Also, found the patch I
wrote: https://gist.github.com/jarshwah/a541857db195f0486b9e
Instead of multiple subclasses, perhaps Func could take on some of the
functionality of transform directly, but that makes no sense for > 1 arity
functions. A nice design should be able to be found.
On Thursda
Transforms are just single argument expressions available using lookup strings.
The biggest complication is that Transforms have slightly different API to
expressions.
I don't see a reason why transforms couldn't be a special case of expression.
There are some API incompabilities, but they shou
FWIW I put a rough patch together awhile ago (I can't find it or the
discussion unfortunately) that attempted to make Transforms usable as
expressions.
They share partial APIs but not the full set which made that first attempt
quite messy/hacky. I'm not aware of any reasons that they couldn't be
Hi list,
The thread about reimplementing the date-based lookups[1] reminded me of
something that was bugging me a couple of days ago.
There's some overlap between Funcs (of arity 1) and Transforms. They seem
to fundamentally do the same thing – wrap an expression in some arbitrary
SQL – but wi
