convert(const wchar_t* input, char* output) const;
  76
  77     /**
  78         Convert input string according to settings passed to Init() in-place,
  79         i.e. write the result to the same memory area.
  80
  81         See the Convert(const char*,char*) const overload for more info.
  82     */
  83     bool Convert(char* str) const;
  84     bool Convert(wchar_t* str) const;
  85
  86     /**
  87         Convert a wxString and return a new wxString object.
  88
  89         See the Convert(const char*,char*) const overload for more info.
  90     */
  91     wxString Convert(const wxString& input) const;
  92     //@}
  93
  94
  95     /**
  96         Similar to GetPlatformEquivalents(), but this one will return ALL
  97         equivalent encodings, regardless of the platform, and including itself.
  98
  99         This platform's encodings are before others in the array.
 100         And again, if @a enc is in the array, it is the very first item in it.
 101     */
 102     static wxFontEncodingArray GetAllEquivalents(wxFontEncoding enc);
 103
 104     /**
 105         Return equivalents for given font that are used under given platform.
 106
 107         Supported platforms:
 108         @li wxPLATFORM_UNIX
 109         @li wxPLATFORM_WINDOWS
 110         @li wxPLATFORM_OS2
 111         @li wxPLATFORM_MAC
 112         @li wxPLATFORM_CURRENT
 113
 114         wxPLATFORM_CURRENT means the platform this binary was compiled for.
 115
 116         Examples:
 117
 118         @verbatim
 119         current platform   enc          returned value
 120         ----------------------------------------------
 121         unix            CP1250             {ISO8859_2}
 122         unix         ISO8859_2             {ISO8859_2}
 123         windows      ISO8859_2                {CP1250}
 124         unix            CP1252  {ISO8859_1,ISO8859_15}
 125         @endverbatim
 126
 127         Equivalence is defined in terms of convertibility: two encodings are
 128         equivalent if you can convert text between then without losing
 129         information (it may - and will - happen that you lose special chars
 130         like quotation marks or em-dashes but you shouldn't lose any diacritics
 131         and language-specific characters when converting between equivalent encodings).
 132
 133         Remember that this function does @b NOT check for presence of
 134         fonts in system. It only tells you what are most suitable
 135         encodings. (It usually returns only one encoding.)
 136
 137         @note Note that argument enc itself may be present in the returned array,
 138               so that you can, as a side-effect, detect whether the encoding is
 139               native for this platform or not.
 140
 141         @note Convert() is not limited to converting between equivalent encodings,
 142               it can convert between two arbitrary encodings.
 143
 144         @note If @a enc is present in the returned array, then it is always the first
 145               item of it.
 146
 147         @note Please note that the returned array may contain no items at all.
 148     */
 149     static wxFontEncodingArray GetPlatformEquivalents(wxFontEncoding enc,
 150                                                       int platform = wxPLATFORM_CURRENT);
 151
 152     /**
 153         Initialize the conversion.
 154
 155         Both output or input encoding may be wxFONTENCODING_UNICODE, but only
 156         if wxUSE_ENCODING is set to 1.
 157
 158         All subsequent calls to Convert() will interpret its argument
 159         as a string in @a input_enc encoding and will output string in
 160         @a output_enc encoding.
 161
 162         You must call this method before calling Convert. You may call
 163         it more than once in order to switch to another conversion.
 164
 165         @a method affects behaviour of Convert() in case input character
 166         cannot be converted because it does not exist in output encoding:
 167
 168         @li @b wxCONVERT_STRICT: follow behaviour of GNU Recode - just copy
 169             unconvertible  characters to output and don't change them
 170             (its integer value will stay the same)
 171         @li @b wxCONVERT_SUBSTITUTE:  try some (lossy) substitutions - e.g.
 172             replace unconvertible latin capitals with acute by ordinary
 173             capitals, replace en-dash or em-dash by '-' etc.
 174
 175         Both modes guarantee that output string will have same length
 176         as input string.
 177
 178         @return @false if given conversion is impossible, @true otherwise
 179                 (conversion may be impossible either if you try to convert
 180                 to Unicode with non-Unicode build of wxWidgets or if input
 181                 or output encoding is not supported).
 182     */
 183     bool Init(wxFontEncoding input_enc, wxFontEncoding output_enc,
 184               int method = wxCONVERT_STRICT);
 185 };
 186