Collect statement

[Former Member]

Hi,

I have 2 internal tables, the first table which has quantity, week(WKYYYY format) matnr werks.

2nd internal table also has the same structure , now i want a flag in 2nd internal table when collect statement happens.

Eg. 1223 abc 10 112010

1223 abc 20 122010

1224 xyz 10 122011

now my 2nd internal table should have a flag like this

1223 abc 30 122010 X

1224 xyz 10 122010 this should not have X set

like this...

Loop at internal table 1.

collect into internal table 2.

endloop.

[kesavadas_thekkillath]

Collect statement will not work for this scenario. You have to find some other logic like reading, adding and modifying.

[Former Member]

Hi

This cannot be done Simultaneously using Collect statement since the 2 table Structures are different and collect will not work.

[roberto_vacca2]

Hi...

That's something that could be useful for you....

DATA: BEGIN OF tab1 OCCURS 0,

key(4) TYPE c,

matnr TYPE mara-matnr,

qty TYPE i,

END OF tab1,

BEGIN OF tab2 OCCURS 0,

key(4) TYPE c,

matnr TYPE mara-matnr,

qty TYPE i,

flag TYPE c,

END OF tab2.

REFRESH: tab1, tab2.

tab1-key = '1223'.

tab1-matnr = 'TEST1'.

tab1-qty = 10.

APPEND tab1.

CLEAR tab1.

tab1-key = '1223'.

tab1-matnr = 'TEST1'.

tab1-qty = 20.

APPEND tab1.

CLEAR tab1.

tab1-key = '1224'.

tab1-matnr = 'TEST2'.

tab1-qty = 15.

APPEND tab1.

CLEAR tab1.

tab1-key = '1225'.

tab1-matnr = 'TEST3'.

tab1-qty = 34.

APPEND tab1.

CLEAR tab1.

tab1-key = '1225'.

tab1-matnr = 'TEST3'.

tab1-qty = 28.

APPEND tab1.

CLEAR tab1.

tab1-key = '1226'.

tab1-matnr = 'TEST4'.

tab1-qty = 2.

APPEND tab1.

DATA: currkey(4) TYPE c.

LOOP AT tab1.

CLEAR tab2.

READ TABLE tab2 WITH KEY currkey.

IF sy-subrc EQ 0 AND

tab1-key EQ currkey.

tab2-flag = 'X'.

MODIFY tab2 FROM tab2 INDEX sy-tabix.

ENDIF.

MOVE-CORRESPONDING tab1 TO tab2.

IF tab1-key NE currkey.

currkey = tab1-key.

CLEAR tab2-flag.

APPEND tab2.

ELSE.

COLLECT tab2.

ENDIF.

ENDLOOP.

[Former Member]

Hi,

Try this code

data : begin of it1,
         fld1,
         fld2(10),
         fld3(10),
         fld4 type i,
       end of it1.

data : begin of it2,
         fld1,
         fld2(10),
         fld3(10),
         fld4 type i,
         fld5,
       end of it1.
data : wa2 like it2.       
sort it1 by fld1 fld2 fld3.       
loop at it1.
  wa2 = it1.
  count = count + 1.
  at end of fld3.
    sum.
    wa2-fld4 = it1-fld4.
    if count gt 1.
      wa2-fld5 = 'X'.
    endif.
    clear count.
    append wa2 to it2.
    clear wa2.
  endat.
endloop.       

Thanks,

Anmol Bhat.

Edited by: anmol112 on Nov 16, 2010 3:48 PM

[Former Member]

Firstly, the first two rows will not be collected into one row because 4th field values are different.

The solution to your question is, instead of adding a flag column to the second itab, add a number column and set 1 for that column in the loop. After all the collection is complete, you should interpret any number greater than 1 as 'X'.

[Former Member]

Very elegant.

Rob

[Former Member]

Try this:

DESCRIBE TABLE a LINES index.

COLLECT wa2 INTO itab2.

DESCRIBE TABLE a LINES index1.

IF index = index1.

collect = 'X'.

ELSEIF index LT index1 AND collect = 'X'.

CLEAR collect.

wa2-flag = 'X'.

MODIFY itab2 FROM wa2 INDEX index TRANSPORTING flag.

CLEAR wa2-flag.

ENDIF.

[chandra_sekhar3]

Hi,

<< Moderator message - Cut and paste response from various sources removed. Plagiarism is not allowed in SCN >>

Edited by: chandrasek on Nov 19, 2010 4:20 PM

Edited by: Rob Burbank on Nov 19, 2010 10:41 AM

[Former Member]

>

> Hi,
> 
> COLLECT can create unique or compressed datasets and should be used precisely for this purpose. If uniqueness or 
> 
> compression are unimportant, or two values with identical default key field values could not possibly occur in your particular 
> 
> task, you should use APPEND instead. However, for a unique or compressed dataset which is also efficient, COLLECT is the 
> 
> statement to use.
> 
> Edited by: chandrasek on Nov 19, 2010 4:20 PM

ow... naughty naughty... i'm pretty sure you didn't write that piece yourself.