topic Re: Sdata field cannot be used where condition in Application Development and Automation Discussions

Sdata field cannot be used where condition

former_member760452 — Mon, 03 Jan 2022 06:24:48 GMT

SELECT *
    FROM edid4
    INTO CORRESPONDING FIELDS OF TABLE lt_edid4
    FOR ALL ENTRIES IN lt_docnum
    WHERE docnum EQ lt_docnum-docnum
       AND segnum EQ '000001'
       AND segnam IN lr_segnam
       AND hlevel EQ '01'
       AND sdata  IN lr_objkey2.

Hi guys i am getting this error can you help me?

The field "SDATA" is a long string or a text and cannot be used in the WHERE condition.

Re: Sdata field cannot be used where condition

matt — Mon, 03 Jan 2022 07:06:05 GMT

You'll need to select without the sdata clause, then loop through lt_edid4 and eliminate records not in lr_objkey2.

Re: Sdata field cannot be used where condition

Sandra_Rossi — Mon, 03 Jan 2022 07:50:38 GMT

In Open SQL, it's not possible directly (see Matthew answer).

Note that the transaction code WE09 permits to find IDocs based on SDATA.

You may also try with native SQL. Below code that you shouldn't use productively, it's just to quickly test if native SQL supports it, if yes you may use AMDP, ADBC or HANA SQL Script (don't know for HANA CDS).

DATA edid4 TYPE edid4.
EXEC SQL PERFORMING select_row_from_edid4.
  SELECT DOCNUM, SDATA FROM EDID4 WHERE SUBSTRING( SDATA, 1, 1 ) = 'A' INTO :EDID4-DOCNUM,:EDID4-SDATA
ENDEXEC.
FORM select_row_from_edid4.
  WRITE: / edid4-docnum, edid4-sdata.
ENDFORM.

Re: Sdata field cannot be used where condition

former_member760452 — Mon, 03 Jan 2022 08:13:52 GMT

SELECT *
  FROM edid4
  INTO CORRESPONDING FIELDS OF TABLE lt_edid4
  FOR ALL ENTRIES IN lt_docnum
  WHERE docnum EQ lt_docnum-docnum
    AND segnum EQ '000001'
    AND segnam IN lr_segnam
    AND hlevel EQ '01'.
*              AND sdata  IN lr_objkey2.
  LOOP AT lt_edid4 INTO DATA(ls_edid).
    DELETE lt_edid4 FROM lr_objkey2.
  ENDLOOP.

According to what you said, will this work for me?

Re: Sdata field cannot be used where condition

matt — Mon, 03 Jan 2022 09:29:27 GMT

Something like that yes.