ABAP 7.40: using function 'segment' instead of command 'split'

[ralf_wenzel_heuristika]

The following coding has to be translated into ABAP 7.40:

DATA:

  source    TYPE string VALUE 'MARA-MATNR=12345',

  tabname   TYPE tabname,

  fieldname TYPE fieldname,

  attribute TYPE string.

  SPLIT:

    source AT '-'

           INTO tabname fieldname,

    fieldname AT '='

              into fieldname attribute.

  " RESULT: tabname = 'MARA', fieldname = 'MATNR', attribute = '12345' !

I do not understand the F1 help of "segment" functions' features....

appendix:

via trial & error:

tabname    = segment( val = source index = 1 sep = '-' ).
fieldname  = segment( val = segment( val = source index = 2 sep = '-' ) index = 1 sep = '=' ).
attribute  = segment( val = source index = 2 sep = '=' ).

any shorter codings?

[Clemenss]

Hi Ralf,

the documentation ABAP Keyword Documentation looka quite clear, But I don't see any advantage over SPLIT in this context.

Your code will run in 740 as it is. Translation is 1 : 1.

I have no idea what the special purpose of segment function may be - don't eve like the sample code for determining min an max as given here ABAP Keyword Documentation.

Regards Clemens

[ralf_wenzel_heuristika]

I am just training 7.40.... So my example may offer no advantages. But it IS an advantage to append a part of a field to an internal table without using help variables, for instance. In my example:

data:

  tabnames type table of tabname. 

append to_upper( segment val = source index = 1 sep = '-' ) ) to tabnames.

instead of

data:

  tabnames type table of tab name,

  helpfield_1 type tabname,

  helpfield_2 type string.

split source at '-'

into helpfield_1  " only used as split target field

     helpfield_2. " helpfield_2 never used, just a dummy!

translate helpfield_1 to upper case.

append helpfield_1 to tabnames.