Re: [bug-grep] Re: grep: -i option not working i cronjobs

[Elliott Hughes]

On Nov 15, 2004, at 06:27, Tim Waugh wrote:

> On Mon, Nov 15, 2004 at 12:56:38PM +0000, Julian Foad wrote:
>
> > Hold on.  Do you think converting to same-case would achieve that
> > one-versus-two-character match?  I don't.
>
> Hmm.
>
> Does the locale interface actually support multichar transliteration?
> In other words, does the de_DE locale data have anything to say about
> this special SS character?
>
> For me at least, in de_DE.UTF-8, "\303\237" is both upper-case and
> lower-case (towlower() and towupper() give the same result).

that's why Java has String.to(Lower|Upper)Case to be used in preference 
to the equivalent methods in Character. but, as i said, Java's regular 
expression code doesn't handle this case by itself.

even String.equalsIgnoreCase works character-by-character. (and yet 
goes to the trouble of comparing the characters, the upper-cased 
characters, and finally the lower-cased characters.)

$ cat > t.java
public class t {
 public static void main(String[] args) {
  
System.out.println("schlie\u00dflich".equalsIgnoreCase("SCHIESSLICH"));
 }
}
$ javac t.java && java t
false
$

i don't have my copy of the Rechtschreibung to see if it says anything 
about upper/lower equivalence and if so, what it says.

--
http://www.jessies.org/~enh/