Re: Possibly misleading documentation of Template Patterns for Date/Time Formatting

pá 17. 4. 2020 v 17:52 odesílatel Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> napsal:

> David Kubecka <davidkubecka366(at)gmail(dot)com> writes:
> > on the official docs
> > https://www.postgresql.org/docs/9.6/functions-formatting.html see the
> table
> > 9-24 and Pattern "Q". The doc (for version 9.6) says:
>
> > quarter (ignored by to_date and to_timestamp)
>
> > All the later versions of the doc (10, 11, 12) miss the "ignored" note
>
> It's still there, just further down:
>
> * In to_timestamp and to_date, weekday names or numbers (DAY, D, and
> related field types) are accepted but are ignored for purposes of
> computing the result. The same is true for quarter (Q) fields.
>
> I think this was changed because we noticed that the docs failed to point
> out the issue for weekday fields, and cramming similar annotations into
> their already-long table entries didn't make sense. So the info got moved
> to the commentary below.
>

I see. I wonder if it wouldn't be better e.g. to reference this info from
all the ignored fields in the table through an asterisk comment just under
the table (thus making the list of unsuppored patterns even more explicit).
Just an idea, It's definitely not an easy task to make a good documentation
:-)

Anyway, could you shed some light if there's any particular reason why TO_DATE(
'2012-4', 'YYYY-Q' ) isn't supported?

> regards, tom lane
>