-
-
Notifications
You must be signed in to change notification settings - Fork 7.9k
Allow image comparison outside tests module #5842
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Allow image comparison outside tests module #5842
Conversation
…ts' module, but issue warning because the code logic seems a bit fragile and may fail.
There is one failure in the Travis build but it seems to be the PEP8 issue that was fixed by #5844. Not sure what's making the AppVeyor builds fail to be honest. |
power cycled the PR to re-kick CI against current master. |
👍 from me once travis passes. This should be non-controversial and I think be backported back to v1.5.x |
appveyor failed on a couple of the job runs with a strange error. I can't restart them. |
Do not worry about appveyor too, it also lies that it passes ;) On Wed, Jan 13, 2016 at 1:24 PM Benjamin Root notifications@github.com
|
Ok. Merging and backporting to v1.5.x. Do I also need to manually backport it to the v2.x branch, or will that happen on its own? |
…e_tests_module Allow image comparison outside tests module
…e_tests_module Allow image comparison outside tests module
backported to v1.5.x as 5240759 |
every so someone does a 1.5.x -> 2.x -> master cascade |
…son_outside_tests_module Allow image comparison outside tests module
This is a replacement for PR #5518 (which should be closed if this is accepted). For now this PR simply removes the
assert
statement so that the@image_comparison
decorator can be used in situations where the test does not live in a parent directory calledtests
(this does not apply to matplotlib but will be useful for third-party users).In the medium term it would be good to re-think and refactor the code in
_image_directories
because isn't very easy to read at a glance if you don't know what's going on already, and the logic seems somewhat fragile (see #3314 for example).