Re: LATERAL, UNNEST and spec compliance

* Tom Lane (tgl(at)sss(dot)pgh(dot)pa(dot)us) wrote:
> However ... David is wrong to claim that it's zero-risk. It's true that
> an SRF can't contain any side-references today, but it can contain an
> outer reference. Consider a case like
>
> SELECT ... FROM a WHERE a.x IN (SELECT ... FROM b, srf(y) WHERE ...)

I see what you mean, but on the other hand, that looks like something we
might actually want to complain about as 'y' is pretty clearly ambiguous
here. I'm a bit surprised that doesn't already throw an error.

> This is a little bit far-fetched, but it could happen. As against that,
> we make incompatible changes in every release, and it does seem like
> assuming LATERAL for functions in FROM would be a usability gain most
> of the time. And special-casing UNNEST to satisfy the standard seems
> *really* ugly.

It's definitely far-fetched, imv. If it's possible, within reason, to
explicitly throw a "please disambiguate 'y'" type of error in those
specific cases, that'd be nice, but I don't think it'd be required. A
mention in the release notes would be sufficient.

Thanks,

Stephen