[BioSQL-l] Re: load_taxonomy.pl

[Aaron J Mackey]

Hi Brian,

I think the reason it resides with biosql is that it really is part of
instantiating a biosql database (i.e. it's an inital data load), whether
you're going to be using biojava, bioperl or biopython (or just using the
database via SQL).  bioperl-db is a rich object-to-relational-db mapping
API (but you're right, it does have database loading scripts that make
use of that API).

That said, one of the criticisms of load_taxonomy.pl is that it not only
requires Perl, but also DBI, DBD drivers, etc; someone wanting to use
biosql with biojava may not have all those things installed (nor know how
to do it).  So a better, more language-agnostic load_taxonomy.pl script
would be a laudable goal (one that simply dumped raw SQL commands that
could be piped into the database); but I haven't had time to do that.

-Aaron

On Thu, 21 Aug 2003, Brian Osborne wrote:

> Aaron,
>
> Would you object to my moving load_taxonomy.pl from biosql to bioperl-db? It
> seems to me that the biosql package is mostly concerned with installation of
> the database, and that the Perl loaders should go in bioperl-db
> (load_seqdatabase and load_ontology are in bioperl-db).
>
> By the way, I used load_taxonomy - ran flawlessly.
>
> Thanks again,
>
> Brian O.
>
>