[Bioperl-l] bioperl-db: adaptors

[Jason Stajich]

On Mon, 8 Jul 2002, Hilmar Lapp wrote:

> Along the tests I'll be fixing adaptors as needed. What I already
> realized is that the virtual absence of any documentation makes a
> programmer's life particularly both easy and enjoyable. Whenever I'll
> fix a method I'll try to at least add arguments and return values
> along with types, provided that I understand what they are supposed to
> be. It'd be helpful if other contributors could (try to) do at least
> that, too.
>
ugh - docfest can't come soon enough.
Thanks for leading the way here, we'll try to encourage documentation more
forcefully here.

> As for adaptor methods, it is generally useful for callers if the
> methods return something (there are only few cases in which there's
> nothing meaningful to return). I'll fix the remove_XXX methods along
> my way (these may be a limited number only) to return the number of
> actually deleted entries (DBI returns that), but for a consistent
> behaviour we probably have to go through these in a concerted effort.
>
> Another thing I've noticed is the style to quote variable values in
> SQL statements by hand instead of having the DBD driver do that (which
> is most likely at least as good at that, but probably many times
> better). Unless there's a particular design decision behind that, I'll
> also fix that where I find myself fixing something already, but for a
> general car detailing there needs to be a concerted effort or more
> time.

Very bad - we should be using late binding as you describe.

>
> Shout if you think this is going to interfere with your client code.
>
> 	-hilmar
>

-- 
Jason Stajich
Duke University
jason at cgt.mc.duke.edu