Dear experts, Dear Jasmin,
recently we are working in great planning project with IP and PAK. We are examining the possibilities of PAK and opposing them to 'traditional' IP functionality.
Since in classic IP context we are working with generic types by design, we wondered how to achieve this with PAK.
OK, so far PAK forced us to create types corresponding to our aggregation levels (C_TH_DATA) as paramteres i_view and i_ref_data.
In IP we are widely working with generic types or even references to generic field-symbols to modularize our coding. Sometimes we need to create types during run-time depending on planned values within c_th_data, create variables of this types, pass them to other methods or function modules and populate them.
In IP this is no problem, as the wohle of its abap oo framework is a ripe product ;-)
Regarding PAK we can see no option, how to achieve this.
We read in Rich Heilman's Blog the planned functionality of table type definitions for parameters.
This looks like a step into this direction.
1) Are there any plans to provide something like the passing of untyped, generic paramteres? Or do we really have to renounce this beautiful functionality of ABAP?
2) In the context of PAK we cannot jump back to the abap stack and proceed our logic with the result passed from the result image of the SQLScript. Does this really mean, that if we have to do this, we cannot use PAK but IP instead (and also cannot activate deep HANA integration for this InfoProviders)? For right now, we are playing aoround with ADBC and AMDPs to achieve this, but this really would mean not to use PAK.
Any helpful discussion is very welcomed!
THX,
Jonas