// UTF-8
// ----------------------------------------------------------------------------
-static wxUint32 utf8_max[]=
+static const wxUint32 utf8_max[]=
{ 0x7f, 0x7ff, 0xffff, 0x1fffff, 0x3ffffff, 0x7fffffff, 0xffffffff };
// boundaries of the private use area we use to (temporarily) remap invalid
const wxUint32 wxUnicodePUAEnd = wxUnicodePUA + 256;
// this table gives the length of the UTF-8 encoding from its first character:
-unsigned char tableUtf8Lengths[256] = {
+const unsigned char tableUtf8Lengths[256] = {
// single-byte sequences (ASCII):
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 00..0F
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 10..1F
return written;
}
- unsigned char c = *p;
- unsigned len = tableUtf8Lengths[c];
- if ( !len )
+ if ( out && !dstLen-- )
break;
- if ( srcLen < len ) // the test works for wxNO_LEN too
- break;
+ wxUint32 code;
+ unsigned char c = *p;
- if ( srcLen != wxNO_LEN )
- srcLen -= len;
+ if ( c < 0x80 )
+ {
+ if ( srcLen == 0 ) // the test works for wxNO_LEN too
+ break;
- if ( out && !dstLen-- )
- break;
+ if ( srcLen != wxNO_LEN )
+ srcLen--;
+ code = c;
+ }
+ else
+ {
+ unsigned len = tableUtf8Lengths[c];
+ if ( !len )
+ break;
- // Char. number range | UTF-8 octet sequence
- // (hexadecimal) | (binary)
- // ----------------------+---------------------------------------------
- // 0000 0000 - 0000 007F | 0xxxxxxx
- // 0000 0080 - 0000 07FF | 110xxxxx 10xxxxxx
- // 0000 0800 - 0000 FFFF | 1110xxxx 10xxxxxx 10xxxxxx
- // 0001 0000 - 0010 FFFF | 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
- //
- // Code point value is stored in bits marked with 'x', lowest-order bit
- // of the value on the right side in the diagram above.
- // (from RFC 3629)
+ if ( srcLen < len ) // the test works for wxNO_LEN too
+ break;
- // mask to extract lead byte's value ('x' bits above), by sequence length:
- static const unsigned char leadValueMask[] = { 0x7F, 0x1F, 0x0F, 0x07 };
+ if ( srcLen != wxNO_LEN )
+ srcLen -= len;
- // mask and value of lead byte's most significant bits, by length:
- static const unsigned char leadMarkerMask[] = { 0x80, 0xE0, 0xF0, 0xF8 };
- static const unsigned char leadMarkerVal[] = { 0x00, 0xC0, 0xE0, 0xF0 };
+ // Char. number range | UTF-8 octet sequence
+ // (hexadecimal) | (binary)
+ // ----------------------+----------------------------------------
+ // 0000 0000 - 0000 007F | 0xxxxxxx
+ // 0000 0080 - 0000 07FF | 110xxxxx 10xxxxxx
+ // 0000 0800 - 0000 FFFF | 1110xxxx 10xxxxxx 10xxxxxx
+ // 0001 0000 - 0010 FFFF | 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
+ //
+ // Code point value is stored in bits marked with 'x',
+ // lowest-order bit of the value on the right side in the diagram
+ // above. (from RFC 3629)
- len--; // it's more convenient to work with 0-based length here
+ // mask to extract lead byte's value ('x' bits above), by sequence
+ // length:
+ static const unsigned char leadValueMask[] = { 0x7F, 0x1F, 0x0F, 0x07 };
- // extract the lead byte's value bits:
- if ( (c & leadMarkerMask[len]) != leadMarkerVal[len] )
- break;
+ // mask and value of lead byte's most significant bits, by length:
+ static const unsigned char leadMarkerMask[] = { 0x80, 0xE0, 0xF0, 0xF8 };
+ static const unsigned char leadMarkerVal[] = { 0x00, 0xC0, 0xE0, 0xF0 };
- wxUint32 code = c & leadValueMask[len];
+ len--; // it's more convenient to work with 0-based length here
- // all remaining bytes, if any, are handled in the same way regardless of
- // sequence's length:
- for ( ; len; --len )
- {
- c = *++p;
- if ( (c & 0xC0) != 0x80 )
- return wxCONV_FAILED;
+ // extract the lead byte's value bits:
+ if ( (c & leadMarkerMask[len]) != leadMarkerVal[len] )
+ break;
- code <<= 6;
- code |= c & 0x3F;
+ code = c & leadValueMask[len];
+
+ // all remaining bytes, if any, are handled in the same way
+ // regardless of sequence's length:
+ for ( ; len; --len )
+ {
+ c = *++p;
+ if ( (c & 0xC0) != 0x80 )
+ return wxCONV_FAILED;
+
+ code <<= 6;
+ code |= c & 0x3F;
+ }
}
#ifdef WC_UTF16
return wxCONV_FAILED;
}
- // if we were really converting, check if we succeeded
- if ( buf )
+ // we did something, check if we really succeeded
+ if ( flags )
+ {
+ // check if the conversion failed, i.e. if any replacements
+ // were done
+ if ( usedDef )
+ return wxCONV_FAILED;
+ }
+ else // we must resort to double tripping...
{
- if ( flags )
+ // first we need to ensure that we really have the MB data: this is
+ // not the case if we're called with NULL buffer, in which case we
+ // need to do the conversion yet again
+ wxCharBuffer bufDef;
+ if ( !buf )
{
- // check if the conversion failed, i.e. if any replacements
- // were done
- if ( usedDef )
+ bufDef = wxCharBuffer(len);
+ buf = bufDef.data();
+ if ( !::WideCharToMultiByte(m_CodePage, flags, pwz, -1,
+ buf, len, NULL, NULL) )
return wxCONV_FAILED;
}
- else // we must resort to double tripping...
+
+ wxWCharBuffer wcBuf(n);
+ if ( MB2WC(wcBuf.data(), buf, n) == wxCONV_FAILED ||
+ wcscmp(wcBuf, pwz) != 0 )
{
- wxWCharBuffer wcBuf(n);
- if ( MB2WC(wcBuf.data(), buf, n) == wxCONV_FAILED ||
- wcscmp(wcBuf, pwz) != 0 )
- {
- // we didn't obtain the same thing we started from, hence
- // the conversion was lossy and we consider that it failed
- return wxCONV_FAILED;
- }
+ // we didn't obtain the same thing we started from, hence
+ // the conversion was lossy and we consider that it failed
+ return wxCONV_FAILED;
}
}
projects / wxWidgets.git / blobdiff