Have a look at this port. It shows a set of cases of 'a'='A' that might contain all the extended character sets that you are looking for. It might not, but it might give you some ideas.

Using the ideas in that post, I wrote a test script:

German.kix

code:

---

BREAK ON
CLS

$=setascii("off")
$TextK = "abcdefghijklmnopqrstuvwxyzäöüß"
$TextG = "ABCDEFGHIJKLMNOPQRSTUVWXYZÄÖÜ"

"source  kleine" $TextK ?
"enlarge kleine" UCASE($TextK) ? ?
"source grosse " $TextG ?
"shrink grosse " LCASE($TextG) ? ?

"little:" ?
chr(228) " " 256+asc("ä") ?
chr(246) " " 256+asc("ö") ?
chr(252) " " 256+asc("ü") ? ?

"big" ?
chr(196) " " 256+asc("Ä") ?
chr(214) " " 256+asc("Ö") ?
chr(220) " " 256+asc("Ü") ? ?

"sz" ?
chr(223) " " 256+asc("ß") ? ?

"subs" ?
"A-a=" 228-196 ?
"O-o=" 246-214 ?
"U-u=" 252-220 ? ?

"comps" ?
if chr(228)=chr(196) "A=a" else "A<>a" endif ?
if chr(246)=chr(214) "O=o" else "O<>o" endif ?
if chr(252)=chr(220) "U=u" else "U<>u" endif ? ?

---

This shows that KiX (3.63 and 2001 beta 2a) believe that Ä=ä and the difference of 32 corresponds with my discoveries in that aforementioned post. I hope I have helped in some way...

cj

p.s. Kix2001 beta can distinguish between Ä and ä with == instead of =

------------------
For more scripts goto cj's scripts page

chrismat@ozemail.com.au