Re: Program Cross reference file

[Vern Hamberg <vhamberg@xxxxxxxxxxxxxxxxxxxxxxxxx>]

This makes me think you could do the DSPFD type(*MBRLIST) against all 
source files you are interested in (could be generated from the DSPOBJD 
list, as the x-ref products do, at least). You might want to correlate 
these lists together somehow.

To find gaps in the names, you can go through them in order, checking that 
each one is in sequence with the previous one. Control level processing in 
RPG could be useful, I would think.

You might be able to cobble together an SQL statement that returns a record 
where the sequence number (substrung and converted to numeric) is more than 
1 higher than the max of that sequence number for all records with the name 
less than the one from the current record. It feels doable with subqueries 
- don't have time to work out the details. You'd group, possibly on the 
first 3 characters, if your naming convention is really well-managed.

HTH

Vern

At 06:39 AM 2/4/2003 -0800, you wrote:
> If you are looking for a "real-time" always-current file, aka QADBXREF
> for files, I don't think you're going to find one for program objects.
>
> Other than DSPOBJD, the only other thing that is "closer" to the OS is
> RTVDSKINF.  Still, it is not real-time.
>
> Frankly, I think you'll find your best solution will be DSPOBJD
> *ALL/*ALL.
>
> Another thought:  Depending on the size of your development team, you
> may also want to ensure that no one else is using a source member name
> that hasn't compiled to an object yet.  So, if you find that SAR123 is
> not an object anywhere on the system and start to develop a program
> with that name, only to find out that the developer next to you is
> using the same name for his new program, well you get the idea.  (Use
> DSPFD *MBRLIST)
>
> hth, Dan
>
> --- SCarter@rsrcorp.com wrote:
> >
> > Did not want to do a dspobjd if there was a file maintained by the
> > OS.