RE: [HACKERS] implementing outer joins - Mailing list pgsql-hackers

From Stupor Genius
Subject RE: [HACKERS] implementing outer joins
Date
Msg-id 000101bda691$d5825880$c697accf@darren
Whole thread Raw
In response to implementing outer joins  (Brett McCormick <brett@work.chicken.org>)
List pgsql-hackers
>
> does anyone have any idea how difficult it would be to implement, and
> perhaps point me in the right direction?
>

I forget where, but there is a place in the code where the two values
of the join are compared to see whether the row qualifies.  Around
this spot there is a check for one of the values being null and if it
is, the function doesn't include the data.  Seems that modifying that
spot or using it as a basis for your code could be a quick start for
the left join.  The right could perhaps then be converted beforehand
in the optimizer to a left to be handled there.

But there's prolly a great and sane reason why this wouldn't work.

Haven't thought about the full outer join yet.  Don't have my machine
completely configured yet and it's hard to stay in to do that with it
being summer and all.

Later,
darrenk

pgsql-hackers by date:

Previous
From: Brett McCormick
Date:
Subject: implementing outer joins
Next
From: Brett McCormick
Date:
Subject: you are lost in a maze of twisting code, all alike