UCase and German characters

[netuser]

Environment: CFMX- Oracle-Windows XP

I'm querying my database for records that matched the German character "?" (If
you dont see this character properly, open up your Windows Word and hold the
the "ctrl" and "shift" keys then hit " & " Let up on all keys, then type "s" .)

But before doing my search, I UCase the input string. Unfortunately UCase
converts it to "SS" and my query does not return the expected results..

Anyone seen this happening ?

[PaulH]

there is no upper case form of "sharp s" so cf is replacing it with ss (which i
think is a rule, i do know that capital case ? was proposed & rejected by the
unicode consortium). can't the oracle query be made case-insensitive?

[netuser]

Paul

Thanks for your response. I can make the query case-insensitive, but I have a
lot of code written which already use UCase. Dint want to impact existing code
just for 1 character.

If the character has been rejected by Unicode consortium, I wonder why Oracle
returns UPPER('?') as ?.

Do you know of any other characters that would have this problem ?

Thanks in Advance.

[PaulH]

oracle has always gone it's own way--there simply isn't an upper case ? and ss
is the rule for when you can't use ?--so in this case i guess oracle's wrong in
what it's doping (though this might be just a case of one of its many charsets
behavior).

as far as i recall that's the only char w/out an upper case form, of course
that's not counting languages w/out any kind of upper case like thai, etc. but
if i get a chance i'll poke around w/icu4j's UCharacter class to see what's
what. what langauges are you dealing with?

[PaulH]

i tested this using icu4j's UCharacter class (using no locale and de_DE locale)
and it returns the same thing as cf/core java, SS for upper case ?. i didn't
have time to run thru many enumerations of captial letters but that appears to
be the only one w/out an upper case form.

<cfscript>
uChar=createObject("java","com.ibm.icu.lang.UChara cter");
germanLocale=createObject("java","com.ibm.icu.util .ULocale").init("de_DE");
sharpS=chr(223);
writeoutput("<b>no locale:</b> #sharpS# #uChar.toUpperCase(sharpS)#
#uChar.toLowerCase(sharpS)#<br>");
writeoutput("<b>de_DE locale:</b> #sharpS#
#uChar.toUpperCase(germanLocale,sharpS)#
#uChar.toLowerCase(germanLocale,sharpS)#<br>");
</cfscript>