split tests into headless and "all"
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Autopilot |
Fix Released
|
High
|
Unassigned | ||
autopilot (Ubuntu) |
Fix Released
|
Undecided
|
Unassigned | ||
Raring |
Won't Fix
|
Undecided
|
Unassigned |
Bug Description
I was talking with didrocks this morning about autopilot packaging and we would like to enable the tests during the distro build. The tests however need to be headless so we need to have two (easy) ways how to run:
1. headless tests
2. tests that require X
3. 1+2
The relevant IRC conversation:
<didrocks> oh? where is it (and can it run in headless mode?)
<mmrazik> I was talking with him about this as well
<didrocks> ah found, autopilot/tests/
<mmrazik> I was told some of the tests require X
<mmrazik> but I/thomi can probably disable that
<mmrazik> let me find the command thomi is using for running them
<didrocks> mmrazik: hum, ok, not required now, but would be nice to have a headless build and tests in the future
<didrocks> mmrazik: so that we can activate them on build
<didrocks> then, another target with both
Related branches
- Francis Ginther: Approve
- PS Jenkins bot: Approve (continuous-integration)
-
Diff: 176 lines (+87/-20)7 files modifiedautopilot/tests/README (+4/-0)
autopilot/tests/functional/__init__.py (+18/-0)
autopilot/tests/functional/test_application_mixin.py (+43/-0)
autopilot/tests/unit/__init__.py (+18/-0)
autopilot/tests/unit/test_application_mixin.py (+1/-20)
debian/control (+1/-0)
debian/rules (+2/-0)
Changed in autopilot: | |
status: | Fix Committed → Fix Released |
I'm taking a different approach here - and splitting the tests into unit tests, which can be run thusly:
autopilot run autopilot. tests.unit
and functional tests, which can be run thusly:
autopilot run autopilot. tests.functiona l
This is a better classification than "headless", since we cannot run tests that have any external dependencies not in the Build-Requires field.