Andrew Dunstan <andrew(dot)dunstan(at)2ndquadrant(dot)com> writes:
> OK, here's the whole series of patches.
I've not tested it at all, but this looks generally sane in a quick
once-over.
A minor quibble is that in 0003, you weren't terribly consistent about
argument order --- in some places you have the FmgrInfo argument added
before the collation argument, and in some places after. I'd suggest
trying to make the argument orders consistent with the fmgr.c support
functions. (I'm generally -1 on blindly adding stuff at the end.)
regards, tom lane