[Date Prev][Date Next][Thread Prev][Thread Next]
[Date Index]
[Thread Index]
- Subject: Re: os.time() and timezones
- From: Sean Conner <sean@...>
- Date: Sat, 14 Apr 2012 16:47:12 -0400
It was thus said that the Great Rena once stated:
> On Fri, Apr 13, 2012 at 14:26, Sean Conner <sean@conman.org> wrote:
> > It was thus said that the Great steve donovan once stated:
> >> On Fri, Apr 13, 2012 at 10:16 AM, Rena <hyperhacker@gmail.com> wrote:
> >> > Ouch.) Does Lua provide a way to turn a table into a Unix timestamp
> >> > under the assumption that the table specifies GMT, not the local
> >> > timezone?
> >
> > Even operating systems have trouble with timezones:
> >
> > http://www.chronos-st.org/Discovering%20the%20Local%20Time%20Zone--Why%20It%27s%20a%20Hard%20Problem.html
> >
> >> I use this kind of logic to find the timezone offset in pl.Date:
> >>
> >> local t = os.time()
> >> local ut = os.date('!*t',t)
> >> local lt = os.date('*t',t)
> >> thour = lt.hour - ut.hour
> >> tmin = lt.min - ut.min
> >>
> >> If this is misguided I would certainly like to know!
> >
> > It is. Try this:
> >
> > gmt = { 0 , 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 ,
> > 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 }
> > loc = { 19 , 20 , 21 , 22 , 23 , 0 , 1 , 2 , 3 , 4 , 5 , 6 ,
> > 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 }
> >
> > for i = 1 , 24 do
> > print(i,loc[i] - gmt[i])
> > end
> >
> > And even when it does work, it still doesn't take into account DST. I found
> > this works:
> >
> > now = os.time()
> > lmt = os.date("*t",now)
> > gmt = os.date("!*t",now)
> > timel = os.time(lmt)
> > timeg = os.time(gmt)
> > zone = os.difftime(timel,timeg)
> >
> > if lmt.isdst then
> > if zone < 0 then
> > zone = zone + 3600
> > else
> > zone = zone - 3600
> > end
> > end
> >
> > -spc (Time is hard! I wonder how The Doctor does it?)
> >
> >
>
> Hmm, are you sure about that DST correction?
Actually, I'm wrong [1]. It should be:
if lmt.isdst then
zone = zone + 3600
end
Man, this stuff is nasty.
-spc (Needs to spend more time testing ... )
[1] Yes, I know---I'm not taking into account time zones that aren't an
hour long.