Extract data from a string

[Former Member]

Hi,

I need to <b>EXTRACT</b> YYYY in the string <u><</u>XX<u>></u>YYYY<u><u><u><\></u></u></u>

However, only the charachters underlined are constant, as in... <\> will be present at the end of the string, but the first <> can have varying data between it.

string examples:

<u><</u>PQRST<u>></u> YYYY <u><\></u>

<u><</u>ABC<u>></u> XXXXYYYY <u><\></u>

I need to extract the 'YYYY' or 'XXXXYYYY' or whatever else that comes after the closure of the first bracket '<*>' and the start of the second bracket '<\>'.

I tried sy-fdpos, but it only gives the START of the <*> part. What really helps me is the CLOSE of the first <> and START of the second <>.

Thanks in advance.

[Former Member]

Hi,

You can split at each ">" and delete the 2 last caracters of the second data.

example :

<test>YYYY<\>

first data : <test

second data : YYYY<\ -> YYYY

[Former Member]

you can use split , that the only way to solve this

[Former Member]

You can also use this:

while lv_string(1) ne '>'.

shift lv_string.

endwhile.

shift lv_string.

if lv_string CA '<'.

lv_result = lv_string(sy-fdpos).

endif.

lv_result will contain the values between the first > and the second <.

Regards,

Michael

[Former Member]

Thanks a bunch for your replies mates,

Absolute peach of a logic there by Michael, was best suited for my smartform. That's the one I used.

All rewarded

[Former Member]

Hi

<XX>YYYY<\>, as you told in this string </> will be in the last.

data: a type i,

str = <XX>YYYY<\> ." (string)

a = strlen( str ).

a = a - 7. " as last is </> -

---

3 chars plus YYYY --- 4 chars. so 7 chars

shift str left by a places.

if it doesn't work,

do a times.

shift str left.

enddo.

now you will get str = YYYY</>.

now ,

str = str(4). " finally you will get str = YYYY.

[rainer_hbenthal]

Regular expressions are the best way of doing that:

REPORT  zregex.

DATA:
  tmp(20) TYPE c,
  pattern(50) TYPE c,
  s1(10) TYPE c,
  s2(10) TYPE c,
  s3(10) TYPE c,
  s4(10) TYPE c,
  s5(10) TYPE c,
  n      TYPE i,
  it_result TYPE match_result_tab.


tmp = '<ABC> XXXXYYYY <>'.
pattern = '.*(<.*>)(.*)(<>).*'.


FIND REGEX pattern
  IN tmp
  MATCH COUNT n
  RESULTS it_result
  SUBMATCHES s1 s2 s3 s4 s5.

write:/ s1, s2, s3.

Message was edited by:

Rainer Hübenthal