Today my tests started to fail even though I haven't changed anything for weeks
(first day back after a period of illness). I have isolated the problem to the
fact that when I call DateTime.toDateMidnight() it *sometimes* changes the
offset from '+01:00' to 'Z'.
As an experienced programmer, the first thing I would be thinking after reading
this is "muppet must be doing something stupid" only I don't think I am :).
The following (groovy) snippet exhibits this behaviour:
--- start
@Test
public void silly() {
def days = 20
DateMidnight now = new DateTime()
DateMidnight nowPlusDays =
now.toDateTime().plusDays(days).toDateMidnight()
def diff = nowPlusDays.millis - now.millis
def expectedMillis = days * 24 * 60 * 60 * 1000
assertEquals expectedMillis, diff
}
--- end
(I realise the above code isn't the optimal for the test but it is
representative of what happens in production)
Now - if I set "days" to be 7, for example, it works as expected. If I set it
to be 20 then it fails. When I debug I can see that the toString method of now
returns "2011-10-11T00:00:00.000+01:00" however nowPlusDays returns
"2011-10-31T00:00:00.000Z". The iChronology of both of them is
"ISOChronology[Europe/London]".
This code has been working for years(!) although in production I am only ever
using this to determine when things should expire so it may well be that I
haven't picked it up. The tests are run on every checking (multiple times a
day) and this is the first time it has failed.
Any ideas - this has come completely out of the blue and I have no idea how to
work around it. I am assuming it is a bug and I couldn't find any FAQ items
related to this...
On an aside - what is the idiomatic way to determine the number of whole days
between two dates? I use DateMidnight as I thought that would simplify things
;).
Many thanks,
Col
------------------------------------------------------------------------------
All the data continuously generated in your IT infrastructure contains a
definitive record of customers, application performance, security
threats, fraudulent activity and more. Splunk takes this data and makes
sense of it. Business sense. IT sense. Common sense.
http://p.sf.net/sfu/splunk-d2d-oct
_______________________________________________
Joda-interest mailing list
Joda-interest@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/joda-interest
