src/common/unichar.cpp

   1 /////////////////////////////////////////////////////////////////////////////
   2 // Name:        src/common/unichar.cpp
   3 // Purpose:     wxUniChar and wxUniCharRef classes
   4 // Author:      Vaclav Slavik
   5 // Created:     2007-03-19
   6 // RCS-ID:      $Id$
   7 // Copyright:   (c) 2007 REA Elektronik GmbH
   8 // Licence:     wxWindows licence
   9 ///////////////////////////////////////////////////////////////////////////////
  10
  11 // ===========================================================================
  12 // headers
  13 // ===========================================================================
  14
  15 // For compilers that support precompilation, includes "wx.h".
  16 #include "wx/wxprec.h"
  17
  18 #ifdef __BORLANDC__
  19     #pragma hdrstop
  20 #endif
  21
  22 #ifndef WX_PRECOMP
  23     #include "wx/strconv.h"  // wxConvLibc
  24 #endif
  25
  26 #include "wx/unichar.h"
  27 #include "wx/string.h"
  28
  29 // ===========================================================================
  30 // implementation
  31 // ===========================================================================
  32
  33 // ---------------------------------------------------------------------------
  34 // wxUniChar
  35 // ---------------------------------------------------------------------------
  36
  37 /* static */
  38 wxUniChar::value_type wxUniChar::From8bit(char c)
  39 {
  40     // all supported charsets have the first 128 characters same as ASCII:
  41     if ( (unsigned char)c < 0x80 )
  42         return c;
  43
  44 #if wxUSE_UTF8_LOCALE_ONLY
  45     wxFAIL_MSG( _T("invalid UTF-8 character") );
  46     return wxT('?'); // FIXME-UTF8: what to use as failure character?
  47 #else
  48     wchar_t buf[2];
  49     if ( wxConvLibc.ToWChar(buf, 2, &c, 1) != 2 )
  50         return wxT('?'); // FIXME-UTF8: what to use as failure character?
  51     return buf[0];
  52 #endif
  53 }
  54
  55 /* static */
  56 char wxUniChar::To8bit(wxUniChar::value_type c)
  57 {
  58     // all supported charsets have the first 128 characters same as ASCII:
  59     if ( c < 0x80 )
  60         return c;
  61
  62 #if wxUSE_UTF8_LOCALE_ONLY
  63     wxFAIL_MSG( _T("character cannot be converted to single UTF-8 byte") );
  64     return '?'; // FIXME-UTF8: what to use as failure character?
  65 #else
  66     wchar_t in = c;
  67     char buf[2];
  68     if ( wxConvLibc.FromWChar(buf, 2, &in, 1) != 2 )
  69         return '?'; // FIXME-UTF8: what to use as failure character?
  70     return buf[0];
  71 #endif
  72 }
  73
  74
  75 // ---------------------------------------------------------------------------
  76 // wxUniCharRef
  77 // ---------------------------------------------------------------------------
  78
  79 #if wxUSE_UNICODE_UTF8
  80 wxUniChar wxUniCharRef::UniChar() const
  81 {
  82     return wxStringOperations::DecodeChar(m_pos);
  83 }
  84
  85 wxUniCharRef& wxUniCharRef::operator=(const wxUniChar& c)
  86 {
  87     wxStringOperations::Utf8CharBuffer utf(wxStringOperations::EncodeChar(c));
  88     size_t lenOld = wxStringOperations::GetUtf8CharLength(*m_pos);
  89     size_t lenNew = wxStringOperations::GetUtf8CharLength(utf[0]);
  90
  91     if ( lenNew == lenOld )
  92     {
  93         // this is the simpler case: if the new value's UTF-8 code has the
  94         // same length, we can just replace it:
  95
  96         iterator pos(m_pos);
  97         for ( size_t i = 0; i < lenNew; ++i, ++pos )
  98             *pos = utf[i];
  99     }
 100     else
 101     {
 102         // the worse case is when the new value has either longer or shorter
 103         // code -- in that case, we have to use wxStringImpl::replace() and
 104         // this invalidates all iterators, so we have to update them too:
 105
 106         wxString& str = *wx_const_cast(wxString*, m_node.m_str);
 107         wxStringImpl& strimpl = str.m_impl;
 108
 109         int iterDiff = lenNew - lenOld;
 110         size_t posIdx = m_pos - strimpl.begin();
 111
 112         // compute positions of outstanding iterators for this string after the
 113         // replacement is done (there is only a small number of iterators at
 114         // any time, so we use an array on the stack to avoid unneeded
 115         // allocation):
 116         static const size_t STATIC_SIZE = 32;
 117         size_t indexes_a[STATIC_SIZE];
 118         size_t *indexes = indexes_a;
 119         size_t iterNum = 0;
 120         wxStringIteratorNode *it;
 121         for ( it = str.m_iterators.ptr; it; it = it->m_next, ++iterNum )
 122         {
 123             wxASSERT( it->m_iter || it->m_citer );
 124
 125             if ( iterNum == STATIC_SIZE )
 126             {
 127                 wxLogTrace( _T("utf8"), _T("unexpectedly many iterators") );
 128
 129                 size_t total = iterNum + 1;
 130                 for ( wxStringIteratorNode *it2 = it; it2; it2 = it2->m_next )
 131                     total++;
 132                 indexes = new size_t[total];
 133                 memcpy(indexes, indexes_a, sizeof(size_t) * STATIC_SIZE);
 134             }
 135
 136             size_t idx = it->m_iter
 137                          ? (*it->m_iter - strimpl.begin())
 138                          : (*it->m_citer - strimpl.begin());
 139
 140             if ( idx > posIdx )
 141                 idx += iterDiff;
 142
 143             indexes[iterNum] = idx;
 144         }
 145
 146         // update the string:
 147         strimpl.replace(m_pos, m_pos + lenOld, utf, lenNew);
 148
 149         // finally, set the iterators to valid values again (note that this
 150         // updates m_pos as well):
 151         size_t i;
 152         for ( i = 0, it = str.m_iterators.ptr; it; it = it->m_next, ++i )
 153         {
 154             wxASSERT( i < iterNum );
 155             wxASSERT( it->m_iter || it->m_citer );
 156
 157             if ( it->m_iter )
 158                 *it->m_iter = strimpl.begin() + indexes[i];
 159             else // it->m_citer
 160                 *it->m_citer = strimpl.begin() + indexes[i];
 161         }
 162
 163         if ( indexes != indexes_a )
 164             delete[] indexes;
 165     }
 166
 167     return *this;
 168 }
 169 #endif // wxUSE_UNICODE_UTF8