ABAP Regular Expression

[former_member813607]

Hi All,

Can someone help me with a regular expression to match and return a pattern from a string

The pattern is anything within parenthesis and three characters long!

Eg: (B6P)

That is I have to get the first matching from the string

Eg2.

From string 'ADFGDFG(4GH)ghjghj(HH6)FGHghfhgfGFHJ(DFGFG)GFHFG sdfg dfgdf

I need to get (4GH)

Eg3

From String 'SDFG(SD6GD)FGDFGDFGsdfgdfg(ghj)(fghgffh)gfhgfhgfDFGDFHG

I need to get (ghi)

Can someone help?. Any quick help will be appreciated heavily

Thanks, Sudeep..

[Paulo_Vantini]

Hi Sudeep,

for string 'ADFGDFG(4GH)ghjghj(HH6)FGHghfhgfGFHJ(DFGFG)GFHFG sdfg dfgdf

just create a string variable with name v_string and use string manipulation like:

DATA: v_string TYPE string,

            v_result TYPE string.

v_string = 'ADFGDFG(4GH)ghjghj(HH6)FGHghfhgfGFHJ(DFGFG)GFHFG sdfg dfgdf'.

v_result = v_string+7(5).

WRITE: v_result.

and for string 'SDFG(SD6GD)FGDFGDFGsdfgdfg(ghj)(fghgffh)gfhgfhgfDFGDFHG' do the same couting the characters and setting the length of the output.

Hope it solves your issue.

Regards.

[former_member813607]

Paulo,

I am looking for a generic solution,, not for the example

Regards, Sudeep..

[Paulo_Vantini]

Sudeep,

you can use that logic to build a subroutine for your problem like:

REPORT  Z_STRING_MANIPULATION.

PARAMETERS: p_string TYPE string,

                          p_count TYPE I,

                          p_length TYPE I.

PERFORM z_return_string USING p_count p_length CHANGING p_string.

WRITE: p_string.

FORM z_return_string USING p_count TYPE I

                                             p_length TYPE I

                                  CHANGING p_return TYPE string.

   p_return = p_return+p_count(p_length).

ENDFORM.

other solution ( if I got your picture ) is as follows:

REPORT  Z_STRING_MANIPULATION.

DATA: pos TYPE i,

       v_dif TYPE i,

       v_return TYPE string,

       pos_2 TYPE i.

PARAMETERS: p_text TYPE string DEFAULT 'SDFG(SD6GD)FGDFGDFGsdfgdfg(ghj)(fghgffh)gfhgfhgfDFGDFHG'.

DO.

SEARCH p_text FOR '('.

IF sy-subrc = 0.

   pos = sy-fdpos.

ENDIF.

SEARCH p_text FOR ')'.

IF sy-subrc = 0.

   pos_2 = sy-fdpos.

ENDIF.

v_dif = pos_2 - pos.

IF  v_dif = 4.

v_return = p_text+pos(5).

WRITE: v_return.

exit.

ELSE.

   pos_2 = pos_2 + 1.

   p_text = p_text+pos_2.

ENDIF.

ENDDO.

Regards.

[Former Member]

Hi Sudeep,

This should be match your requirment:

DATA: str TYPE string.

  str = 'SDFG(SD6GD)FGDFGDFGsdfgdfg(ghj)(fghgffh)gfhgfhgfDFGDFHG'.
  FIND FIRST OCCURRENCE OF REGEX
   '\(.{3}\)'
    IN str
    IGNORING CASE
    MATCH OFFSET moff
    MATCH LENGTH mlen.

  WRITE:str+moff(mlen).

Regards,

Archer

[former_member813607]

Archer

That's exactly what I was looking for brother, you solved my problem!!

Thanks, Sudeep..

[Former Member]

Regex may need to be modified of your string can have this type of data.

For input string 'a(bc))ef(ghi)' , first match would be (bc)) , not (ghi).

See this snippet. Between parenthesis, we need to ensure that there are no parenthesis.

DATA: str     TYPE string,

      result  TYPE string.

str = 'asdf(asdf)asdf(fds)abc'.

FIND REGEX '(\([^()]{3}\))' IN str SUBMATCHES str.

WRITE: sy-subrc, result.

str = 'a(bc))ef(ghi)'.

FIND REGEX '(\(.{3}\))' IN str SUBMATCHES result.

WRITE: sy-subrc, result.

[Former Member]

Yes, you are right. Anyway, this decide by what kind of data he have.