kannonier8/blender
1
0
Fork 0
forked from bartvdbraak/blender
Code Pull Requests Activity

utf8 OSX - cleanup

Browse Source 
I still think utf8_buf can be 5 (4 bytes + '\0'), but even 6 may not be enough to what is coming next (NFC - precomposedStringWithCanonicalMapping)
incorporating ascii as a subset of utf8. I don't think we need to re-encode it.
U+0000 ~ U+00FF - latin1 set
This commit is contained in:
Dalai Felinto 2011-10-22 09:28:10 +00:00
parent 8f4a015684
commit cac4fde224
1 changed files with 10 additions and 33 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

43
intern/ghost/intern/GHOST_SystemCocoa.mm
View File

@ -1655,15 +1655,8 @@ GHOST_TSuccess GHOST_SystemCocoa::handleKeyEvent(void *eventPtr)
return GHOST_kFailure;
}
/* unicode input - not entirely supported yet
* but we are getting the right byte, Blender is not drawing it though
* also some languages may need special treatment:
- Japanese: romanji is used as input, and every 2 letters OSX converts the text
to Hiragana/Katakana.
- Korean: one add one letter at a time, and then the OSX join them in the equivalent
combined letter.
*/
char utf8_buf[6]= {'\0'};
ascii = 0;
switch ([event type]) {
@ -1678,32 +1671,18 @@ GHOST_TSuccess GHOST_SystemCocoa::handleKeyEvent(void *eventPtr)
keyCode = convertKey([event keyCode],0,
[event type] == NSKeyDown?kUCKeyActionDown:kUCKeyActionUp);
/* ascii */
/* handling both unicode or ascii */
characters = [event characters];
if ([characters length]>0) { //Check for dead keys
//Convert characters to iso latin 1 encoding
convertedCharacters = [characters dataUsingEncoding:NSISOLatin1StringEncoding];
if ([convertedCharacters length]>0)
ascii =((char*)[convertedCharacters bytes])[0];
else
ascii = 0; //Character not available in iso latin 1 encoding
}
else
ascii= 0;
/* unicode */
if ([characters length]>0) {
convertedCharacters = [characters dataUsingEncoding:NSUTF8StringEncoding];
if ([convertedCharacters length]>0) {
utf8_buf[0] = ((char*)[convertedCharacters bytes])[0];
utf8_buf[1] = ((char*)[convertedCharacters bytes])[1];
utf8_buf[2] = ((char*)[convertedCharacters bytes])[2];
utf8_buf[3] = ((char*)[convertedCharacters bytes])[3];
utf8_buf[4] = ((char*)[convertedCharacters bytes])[4];
utf8_buf[5] = ((char*)[convertedCharacters bytes])[5];
for (int x = 0; x < [convertedCharacters length]; x++) {
utf8_buf[x] = ((char*)[convertedCharacters bytes])[x];
}
else {
utf8_buf[0] = '\0';
/* ascii is a subset of unicode */
if ([convertedCharacters length] == 1) {
ascii = utf8_buf[0];
}
}
@ -1714,9 +1693,7 @@ GHOST_TSuccess GHOST_SystemCocoa::handleKeyEvent(void *eventPtr)
pushEvent( new GHOST_EventKey([event timestamp]*1000, GHOST_kEventKeyDown, window, keyCode, ascii, utf8_buf) );
//printf("Key down rawCode=0x%x charsIgnoringModifiers=%c keyCode=%u ascii=%i %c utf8=%s\n",[event keyCode],[charsIgnoringModifiers length]>0?[charsIgnoringModifiers characterAtIndex:0]:' ',keyCode,ascii,ascii, utf8_buf);
} else {
// XXX Font Object bug - backspace or adding new chars are being computed twice (keydown and keyup)
utf8_buf[0] = '\0';
pushEvent( new GHOST_EventKey([event timestamp]*1000, GHOST_kEventKeyUp, window, keyCode, ascii, utf8_buf) );
pushEvent( new GHOST_EventKey([event timestamp]*1000, GHOST_kEventKeyUp, window, keyCode, 0, '\0') );
//printf("Key up rawCode=0x%x charsIgnoringModifiers=%c keyCode=%u ascii=%i %c utf8=%s\n",[event keyCode],[charsIgnoringModifiers length]>0?[charsIgnoringModifiers characterAtIndex:0]:' ',keyCode,ascii,ascii, utf8_buf);
}
break;