24-May-2013 18:38, Manu пишет:
On 24 May 2013 19:49, Jacob Carlborg <d...@me.com <mailto:d...@me.com>>
wrote:
On 2013-05-23 23:42, Joseph Rushton Wakeling wrote:
I'm also in agreement with Manu. There may well already be bugs
for some of
them -- e.g. there is one for toUpperInPlace which he referred
to, and the
source of the allocation is clear and is even responsible for
other bugs:
http://d.puremagic.com/issues/__show_bug.cgi?id=9629
<http://d.puremagic.com/issues/show_bug.cgi?id=9629>
toUpper/lower cannot be made in place if it should handle all
Unicode. Some characters will change their length when convert
to/from uppercase. Examples of these are the German double S and
some Turkish I.
ß and SS are both actually 2 bytes, so it works in UTF-8 at least! ;)
Okay, here you go - an UTF-8 table of cased sin :)
Codepoint - upper-case - lower-case
0x01e9e : 0x000df - 3 : 2
0x0023a : 0x02c65 - 2 : 3
0x0023e : 0x02c66 - 2 : 3
0x02c7e : 0x0023f - 3 : 2
0x02c7f : 0x00240 - 3 : 2
0x02c6f : 0x00250 - 3 : 2
0x02c6d : 0x00251 - 3 : 2
0x02c70 : 0x00252 - 3 : 2
0x0a78d : 0x00265 - 3 : 2
0x0a7aa : 0x00266 - 3 : 2
0x02c62 : 0x0026b - 3 : 2
0x02c6e : 0x00271 - 3 : 2
0x02c64 : 0x0027d - 3 : 2
0x01e9e : 0x000df - 3 : 2
0x02c62 : 0x0026b - 3 : 2
0x02c64 : 0x0027d - 3 : 2
0x0023a : 0x02c65 - 2 : 3
0x0023e : 0x02c66 - 2 : 3
0x02c6d : 0x00251 - 3 : 2
0x02c6e : 0x00271 - 3 : 2
0x02c6f : 0x00250 - 3 : 2
0x02c70 : 0x00252 - 3 : 2
0x02c7e : 0x0023f - 3 : 2
0x02c7f : 0x00240 - 3 : 2
0x0a78d : 0x00265 - 3 : 2
0x0a7aa : 0x00266 - 3 : 2
And this is only with 1:1 mapping.
Generated by:
void main(){
import std.uni, std.utf, std.stdio;
char buf[4];
foreach(dchar ch; unicode.Cased_Letter.byCodepoint){
dchar upper = toUpper(ch);
dchar lower = toLower(ch);
int uLen = encode(buf, upper);
int lLen = encode(buf, lower);
if(uLen != lLen)
writefln("0x%05x : 0x%05x - %d : %d", upper, lower, uLen,
lLen);
}
}
--
Dmitry Olshansky
