Re: [Gnumed-devel] removing test data

[Karsten Hilbert]

On Fri, Aug 07, 2009 at 05:36:11PM -0700, Jim Busser wrote:

> >We cannot
> >add that to the upgrade/bootstrap because it would unfixably
> >break the data consistency tests. People will have to run it
> >manually.
> 
> If I understand correctly, we hope / expect that this script will
> work provided that a person would run it *manually* (from the
> command line), but it should be ok to run it without any other
> special preparation steps, even immediately after running upgrade /
> bootstrap?

That is correct.

> In which case is it any different if one would run a script that would
>       - call bootstrap, and then
>       - call the drop script
> ??

No. The difference is between "after" and "integrate with".

> >I will provide two more scripts: one that drops data for the
> >"Hilbert" test persons and one that drops Kirk and McCoy.
> 
> I imagine that dropping this other data requires dropping records
> from a wider variety of tables. Once the task of crafting the sql is
> complete, is there a reason the additional code cannot be appended
> together, into a single script?

Yes. There exist databases in the wild which hold real
patient data for the Hilbert patients. We better not drop
those.

Kirk and McCoy go together. People should make a decision on
whether to drop them. They are "more valuable" in terms of
test data.

> Might a drop_demo_patients script tie together (run) each of the
> component scripts?

It could.

Karsten
-- 
GPG key ID E4071346 @ wwwkeys.pgp.net
E167 67FD A291 2BEA 73BD  4537 78B9 A9F9 E407 1346