Dmitry Gozman
cb30cb4eb7
feat(test runner): add --trace cli option ( #16277 )
2022-08-05 09:20:39 -07:00
Dmitry Gozman
2fa48483af
test: unflake "should retain traces for interrupted tests" ( #16235 )
...
Ensure both tests start before one of them finishes.
2022-08-04 09:18:18 -07:00
Dmitry Gozman
03fe75251b
fix(test runner): show tests as interrupted when maxFailures stops them ( #16178 )
...
Previously, we marked these tests as skipped, and it was sometimes
confusing, because they did actually run and produce some output/artifacts.
2022-08-03 15:25:25 -07:00
Dmitry Gozman
6a60de1b2f
test: add failing test for traces after interruption ( #15796 )
2022-07-19 13:50:52 -07:00
Pavel Feldman
030e7d211c
chore(har): allow replaying from zip har ( #14962 )
2022-06-17 16:11:22 -07:00
Ross Wollman
3974d1b359
revert( #14797 ): fix(test runner): collect artifacts when calling browser.close()
( #14846 )
...
Revert "fix(test runner): collect artifacts when calling `browser.close()` (#14797 )"
This reverts commit c7a28ac7e9
.
Looks like it broke a bunch of tracing tests:
* Good: https://github.com/microsoft/playwright/runs/6838098316?check_suite_focus=true
* First Bad: https://github.com/microsoft/playwright/runs/6838104691?check_suite_focus=true
* Still bad on HEAD (88664c39c9
): https://github.com/microsoft/playwright/runs/6868333846?check_suite_focus=true
2022-06-13 17:20:59 -07:00
Dmitry Gozman
c7a28ac7e9
fix(test runner): collect artifacts when calling browser.close()
( #14797 )
...
Previously, we only collected artifacts on context closure.
However, in serial mode it is possible to close the browser instead.
2022-06-10 14:11:34 -07:00
Yury Semikhatsky
2bcdf68ef5
fix(runner): do not override trace files in afterAll ( #14529 )
2022-05-31 15:21:51 -07:00
Dmitry Gozman
c3beb71b07
fix(test runner): do not run automatic fixtures for beforeAll hooks ( #14104 )
...
There are a few issues this covers:
- Some fixtures like `page` and `context` are not allowed in `beforeAll`
hooks, so using them in automatic fixture makes it throw.
- Running automatic fixture solely for `afterAll` is unexpected.
This currently happens when `afterAll` is run for cleanup after
fixture timeout/throw.
For built-in playwright fixture, we keep `'all-hooks-included'` auto mode.
Added a doc explaining the execution order.
2022-05-13 11:17:20 +01:00
Pavel Feldman
bde7bf4ea9
chore: move registry & dispatchers to under server ( #13370 )
2022-04-06 22:21:27 -07:00
Pavel Feldman
d14ecec612
chore: remove flaky-by-design test ( #12434 )
2022-03-01 08:52:52 -08:00
Yury Semikhatsky
706c897031
feat(fixtures): respect tracing config for APIRequestContext ( #11954 )
2022-02-09 08:54:09 -08:00
Andrey Lushnikov
2336692e8a
feat: support clarification message for expect ( #11735 )
...
The clarification message is displayed in the HTML report as the name of the step:
![image](https://user-images.githubusercontent.com/746130/151852652-48194140-5ea4-439d-afee-12583a8caf71.png )
It is also shown in terminal output:
![image](https://user-images.githubusercontent.com/746130/151852666-5c956ef1-6e94-4bc2-8e55-b58688dfc7e0.png )
Fixes #7816
2022-01-31 17:14:59 -08:00
Pavel Feldman
7b64161a37
feat(test-runner): allow specifying fine-grained trace options ( #10147 )
2021-11-08 15:39:58 -08:00
Dmitry Gozman
3c1aaa5338
fix(trace): do not call tracing.stopChunk() twice ( #10054 )
2021-11-04 21:08:42 -07:00