Re: stawidth inconsistency with all NULL columns

Joe Conway <mail(at)joeconway(dot)com> writes:
> else if (null_cnt > 0)
> {
> /* We found only nulls; assume the column is entirely null */
> stats->stats_valid = true;
> stats->stanullfrac = 1.0;
> if (is_varwidth)
> stats->stawidth = 0; /* "unknown" */
> else
> stats->stawidth = stats->attrtype->typlen;
> stats->stadistinct = 0.0; /* "unknown" */
> }
> 8<-----------------

> So apparently intentional, but seems gratuitously inconsistent. Could
> this cause any actual inconsistent behaviors? In any case that first
> comment does not reflect the code.

Are you suggesting that we should set stawidth to zero even for a
fixed-width datatype? That seems pretty silly. We know exactly what
the value should be, and would be if we'd chanced to find even one
non-null entry.

regards, tom lane