On 19 January 2014 11:43, Marko Tiikkaja <marko@joh.to> wrote:
> New version attached, without the doc change.
>
This looks good to me.
- applies cleanly.- compiles with no warnings.- passes a sensible set of new regression tests.- implements the agreed
behaviour,per SQL spec.- I can't think of any corner cases to break it.
I think this is ready for committer, although I would also like to see
the doc changes to make the table of array function descriptions a bit
more explicit about corner cases.
Also, does this mean that we can now claim full support for SQL
feature S091 "Basic array support"?
Regards,
Dean