On 2/15/07, Bruce Momjian <bruce@momjian.us> wrote:
> Bruce Momjian wrote:
> > Brendan Jurd wrote:
> > > On 2/15/07, Bruce Momjian <bruce@momjian.us> wrote:
> > > > Tom Lane wrote:
> > > > > Bruce Momjian <bruce@momjian.us> writes:
> > > > > > Peter Eisentraut wrote:
> > > > > >> I don't think Oracle even has that. But personally I'd like to see
> > > > > >> errors for invalid pattern combinations.
> > > > >
> > > > > > What do we do with other invalid pattern combinations in to_char() now?
> > > > >
> > > > > Mostly, we return bogus results :-(. The formatting.c code in general
> > > > > doesn't seem very robust.
> > > >
> > > > Yep, seems every release I am in there cleaning up some mistake repeated
> > > > multiple times in the code. It needs a good cleaning.
> > >
> > > I'm happy to volunteer to do something about the invalid field
> > > combinations, but I suspect an overhaul of formatting.c is more than I
> > > can currently chew. I figure it would be a bit misguided of me to put
> > > together a patch for invalid field combinations if somebody is about
> > > to do a rewrite of much of the code?
> >
> > Yea, I was just throwing out a note that someday if someone has time,
> > that file need a good sweeping.
>
> Sorry, I wasn't clear. No one is currently working on overhauling
> formatting.c, so if you want to submit _any_ patch to improve the file,
> please do. :-)
It'll be a pleasure Bruce. I think it would be best to wait until the
existing ISO week date patch has gone through before working on it
though.
Perhaps a TODO item is in order?