Andrew Gierth <andrew(at)tao11(dot)riddles(dot)org(dot)uk> writes:
> My inclination is to fix this in the planner rather than the executor;
> there seems no good reason to actually hash a duplicate column more than
> once.
Sounds reasonable --- but would it make sense to introduce some
assertions, or other cheap tests, into the executor to check that
it's not being given a case it can't handle?
regards, tom lane