special characters

[Anuz]

Hi All,

I have a situation where in strings can have special characters i.e. names with Umlaut's.

For example :

Is there a way for example :CREACIONES PAZ RODRíGUEZ S L replaced with CREACIONES PAZ RODRA­GUEZ S L ?

But it can any character with umlaut's.

data aaa;
ORGNAME='PAZ RODRíGUEZ S L';
Run;

[s_lassen]

It may depend on your character set settings, but I had some problems with your example The obvious solution seems to be the good old TRANSLATE function, but that does not work, as 'Ã' seems to register as more than one character in my current SAS session:

 70         data _null_;
 71           set aaa;
 72           newname=translate(orgname,'A','Ã');
 73           put _all_;
 74         run;
 
 ORGNAME=PAZ RODRíGUEZ S L newname=PAZ RODRA ­GUEZ S L _ERROR_=0 _N_=1

The TRANWRD function works, on the other hand:

 70         data _null_;
 71           set aaa;
 72           newname=tranwrd(orgname,'Ã','A');
 73           put _all_;
 74         run;
 
 ORGNAME=PAZ RODRíGUEZ S L newname=PAZ RODRA­GUEZ S L _ERROR_=0 _N_=1

- only problem with that is that you may have to use several calls to get rid of different "strange" characters.

KTRANSLATE works better, as it can translate double-byte to single-byte characters (or the other way round):

 70         data _null_;
 71           ORGNAME='PAZ RODRíGUEZ S Lã';
 72           newname=ktranslate(orgname,'Aa','Ãã');
 73           put _all_;
 74         run;
 
 ORGNAME=PAZ RODRíGUEZ S Lã newname=PAZ RODRA­GUEZ S La _ERROR_=0 _N_=1

[ErikLund_Jensen]

Hi @Anuz 

No, I don't think SAS is able to recognise the "basic" character, but you can build your own translation table, like this:

108
109  data aaa;
110    ORGNAME='PAZ RODRíGUEZ S L';
111    NEWNAME = translate(ORGNAME,'A','Ã');
112    put ORGNAME=;
113    put NEWNAME=;
114  Run;

ORGNAME=PAZ RODRíGUEZ S L
NEWNAME=PAZ RODRA­GUEZ S L

You can supply a whole list of to-characters and from-characters in the same translate, see the. manual.

[PaigeMiller · Accepted solution]

You can use the BASECHAR function.

[Anuz]

@PaigeMiller  That resolved the issue. that is exactly what i was after. Thank you