Re: [HACKERS] Another reason to redesign querytree representation

> Bruce Momjian <maillist(at)candle(dot)pha(dot)pa(dot)us> writes:
> > Did you want to do the change for UNION, or were you just suggesting it
> > be done? I can easily add it to the TODO list.
> > Done:
>
> > * redesign UNION structures to have separarate target lists.
>
> Actually, it's not so much UNION that's busted as it is INSERT.
> The parser problems could be dealt with by having a two-level structure
> for INSERT ... SELECT ..., so that the targetlist for the eventual
> INSERT could be described without changing the semantics of the
> underlying SELECT.
>
> There might be other extensions needed for rules (paging Jan...) but
> as far as what I've been looking at goes, the TODO entry could be just
>
> * redesign INSERT ... SELECT to have two levels of target list.

Removed:
* Be smarter about promoting types when UNION merges different data types
* SELECT ... UNION ... GROUP BY fails if column types disagree
* INSERT ... SELECT ... UNION is not reliable

And added:

* redesign INSERT ... SELECT to have two levels of target list

--
Bruce Momjian | http://www.op.net/~candle
maillist(at)candle(dot)pha(dot)pa(dot)us | (610) 853-3000
+ If your life is a hard drive, | 830 Blythe Avenue
+ Christ can be your backup. | Drexel Hill, Pennsylvania 19026