Hi Dimitrij
Your logic sounds good
Step1-You may build two internal tables, A and B lets say
Step 2- loop and compare the two based on a primary key and build the diffrences table call it C.
These tables can be of same type
You can use hash table to have unıque or you can have a normal standard table
sort the internal table and do a binary search will give you fast access.
loop at a
read table b with key x = a.x.
if sy-subrc eq 0.
else.
append initial line to C
endif.
endloop.