How to avoid NAT1310?

[Matthias1 (Matthias)]

The Natural-Documentation of ON ERROR says:

When a subroutine structure is built by using CALLNAT, PERFORM or FETCH RETURN, each module may contain an ON ERROR statement.
When an error occurs, Natural will automatically trace back the subroutine structure and select the first ON ERROR statement encountered in a subroutine for execution.

and the Documentation of RETRY says:

The RETRY statement must be placed in the object that causes the Error 3145.

That means: If a NAT3145 happens in a module with no ON ERROR-Block, Natural traces back all structure-levels. And if an ON-ERROR-Block is found in one of the calling programs, this one is taken for error handling. But there, I can’t do a RETRY, otherwise a NAT1310 occurs.

So my question is: How can I avoid this situation from the ON-ERROR’s point of view? Is it possible to code the following way:

ON ERROR
  IF *ERROR-NR = 3145
    IF this-block-is-found-by-back-tracing  /* <-- what to code here?
      WRITE 'Record is in hold! Try later!'
      ESCAPE ROUTINE /* or whatever
    ELSE
      INPUT 'Hit ENTR to retry'
      RETRY
    END-IF
  END-IF
END-ERROR

Maybe there’s something with IF *PROGRAM = ‘…’ or IF *LEVEL = … But I got no idea how to solve this.

I know, that it’s possible to give every module an ON-ERROR-Block to avoid a NAT1310. But that would be a real time-consuming job for me (we got > 10000 Modules).

Regards,

Matthias

[Wilfried_B]

USR2001N will give you detailed information about the last error. Only if level and program name match, you may issue a retry (this user exit will provide you with the missing system variables *error-program and *error-level)

[Matthias1 (Matthias)]

Thanks! USR2001N is a quite simple and working solution. It’s a shame that I didn’t know this by myself. :oops:

The USR2001N-Parameter P_ERROR_PROGRAM delivers an alphanumeric 32-byte field. But I tested it and it seems, that it’s always filled with the short name of the module. This is important in case of subroutine-modules, which have a long and a short name… But P_ERROR_LEVEL should work anyway.