Interesting itab(*) syntax in a custom program

[mike_m1]

Hi fellow ABAP developers!

I have just spotted a bit of ABAP in a bespoke report program that has me flumoxed.

This is it (with some anonymization);

*---------------------------------------------------------------------------------------------------- 

TYPES:   tt_chg_ptrs_in   TYPE STANDARD TABLE OF bdcp,

                tt_ztable            TYPE SORTED TABLE     OF ztable WITH UNIQUE KEY matnr.

DATA:   l_it_chg_ptrs_in TYPE tt_chg_ptrs_in, 

               l_it_ztable          TYPE tt_ztable.

    IF     l_it_chg_ptrs_in(*) IS NOT INITIAL

     OR   l_it_ztable[]   IS NOT INITIAL.   

         .

         .

    ENDIF.

*----------------------------------------------------------------------------------------------------

The (*) appendage appears to behave in the same way as the [] variant when refering to the complete set of itab entries. I checked this out in debug just to be certain that the l_it_chg_ptrs_in(*) reference returned the same data as l_it_chg_ptrs_in[].

Is this a new (or old) lanuage element?

Who can shed some light on this for me?

Cheers all,

Mike

[RaymondGiuseppi]

Welcome to 7.40, this syntax is explained in table_exp - Table Expressions. (can be found from changes in Release 7.40 in ABAP - Release-Specific Changes.)

Regards,

Raymond

[mike_m1]

Thanks for the speedy reply Raymond,

This client is currently on 620.

Sadly, I couldn't see a reference to the itab(*) syntax in the table_exp - Table Expressions keyword documentation. 😞

Could you be more specific please? Where in this document would I find the explanation you alluded to?

Regards,

Mike

[RaymondGiuseppi]

Sorry, I did not look carefully enough, mistake [] for ()...

For this syntax I'm perplex too, seems to be allowed for structure, variable and internal tables, but not exactly (no redefinition of type) as var+0(*) or var+offset( *) if only character fields.

Regards,

Raymond

[mike_m1]

No worries Raymond.

I'd love to hear from anyone else who knows why this (*) syntax exists.

Mike

[Former Member]

Weird,

I just checked what this syntax does in de debugger, but it seems to be exactly the same as without (*)

I'm guessing this is used in other programming languages and the abap compiler just doesn't mind.

Kind regards, Rob Dielemans

[mike_m1]

Cheers Rob.

I appreciate your time.

[Former Member]

Hey Mike,

Try the coffee corner to speculate on why would SAP allow this syntax?

There are a couple of others, you can use >< instead of <> for instance.

Cheers!

[retired_member]

you can use >< instead of <> for instance

But only outside classes 

http://help.sap.com/abapdocu_740/en/index.htm?file=abenobsolete_logexp_op.htm

[ThomasZloch]

Maybe would know, let's see if he can be paged...

Thomas

[retired_member]

Well, well, a full text search for (*) in the ABAP documentation leads us to substring access.

• If off is smaller than the length of dobj, an asterisk (*) can be 
  specified for len. The upper limit of dobj then determines the upper limit of the memory area. 

And guess what, you can also write struct+0(*) or itab+0(*).

With other words, an Offset-/Length addition +0(*), +0, or (*) that denotes the full field seems to be simply ignored by the compiler. And that's why you can add it even to fields where a substring access is normally forbidden - an itab+1(*) gives a syntax error..

Cheers!

Horst

[mike_m1]

Thank you Horst. My curiosity is satisfied (for the moment, at least).

Mike