Loop question

[Former Member]

hello,

I have a requirement to make records unique before updating a table. the scenario is as follows.

table - fields A to G ( A, B, C and D are primary )

The data provider changed so they cannot provide an entry for field D, meaning a combination of A B and C may not be unique.

Eg : A B C D

-

---

X01 032 202

X02 032 002

X02 032 002

X04 032 662

Records 2 and 3 are not unique, I need to make it unique by incrementing the value of D by 1 for each record before updating the table.

how do I loop within a combination of records in this case to update the field D ?

[Former Member]

Hi,

1 way is:

sort itab by a b c d

loop at itab into wa.

loop at itab where itab-a eq wa-a and itab-b eq wa-b and itab-c eq wa-c and itab-d eq wa-d.

add 1 to d.

modify itab.

exit.

endloop.

endloop.

Regards

Nicole

[Former Member]

Hello,

You can keep value of D as 1 for every entry. Also you need to check that a particular combination is exsisting already in the table. if yes then increment value of D.

eg:

sort itab by a b c.
loop at t_itab into f_itab.
at new a.
   f_itab-d = 1.
end at.
read table t_itab into f_itab1      "<-- different work area.
    with key a = f_itab-a  b = f_itab-b  c = f_itab-c binary search.
if sy-subrc is initial.
  add 1 to f_itab-d.
endif.
append f_itab to t_itab1.
clear f_itab.
delete t_itab index sy-tabix.

endloop.
" Now required unique entries are available in table t_itab1

Hope this helps you.

Regrads,

Sachinkumar Mehta.

[Former Member]

Hi,

Take one more internal table itab2.

move the contents of itab1 to itab2.

itab2[] = itab1[].

now go for LOOP.

data v_count type i.

loop at itab1.

clear v_count.

loop at itab2 where a = itab1-a and b = itab1-b and c = itab1-c and d = itab1-d.

v_count = v_count + 1.

itab2-d = v_count.

modify itab2.

endloop.

endloop.

Finally itab2 contains unique records.

Regards,

Kumar Bandanadham

[Former Member]

Hi Suker,

Please check the test program to apply the logic before updating the table.

DATA:
       l_itab TYPE TABLE OF ztest_str11,
       l_str  TYPE ztest_str11, " Structure with field1, field2...
       l_str_temp TYPE ztest_str11,
       l_count  TYPE i VALUE 1.

CLEAR l_str.
l_str-field1  = 12.
l_str-field2  = 11.
l_str-field3  = 1.
l_str-field4  = ''.
l_str-field5  = ''.
APPEND l_str TO l_itab.

CLEAR l_str.
l_str-field1  = 12.
l_str-field2  = 11.
l_str-field3  = 2.
l_str-field4  = ''.
l_str-field5  = ''.
APPEND l_str TO l_itab.

CLEAR l_str.
l_str-field1  = 12.
l_str-field2  = 11.
l_str-field3  = 1.
l_str-field4  = ''.
l_str-field5  = ''.
APPEND l_str TO l_itab.

CLEAR l_str.
l_str-field1  = 12.
l_str-field2  = 11.
l_str-field3  = 1.
l_str-field4  = ''.
l_str-field5  = ''.
APPEND l_str TO l_itab.

CLEAR l_str.
l_str-field1  = 12.
l_str-field2  = 11.
l_str-field3  = 2.
l_str-field4  = ''.
l_str-field5  = ''.
APPEND l_str TO l_itab.

CLEAR l_str.

SORT l_itab BY field1 field2 field3 field4.

LOOP AT l_itab INTO l_str.
  IF l_str_temp <> l_str.
    l_count = 1.
    l_str_temp = l_str.
    l_str-field4 = l_count.
    MODIFY l_itab FROM l_str.
  ELSE.
    l_count = l_count + 1.
    l_str-field4 = l_count.
    MODIFY l_itab FROM l_str.
  ENDIF.
  CLEAR l_str.
ENDLOOP.

After your internal table has the correct data, then update the entries in the table.

Regards,

George

[Former Member]

As long as A, B and C are not numerical fields and field D is, create an additional internal table with the same definition.

LOOP AT your_table INTO wa_your_table.

     wa_your_table-D = 1.
     COLLECT wa_your_table INTO a_new_table.

ENDLOOP.

As long as A, B and C are the same, the value of D will be incremented with 1 by the COLLECT statement.