Archives of the TeradataForum

Message Posted: Tue, 19 Dec 2006 @ 11:39:08 GMT

	<Prev	Next>		<<First	<Prev	Next>	Last>>

<-- Anonymously Posted: Tuesday, December 19, 2006 00:58 -->

Many thanks for the reply and sorry for not putting it across properly.

The below is what I meant. The unicode Values for the characters are '0E20' and '0E40' ( My other question would be is there a teradata function to use to get the character given the unicode value? )

     CREATE SET TABLE test.rd1 (
           a CHAR(10) CHARACTER SET UNICODE NOT CASESPECIFIC,
           b CHAR(10) CHARACTER SET UNICODE NOT CASESPECIFIC)
     PRIMARY INDEX ( a );

     insert into RD1 values ('?', '?') -- the characters corresponding to
     unicode '0E20' and --'0E40'

     sel * from rd1 where a=b
     Answer :  ('?', '?')  -- the row just inserted is output


     CREATE SET TABLE test.rd2   (
           a CHAR(10) CHARACTER SET UNICODE CASESPECIFIC,
           b CHAR(10) CHARACTER SET UNICODE CASESPECIFIC)
     PRIMARY INDEX ( a );

     insert into RD2 values ('?', '?') -- the characters corresponding to
     unicode '0E20' and --'0E40'
     sel * from rd2 where a=b
     Answer :  No rows selected.

Teradata should ideally be returning no rows for the not casespecific case as well.One solution would be to use the function CASESPECIFIC in Selects. Other would be to make all columns casespecific by default, but 'not casespecific' seems an important business requirement.

Thanks for all your opinions

Regards

	<Prev	Next>		<<First	<Prev	Next>	Last>>