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     #include "wx/log.h"
  25 #endif
  26
  27 #include "wx/unichar.h"
  28 #include "wx/string.h"
  29
  30 // ===========================================================================
  31 // implementation
  32 // ===========================================================================
  33
  34 // ---------------------------------------------------------------------------
  35 // wxUniChar
  36 // ---------------------------------------------------------------------------
  37
  38 /* static */
  39 wxUniChar::value_type wxUniChar::FromHi8bit(char c)
  40 {
  41 #if wxUSE_UTF8_LOCALE_ONLY
  42     wxFAIL_MSG( "invalid UTF-8 character" );
  43     return wxT('?'); // FIXME-UTF8: what to use as failure character?
  44 #else
  45     wchar_t buf[2];
  46     if ( wxConvLibc.ToWChar(buf, 2, &c, 1) != 2 )
  47     {
  48         wxFAIL_MSG( "invalid multibyte character" );
  49         return wxT('?'); // FIXME-UTF8: what to use as failure character?
  50     }
  51     return buf[0];
  52 #endif
  53 }
  54
  55 /* static */
  56 char wxUniChar::ToHi8bit(wxUniChar::value_type c)
  57 {
  58 #if wxUSE_UTF8_LOCALE_ONLY
  59     wxFAIL_MSG( "character cannot be converted to single UTF-8 byte" );
  60     return '?'; // FIXME-UTF8: what to use as failure character?
  61 #else
  62     wchar_t in = c;
  63     char buf[2];
  64     if ( wxConvLibc.FromWChar(buf, 2, &in, 1) != 2 )
  65     {
  66         wxFAIL_MSG( "character cannot be converted to single byte" );
  67         return '?'; // FIXME-UTF8: what to use as failure character?
  68     }
  69     return buf[0];
  70 #endif
  71 }
  72
  73
  74 // ---------------------------------------------------------------------------
  75 // wxUniCharRef
  76 // ---------------------------------------------------------------------------
  77
  78 #if wxUSE_UNICODE_UTF8
  79 wxUniChar wxUniCharRef::UniChar() const
  80 {
  81     return wxStringOperations::DecodeChar(m_pos);
  82 }
  83
  84 wxUniCharRef& wxUniCharRef::operator=(const wxUniChar& c)
  85 {
  86     wxStringOperations::Utf8CharBuffer utf(wxStringOperations::EncodeChar(c));
  87     size_t lenOld = wxStringOperations::GetUtf8CharLength(*m_pos);
  88     size_t lenNew = wxStringOperations::GetUtf8CharLength(utf[0]);
  89
  90     if ( lenNew == lenOld )
  91     {
  92         // this is the simpler case: if the new value's UTF-8 code has the
  93         // same length, we can just replace it:
  94
  95         iterator pos(m_pos);
  96         for ( size_t i = 0; i < lenNew; ++i, ++pos )
  97             *pos = utf[i];
  98     }
  99     else
 100     {
 101         // the worse case is when the new value has either longer or shorter
 102         // code -- in that case, we have to use wxStringImpl::replace() and
 103         // this invalidates all iterators, so we have to update them too:
 104
 105         wxString& str = *wx_const_cast(wxString*, m_node.m_str);
 106         wxStringImpl& strimpl = str.m_impl;
 107
 108         int iterDiff = lenNew - lenOld;
 109         size_t posIdx = m_pos - strimpl.begin();
 110
 111         // compute positions of outstanding iterators for this string after the
 112         // replacement is done (there is only a small number of iterators at
 113         // any time, so we use an array on the stack to avoid unneeded
 114         // allocation):
 115         static const size_t STATIC_SIZE = 32;
 116         size_t indexes_a[STATIC_SIZE];
 117         size_t *indexes = indexes_a;
 118         size_t iterNum = 0;
 119         wxStringIteratorNode *it;
 120         for ( it = str.m_iterators.ptr; it; it = it->m_next, ++iterNum )
 121         {
 122             wxASSERT( it->m_iter || it->m_citer );
 123
 124             if ( iterNum == STATIC_SIZE )
 125             {
 126                 wxLogTrace( _T("utf8"), _T("unexpectedly many iterators") );
 127
 128                 size_t total = iterNum + 1;
 129                 for ( wxStringIteratorNode *it2 = it; it2; it2 = it2->m_next )
 130                     total++;
 131                 indexes = new size_t[total];
 132                 memcpy(indexes, indexes_a, sizeof(size_t) * STATIC_SIZE);
 133             }
 134
 135             size_t idx = it->m_iter
 136                          ? (*it->m_iter - strimpl.begin())
 137                          : (*it->m_citer - strimpl.begin());
 138
 139             if ( idx > posIdx )
 140                 idx += iterDiff;
 141
 142             indexes[iterNum] = idx;
 143         }
 144
 145         // update the string:
 146         strimpl.replace(m_pos, m_pos + lenOld, utf, lenNew);
 147
 148         // finally, set the iterators to valid values again (note that this
 149         // updates m_pos as well):
 150         size_t i;
 151         for ( i = 0, it = str.m_iterators.ptr; it; it = it->m_next, ++i )
 152         {
 153             wxASSERT( i < iterNum );
 154             wxASSERT( it->m_iter || it->m_citer );
 155
 156             if ( it->m_iter )
 157                 *it->m_iter = strimpl.begin() + indexes[i];
 158             else // it->m_citer
 159                 *it->m_citer = strimpl.begin() + indexes[i];
 160         }
 161
 162         if ( indexes != indexes_a )
 163             delete[] indexes;
 164     }
 165
 166     return *this;
 167 }
 168 #endif // wxUSE_UNICODE_UTF8