Like this ? REGEXP_REPLACE(str,'[^[a-z,A-Z]]*',' ')

Hi,

user5116754 wrote:
How to replace all this German letters like (ü,ö,ä...)


select regexp_replace('Jürgen könnte übermorgen den Text ändern.','[ÖÜÄöüäß]', '(oeaeuess') NEW_TEXT from dual;

Of course this doesn't work...

All suggestions are welcome.

I think you'll need nested operations to do that. For example:

SELECT      TRANSLATE ( REGEXP_REPLACE ( REPLACE ( 'Jürgen könnte übermorgen den Text ändern.'
                              , 'ß'
                              , 'ss'
                              )
                       , '([ÖÜÄöüä])'
                       , '\1e'
                       )
            , 'ÖÜÄöüä'
            , 'OUAoua'
            )      AS new_text 
FROM     dual;

user5116754 wrote:
Yesssss, it works. Thank you Frank.
...but a simple regexp_replace - statement can't achieve the same???

Sorry, I don't think so. Maybe you could use a single statement to replace all the umlauts, since there is a common pattern to all of them, but the ess-zed would still need a separate operation.
You really want something like TRANSLATE that works on strings, not single characters, but, unfortunately, that doesn't exist. See the following thread for another work-around:
SQL Query

The solution I posted always replace the umlaut with a lower-case 'e'. If you want to detect all-upper-case words, and change 'ÜBER' into 'UEBER' (rather than 'UeBER'), then use PL/SQL.