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