Consistently include test setup and teardown in duration measurement#380
Consistently include test setup and teardown in duration measurement#380localheinz merged 1 commit intoergebnis:mainfrom
Conversation
664e23d to
3fc2545
Compare
Codecov ReportAttention:
Additional details and impacted files@@ Coverage Diff @@
## main #380 +/- ##
============================================
- Coverage 70.54% 69.86% -0.68%
Complexity 110 110
============================================
Files 26 26
Lines 516 521 +5
============================================
Hits 364 364
- Misses 152 157 +5 ☔ View full report in Codecov by Sentry. |
f38752b to
15c3c4d
Compare
6210507 to
e910fd3
Compare
78efb66 to
843b68a
Compare
|
Two things:
What do you think? |
e22cf77 to
694e831
Compare
I belive it should. When failed test failed fast, it is never reported. And in general, tests do not fail slower than when passed fully. But if a failed test took longer than a maximum duration, it should be reported to make the developer aware there is also some slowness issue. Especially when run remotely/in CI.
I am looking into the PHPUnit v8/9 hooks to make the behaviour consistent. |
ee2f558 to
0eb24f2
Compare
b06a140 to
8070633
Compare
33d6fe2 to
c9c42f7
Compare
945837c to
5eea786
Compare
5eea786 to
9a06d32
Compare
|
Rebased. |
|
Don't worry about rebasing, I'm working on it and will finish it in a moment! |
e744cbb to
dee1111
Compare
Co-authored-by: Andreas Möller <am@localheinz.com> Co-authored-by: Michael Voříšek <mvorisek@mvorisek.cz>
dee1111 to
e418bd4
Compare
|
Thank you, @mvorisek! |
fix #378, measure the whole
TestCase::runBare()durationalso measure any test, not only the successful ones, measuring failed/errored tests is helpful especially for tests using IO or external services