Re: [HACKERS] empty concatenate

Thus spake Tom Lane
> "D'Arcy" "J.M." Cain <darcy@druid.net> writes:
> >> ! Why is textcat() (and other) function called if result from this 
> >> function is ignored, it is bad spending (my CPU is not boredom). See 
> >> my 'C' example in my first letter...  
> 
> > This is the issue no matter which side of the debate you are on.
> 
> "Debate"?  There's no debate --- everybody agrees that the current
> fmgr interface doesn't handle NULLs reasonably.  It's just a matter
> of finding time to fix it.  It's a fairly large project, given the
> amount of code that needs to be touched.

Well, it may have been a lopsided (and friendly) debate but there was
definitely two sides.  The one (which I assume you mean as the one
that "everyone" accepts says to stick to SQL conformance and fix it
so that the functions are just never called.  The other said to have
the functions called then use the value returned so that each function
could decide what to do with NULLs.

-- 
D'Arcy J.M. Cain <darcy@{druid|vex}.net>   |  Democracy is three wolves
http://www.druid.net/darcy/                |  and a sheep voting on
+1 416 425 1212     (DoD#0082)    (eNTP)   |  what's for dinner.