On 2017-01-30 17:24:31 -0500, Tom Lane wrote:
> Make it work like Agg and WindowFunc. To wit, dump the actually special
> function calls (the set-returning functions) into a list that's internal
> to the FunctionScan node, and then anything above those goes into scalar
> expressions in the node's tlist, which refer to the SRF outputs using
> Vars or things morally equivalent to Vars.
Hm. That should be fairly doable. (I'd advocate very strongly against
building that list via ExecInitExpr, but that's an implementation
detail). We'd evaluate SRFs early, but that's just consistent with
targetlist SRFs.
Wonder if we there's an argument to be made for implementing this
roughly similarly to split_pathtarget_at_srf - instead of injecting a
ProjectSet node we'd add a FunctionScan node below a Result node.
- Andres