Re: [Gnumed-devel] Consumable substances

[Karsten Hilbert]

On Mon, Sep 05, 2011 at 02:31:50AM -0700, Jim Busser wrote:

> > - you may need to investigate "\copy" (which
> >  client/pycommon/gmPsql.py only partially and non-testedly
> >  implements)
> 
> Works under psql.

I know but that does not matter (except for it being a
fairly good test) because data packs are run under
gmPsql.py.

> Presumably I could test it once the data packs are online.

Yes. I will upload your supplied test pack.

> > over using the SQL level "copy" command (which
> >  has certain conceptual limitations in that data files need
> >  to live on the machine the copy is actually run on rather
> >  than from where it is invoked - although no matter in local
> >  installs)
> 
> Whichever GNUmed admin user who would be running GNUmed and selecting the menu
> 
>       Install data pack
> 
> in any multiuser praxis would be unlikely to be using the
> actual server machine. We may be accustomed to local being
> ok on account of some developers using GNUmed clinically on
> laptops as standalone machines with no other users
> connecting.

Hence we either need use of direct SQL or else use the SQL
copy (we can't use \copy because we don't use psql).

> > - I am somewhat unsure about the meaning of ".exit" at the last-but-one line
> 
> This was a precaution which I had used with SQlite in case
> I had wanted to retain any obsoleted code an the bottom of a
> script. Does Postgres accept this (or other) method of
> signalling to stop processing the file?

It would need to be supported by gmPsql.py which is
processing the file.

Offhand I am not aware of a method to effect that.

Karsten
-- 
GPG key ID E4071346 @ gpg-keyserver.de
E167 67FD A291 2BEA 73BD  4537 78B9 A9F9 E407 1346