aboutsummaryrefslogtreecommitdiffstats
path: root/spec/lib/timezone_fixes_spec.rb
Commit message (Collapse)AuthorAgeLines
* Remove timezone monkeypatch fixes because they're not necessary in Rails 3 ↵Matthew Landauer2013-01-25-0/+5
| | | | anymore
* read_attribute does timezone conversion in rails 3. So using ↵Matthew Landauer2013-01-25-13/+8
| | | | attributes_before_type_cast instead
* Wrap specs on the extraction of RFC-822 headers in code that sets the ENV ↵Louise Crow2012-12-11-16/+0
| | | | timezone. TMail renders headers using localtime, which is not ideal, but we're migrating away from it anyway, so I'm not sure it's worth delving into the internals of TMail to fix it.
* Fix failing timezone specs. As we are now setting a value for timezone in ↵Louise Crow2012-10-25-8/+40
| | | | | | config.timezone, this switches on ActiveRecord::time_zone_aware_attributes, so values are returned in the time zone specified (although still saved in whatever is specified in ActiveRecord.default_timezone in the db - in our case UTC). Previously with no timezone set, values would have been returned in UTC. So add some lines to look at the zone the data is saved in before the time_zone_aware_attributes kick in, and some lines to look at the values afterwards. Note that expected raw_saved time values are the same as the previous expected values for saved time (when there was no attribute time zone awareness), but that the time zone of values coming back on the model is set by config.time_zone and then overridden by Time.use_zone.
* Fix spec - based on the spec description, what's being tested and the other ↵Louise Crow2012-10-25-12/+12
| | | | specs, I assume that the activerecord default timezone is supposed to be set to :local as a precondition. Doing so makes the spec pass and make sense.
* Rename models and tables exim -> mail_serverMatthew Landauer2012-10-11-9/+9
|
* Make path work in Ruby 1.9Francis Irving2010-09-15-1/+1
|
* Rename spec files to libs, and to end _specFrancis Irving2009-12-03-0/+81