Until now, in an NUP the data is copied to field wa or to table line itab as defined by the structure of the table work area dbtab without taking its structure into consideration. Only the length and alignment are checked.

SELECT * FROM dbtab ... INTO wa ...
SELECT * FROM dbtab ... INTO TABLE itab  ...
SELECT * FROM dbtab ... APPENDING TABLE itab ...

FETCH NEXT CURSOR c ... INTO wa.
FETCH NEXT CURSOR c ... INTO TABLE itab.
FETCH NEXT CURSOR c ... APPENDING TABLE itab.

INSERT INTO dbtab ... FROM wa.
INSERT dbtab ... FROM wa.
INSERT dbtab ... FROM TABLE itab.

UPDATE dbtab ... FROM wa.
UPDATE dbtab ... FROM TABLE itab.

MODIFY dbtab ... FROM wa.
MODIFY dbtab ... FROM TABLE itab.

DELETE dbtab FROM wa.
DELETE dbtab FROM TABLE itab.

The following rules are now valid in a UP:

• If the work area or the line of the internal table is a structure, there is also a check if the fragment views of the work area and the database table are the same up to the length of the database table. If the work area is a single field, the field must be have character-type and the database table must have character-type only. These requirements are valid for all the commands mentioned above.

Only the types C, N, D, T, – and flat structures of these types – are now valid for the version field in any statement that processes database tables (READ, MODIFY, DELETE, LOOP) and uses the VERSION addition. Otherwise, a warning is triggered in an NUS, and a syntax error in a US.