On Jul 23, 8:52 pm, MRAB <pyt...@mrabarnett.plus.com> wrote: > > dt_twothirty=dt_localtime.replace(hour=settings.UPDATE_TIME_HOURS,minute=se > > ttings.UPDATE_TIME_MINS,second=0,microsecond=0) > > You're changing the time of day, but not the date. You might want to add > a day to the shutdown time if it's earlier than the current time.
I started out by finding the right date (I used a less-than test and toordinal and fromordinal() ). However, after some trials, I came to believe that I don't need to find the right date. The part of that calculation I need, and later refer to, is the .seconds attribute. I perceive that the way TimeDelta objects are laid out, the seconds attribute will be the same, regardless of whether I calculate it using 2:30 today or first finding which is the right date and using its 2:30. That is, as I understood it, if it is now 2:29 then the .seconds attribute will be 60. If it is now 2:31 then the .seconds attribute will be 24*60*60-60. I believe this holds regardless of which day I use. >>> import time,datetime >>> x=datetime.datetime(2010,7,23) >>> dt_localtime=x.fromtimestamp(time.time()) >>> dt_localtime datetime.datetime(2010, 7, 24, 7, 17, 46, 122642) >>> dt_twothirty=dt_localtime.replace(hour=2,minute=30,second=0,microsecond=0) >>> print dt_twothirty 2010-07-24 02:30:00 >>> dd_diff=dt_twothirty-dt_localtime >>> print dd_diff -1 day, 19:12:13.877358 >>> dt_tomorrow_twothirty=dt_localtime.replace(day=25,hour=2,minute=30,second=0,microsecond=0) >>> print dt_tomorrow_twothirty 2010-07-25 02:30:00 >>> dd_tomorrow_diff=dt_tomorrow_twothirty-dt_localtime >>> print dd_tomorrow_diff 19:12:13.877358 Tested it, of course. Not that I haven't gotten things wrong in the past, even though I tested them. :-} Jim -- http://mail.python.org/mailman/listinfo/python-list
