RE: Question using jdbc/jt400

From: MidRange

I am working on an application using jt400 jdbc to access a db on
the i5. The application moves backward and forward through a couple
of files (in lockstep) randomly, based on user input. What is the
best way to accomplish this? I can just create a statement and
result set with each iteration, but it seems there ought to be a way
to use a cursor or something to traverse the tables more
economically without opening and closing result sets constantly.
However, I cannot find anything that allows me to do that. I am
only an egg.;) Can someone enlighten me? Thanks!

Hi Rick! The toolbox support two distinct styles of file access. The first
is the standard SQL-oriented access of JDBC, which allows you to create a
cursor and then move through it, but repositioning randomly (especially
based on key) is not particularly easy. You can't, for example, create a
result set starting at a given key value and then position yourself
BACKWARDS one record (along the lines of a SETGT, READP).

The second technique is called Record Level Access, or RLA, and it basically
provides standard access to your files, by RRN for non-keyed files and by
key for keyed files. You can open a logical view and then position yourself
by key anywhere in the file exactly like a SETLL (well almost exactly; there
is a subtle difference on where the cursor gets positioned, but for most
purposes it works just the same).

I wrote a course comparing the two a while back, but in general it's pretty
straightforward. You can find documentation on the standard JDBC stuff just
about anywhere; it's pretty much industry standard. IBM has documentation
on com.ibm.as400.access.KeyedFile and com.ibm.as400.access.SequentialFile
for the RLA stuff.

Joe