PG Bug reporting form <noreply@postgresql.org> writes:
> On line 223 there is a check of string 'name' and in this case it is
> initiallized and then is being checked for null once more with no reason.
This code is taken from the upstream tzcode distribution, in which
TZDEFAULT can be NULL. (Or so I assume anyway without having looked.)
I'm disinclined to diverge from their code just to satisfy some static
analyzer complaining about useless code --- any reasonable C compiler
will discard that code, so it's not a problem.
> The code becomes quite buggy.
If there is actually a bug here, I'm not seeing it.
regards, tom lane