All three file formats allow multiple keys to access data
records and duplicate key values for specified keys. You can access
records by various keys using constructs within the programming
language. You can also use KSAM intrinsics to access records in
various sequences.
Record retrieval can be by direct match of specific key value,
by generic (or partial) key value, or by approximate match. Access
of data records by physical record location may or may not match
the primary key sequence, depending upon the order in which records
were initially loaded.