Amebis/wxWidgets
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

optimized wxMBConvStringUTF8::ToWchar() for ASCII characters

Browse Source 
git-svn-id: https://svn.wxwidgets.org/svn/wx/wxWidgets/trunk@48427 c3d73ce0-8a6f-49c7-b76d-6d57e0e08775
This commit is contained in:
Václav Slavík
2007-08-28 10:40:40 +00:00
parent c59a09cfea
commit 5367a38ad3
1 changed files with 61 additions and 46 deletions
Download Patch File Download Diff File Expand all files Collapse all files

39
src/common/strconv.cpp
View File

@@ -778,7 +778,24 @@ wxMBConvStrictUTF8::ToWChar(wchar_t *dst, size_t dstLen,
return written;
}
if ( out && !dstLen-- )
break;
wxUint32 code;
unsigned char c = *p;
if ( c < 0x80 )
{
if ( srcLen == 0 ) // the test works for wxNO_LEN too
break;
if ( srcLen != wxNO_LEN )
srcLen--;
code = c;
}
else
{
unsigned len = tableUtf8Lengths[c];
if ( !len )
break;
@@ -789,23 +806,20 @@ wxMBConvStrictUTF8::ToWChar(wchar_t *dst, size_t dstLen,
if ( srcLen != wxNO_LEN )
srcLen -= len;
if ( out && !dstLen-- )
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)
// 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)
// mask to extract lead byte's value ('x' bits above), by sequence length:
// mask to extract lead byte's value ('x' bits above), by sequence
// length:
static const unsigned char leadValueMask[] = { 0x7F, 0x1F, 0x0F, 0x07 };
// mask and value of lead byte's most significant bits, by length:
@@ -818,10 +832,10 @@ wxMBConvStrictUTF8::ToWChar(wchar_t *dst, size_t dstLen,
if ( (c & leadMarkerMask[len]) != leadMarkerVal[len] )
break;
wxUint32 code = c & leadValueMask[len];
code = c & leadValueMask[len];
// all remaining bytes, if any, are handled in the same way regardless of
// sequence's length:
// all remaining bytes, if any, are handled in the same way
// regardless of sequence's length:
for ( ; len; --len )
{
c = *++p;
@@ -831,6 +845,7 @@ wxMBConvStrictUTF8::ToWChar(wchar_t *dst, size_t dstLen,
code <<= 6;
code |= c & 0x3F;
}
}
#ifdef WC_UTF16
// cast is ok because wchar_t == wxUint16 if WC_UTF16