Aletrnatives for Nested Loop to same table

[Former Member]

Hi,

Is there any alternative for this logic below.....

Table1 consist of date field. Date field in a particular record is checked with all other previous record and then some logic is proccessed.

Normal way is this...

Loop at table1.

     l_index = sy-tabix.

     Loop at table2(copy of table1) untill l_index.

          some logic.

     Endloop.

Endloop.

Thanks....

[Former Member]

Hi Pradosh,

The above code is a good technique when performance is taken into consideration and u can proceed with it. Called Parallel cursor technique and it is an alternative for Nested Loops.

Could u plz tell if u faced any issue with this one ......

Thanks

Vivek

[Former Member]

This message was moderated.

[Former Member]

hi,

please find exact code as below

data:lv_index type sy-tabix.

Loop at table1.

     Loop at table2(copy of table1) from lv_index.

      if table1-field <> table2-field

         l_index = sy-tabix.

         exit.

     endif.

          some logic.

       Endloop.

Endloop.

[Former Member]

Hi Krishna,

Thanks for u r concern.........Your approach is correct but my scenario is different.

I need to compare the date field of the table2 with all other previous record's date field of table1

and put into another table(result, I need).

I need alternative other than Nested Loop.....

or else I have to implement this itself...

Regards

[former_member189845]

Hi Apta,

This is the More sophisticated loop in which parallel cursors .

For EX:

* Entries: 100 (ITAB1), 1000 (ITAB2)

* Line width: 100

* Both tables sorted by key K

I = 1.

LOOP AT ITAB1 INTO WA1.

  LOOP AT ITAB2 INTO WA2 FROM I.

    IF WA2-K <> WA1-K.

      I = SY-TABIX.

      EXIT.

    ENDIF.

    " ...

  ENDLOOP.

ENDLOOP.

this is most efficeint way of coding and measure runtime is only 183 micro seconds.

Thanks,

Siva

[uppu_narayan]

hi Pradosh,

      According to me your requirement doesn't require a parallel cursor technique.......the logic you have written is perfect to your requirement.....

thanks and regards,

narayan

[Former Member]

Hi Pradosh,

Suppose TABLE2 is copy of TABLE1 and you have 1 additional field in table2 called  ZINDEX of TYPE SY-TABIX.

Also assuming that number of data in both internal tables are same (since table2 is reflecting table1):

DATA: ITAB1 TYPE STANDARD TABLE OF TABLE1,

           ITAB2 TYPE STANDARD TABLE OF TABLE2,

           WA1 TYPE TABLE1,

           WA2 TYPE TABLE1.

DATA: INDEX_COUNT TYPE SY-TABIX.     

LOOP AT ITAB1 INTO WA1.

INDEX_COUNT = SY-TABIX.

WA2-ZINDEX = INDEX_COUNT.

MODIFY ITAB2 INDEX  INDEX_COUNT  FROM WA2 TRANSPORTING ZINDEX.

READ TABLE ITAB2 WITH KEY ZINDEX = INDEX_COUNT

IF SY-SUBRC = 0.

perform logic and changes in itab_2

ENDIF.

ENDLOOP.

*******************************

So when SY-TABIX of ITAB1 is 1, changes are made in ITAB2 at index 1, when SY-TABIX of ITAB2 is 2, logic for ITAB2 is executed at index 2 and so on.. And so logic will be applied on each line of ITAB2 till the current index.

(let me know if i am wrong).

Regards,

Khushboo.

[Former Member]

Thanks for your concern...

You need to understand my requirement....again

Your logic is perfect but it doesn't meet my requirement..

Regards

[Former Member]

hi

Please Check below link and find your solution

http://wiki.sdn.sap.com/wiki/display/Snippets/ABAP+Code+for+Parallel+Cursor+-+Loop+Processing

Thanks & Regards

RKarmakar