On Fri, 20 Feb 2004, Tim Bunce wrote:
> On Fri, Feb 20, 2004 at 10:30:44AM -0500, Rudy Lippan wrote:
> >
> > - Rewrote the primary_key_info() and primary_key() methods to
> > cleanly handle multi-column primary keys. Also added a
> > "onerow"
> > attribute to allow primary_key_info() to return a single
> > row containing multiple-column information. [Greg Sabino Mullane]
>
> Out of interest, why did you need a custom primary_key() method?
>
I think this was done while I was offline, so I am going by the code, but
I am CCing Greg who knows this section of code better than I do.
1. It looks it was done so that primary_key() could pass the {onerow=>2}
attribute in to primary key_info, but I don't really see a compelling
reason for doing it this way.
2. In {onerow=>2} above, onerow should be pg_onerow, and '2' should
probably be an exported constant rather than a magic number.
> > - Added the "noprefix" attribute to prevent the tables() method
> > from prepending the schema name. [Greg Sabino Mullane]
>
> pg_noprefix?
No it is not, but it should be. There I updated my copy.
Rudy
