Re: Fw: Debugging - stopping at F-Spec

The debugger stops on every field in read from the file because of the way
RPG handles "native" IO. Records are not read directly into your fields.
They are read into a record buffer and then each field must be moved to
the field in your program--one at a time. (You can see the effects of
this more clearly if you debug in the *LISTING view.) The best way to
prevent this is to code the *NODEBUGIO option. You can also improve the
processing by specifying a data structure to use on your F-spec. Then RPG
will read directly into the data structure and there is only one step at
the f-spec. Someone wrote a good analysis on this, but I can't put my
finger on it right now.

wdsci-l-bounces@xxxxxxxxxxxx wrote on 06/08/2012 04:10:26 AM:

----- Message from Paul Bailey <PaulBailey@xxxxxxxxxx> on Fri, 8 Jun
2012 09:09:03 +0100 -----

To:

Rational Developer for IBM i / Websphere Development Studio Client
for System i & iSeries <wdsci-l@xxxxxxxxxxxx>

Subject:

Re: [WDSCI-L] Fw: Debugging - stopping at F-Spec


Karleen has the best option, IMO, because the additional *SrcStmt is
very useful for working out which source line a runtime error
appears on. You can specify the *nodebugio and *srcstmt in the
default compiler options as well.

But to add a little to everyone else's answers, the "indeterminate"
in this statement: "need to hit f6 an indeterminable number of
times" is easily explained by the number of fields on the file in
the F-spec - the debugger stops for every single field read from the
file. Don't ask me why, but it was always true until I started
preventing debugIO as well.


-Paul