Simplification of import of serialized data for RFC usage
As it was shown in previous example it’s possible to supply any data into RFC enabled FORM. But, at the same time direct usage of IMPORT clause requires explicit specification of the ID’s in data buffer and also involves some manual work to define types. Fortunately there is a special class CL_ABAP_IMPEXP_UTILITIES that could handle all dirty work.
The only limitation I’ve found – it works pretty good with DDIC types, but for custom defined types it could fail without any detailed explanation.
This greatly simplifies extension of the methods and makes code much more compact and readable.
I’ve tested this example with 50000 lines and seems ABAP can handle such amount of data. Generally for the parallel processing this should be enough.
So, here is the sample code:
REPORT Z_BINARY_TRANSFORM2. data: x_b2 type xstring, lt_abc type STANDARD TABLE OF t000, wa_t000 type t000. INITIALIZATION. * Fill table with sample data wa_t000-mandt = sy-mandt. wa_t000-mtext = 'SCN Demo #1'. wa_t000-ort01 = 'Moscow'. append wa_t000 to lt_abc. wa_t000-mandt = sy-mandt. wa_t000-mtext = 'SCN Demo #2'. wa_t000-ort01 = 'Tokio'. append wa_t000 to lt_abc. * Export data into buffer export tadir = lt_abc to DATA BUFFER x_b2 COMPRESSION on. * Check the result perform describe_buffer using x_b2. form describe_buffer using in_buffer type xstring. data: lt_datatab type tab_cpar. FIELD-SYMBOLS: <fs> like line of lt_datatab, <fs2> type any table, <fs3> type any. lt_datatab = cl_abap_expimp_utilities=>dbuf_import_create_data( dbuf = in_buffer ). loop at lt_datatab assigning <fs>. write : <fs>-name. " Here is the name of the currently processed tab in buffer new-line. assign <fs>-dref->* to <fs2>. loop at <fs2> assigning <fs3>. write : <fs3>. new-line. endloop. endloop. endform.
Thanks Alexey for sharing this