How to avoid Delete after select error in Code Inspector

[svasi_08]

I have written below logic, its showing as error in SCI as "DELETE statement for result of SELECT statement found" ,

how can I fix this ?

  IF lt_inact_data[] IS NOT INITIAL.
    SELECT * FROM zpaheadcount INTO TABLE @DATA(lt_hcount)
    FOR ALL ENTRIES IN @lt_inact_data
    WHERE kostl = @lt_inact_data-cc(4).
  ENDIF.

  SORT lt_hcount BY kostl.

  DELETE ADJACENT DUPLICATES FROM lt_hcount COMPARING kostl.

What should be written to avoid this Code insepector error? I mean how can I replicated delete under

[Sandra_Rossi]

When I see your code, I wonder what lt_hcount contains if lt_inact_data is initial. I guess Code Inspector has the same interrogation...

[Szczerbowski]

Hi,

With the code pushdown, you should probably force the Sort/delete on the database.
So select distinct and order by in the SQL rather than ABAP.

And if your code needs to stay like this, use a pragma dedicated to hiding this error (usually the ATC error description mentions which pragma hides the error).

Regards,
Michal

[svasi_08]

Hi Sandra,

"When I see your code, I wonder what lt_hcount contains if lt_inact_data is initial. I guess Code Inspector has the same interrogation..." 

Dont think thats the issue. Good point any way I have moved code inside the if else condition, but still error exists.

Hi Michal

Hi,
With the code pushdown, you should probably force the Sort/delete on the database.
So select distinct and order by in the SQL rather than ABAP.

And if your code needs to stay like this, use a pragma dedicated to hiding this error (usually the ATC error description mentions which pragma hides the error).

I am not sure how to get only specific data based on the order when I am using distinct and order by, any code help is highly appreiacted.

and there is no Pragma for this to hide, thank you.

[svasi_08]

Lifting for answers

[Sandra_Rossi]

Don't you think this would be better:

  IF lt_inact_data[] IS NOT INITIAL.
    SELECT * FROM zpaheadcount INTO TABLE @DATA(lt_hcount)
      FOR ALL ENTRIES IN @lt_inact_data
      WHERE kostl = @lt_inact_data-cc(4).
    IF sy-subrc = 0.
      SORT lt_hcount BY kostl.
      DELETE ADJACENT DUPLICATES FROM lt_hcount COMPARING kostl.
    ENDIF.
  ENDIF.

[svasi_08]

Hi Sandra,

Looks better the code but not helping to erase the code inspector error.

[svasi_08]

Looking for some answers, thank you.

[FredericGirod]

it would be helpful to have the exact structure of this db table.

maybe you could play with subquery

SELECT *  
       FROM zpaheadcount
       WHERE [the_key] EQ ( SELECT [the_key] 
                                    FROM zpaheadcount
                                    FOR ALL ENTRIES IN ...
                                    WHERE kostl = ....  )
      ORDER BY kostl
      INTO TABLE @data(something_with_real_name).
                                    
 

[TheGokke]

select distinct?

[MateuszAdamus]

FOR ALL ENTRIES already forces a SELECT DISTINCT.

Kind regards,
Mateusz

[FredericGirod]

FOR ALL ENTRIES forces distinct based on the fields list, as there is a SELECT * there is no distinct

[Szczerbowski]

Well, as other people indicated, FOR ALL already implies distinct, so...

Do you really need all them columns, or just a specific one and cost center for key?
Why do you have duplicates in that table for CC, and why don't you care which line you get?

Then maybe group by would help, but then you need to change that for all into ranges IN maybe?

Regards,
Michał