Test create stats report #2170
Test create stats report #2170
Conversation
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## master #2170 +/- ##
==========================================
+ Coverage 56.87% 57.36% +0.48%
==========================================
Files 73 73
Lines 17351 17351
Branches 3368 3368
==========================================
+ Hits 9869 9954 +85
+ Misses 6625 6529 -96
- Partials 857 868 +11
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
|
Git conflicts ☝️ |
test/unit/test_diagram.py
Outdated
| import datetime | ||
| from pathlib import Path | ||
|
|
||
| from mock.mock import patch |
There was a problem hiding this comment.
tepmfile and patch can be replaced by the built-in fixtures that come with pytest: https://docs.pytest.org/en/6.2.x/fixture.html#pytest-fixtures-explicit-modular-scalable
For tempfile there's tmp_path which returns a Path already for you (I think you may not need to import Path` here, maybe.
And for patch, the pytest_mock lib we use has mocker, https://pytest-mock.readthedocs.io/en/latest/usage.html
Both work the same, but with pytest fixture the code is a bit shorter (one of the main selling points of pytest over unittest was to write less to have the same -- no classes, just functions, no need to import a lot of things, just use fixtures, etc...)
There was a problem hiding this comment.
I've recently written a few tests for stats and monitor in #2322 , and there you can see the mocker.patch fixture being used.
test/unit/test_diagram.py
Outdated
| jobs_by_association.return_value = [JobData(),JobAggData()] | ||
| create_status = diagram.create_stats_report(expid, jobs_data, {}, temp_dir,True,True,False,None, | ||
| None,None) | ||
| assert create_status is False |
There was a problem hiding this comment.
😆 this is not really fair. This is the function you mentioned that receives a bool and returns that same bool. So this test will pass whether we break the create_stats_report or not (the focus when writing tests must be to prevent bugs, not only coverage).
I think a better test would be one that either produces a plot somewhere in a temp dir, and then we assert the plot was created, has more than 0 bytes, and maybe that's that. Or then mock just the matplotlib part, and confirm the mock was called as expected.
|
Waiting for approval of the issue #2169 to refine this issue |
VindeeR
force-pushed
the
test_create_stats_report
branch
5 times, most recently
from
1e75b9b to
a73db24
Compare
VindeeR
force-pushed
the
test_create_stats_report
branch
4 times, most recently
from
c3258ed to
9770e67
Compare
…g to ensure it works as intended
VindeeR
force-pushed
the
test_create_stats_report
branch
from
9770e67 to
0d764a5
Compare
|
@kinow I used the new fixture to develop the test and a bit of tidiness of the code too Thanks for the master class yesterday, I'm not sure I would manage this branch this easily without it ❤️ |
It was fun!!! Thanks for the different Thursday! |
Create a test case for the function create_stats_report asserting if functions is creating files properly with the information from the jobs