Open jerrymarino opened 5 years ago
Hmm... I suspect the output parser is getting confused from EarlGray's output. Do you have an app where this reproduces?
Thanks @ob - it's totally possible, I'm not super familiar with the expected behavior
One other interesting bit from a similar log:
Test Suite 'XX' failed at 2019-04-09 15:26:56.965.
Executed 0 tests, with 1 failure (1 unexpected) in 0.000 (4.644) seconds
XCTestOutputBarrierTest Suite 'XX.xctest' failed at 2019-04-09 15:26:56.979.
Executed 0 tests, with 1 failure (1 unexpected) in 0.000 (4.660) seconds
XCTestOutputBarrierTest Suite 'Selected tests' failed at 2019-04-09 15:26:56.981.
Executed 0 tests, with 1 failure (1 unexpected) in 0.000 (4.667) seconds
XCTestOutputBarrierXCTestOutputBarrier
I really don't get what this means in XCTest
terms: 0
tests with 1
failure.
This bluepill
invocation was running with --repeat-count 0
so perhaps thats' expected behavior.
I also didn't see any entries in the log file which would trigger -[BPExecutionPhaseProtocol onTestCaseFailed..]
or -[BPExecutionPhaseProtocol onTestCasePassed..]
.
None the less, I ran a similar log file through the test harness and am seeing this:
{87381} 20190409.170742 [34m[ INFO ][0m (BLUEPILL) BPTree parser completed
{87381} 20190409.170742 [33m[ DEBUG ][0m (BLUEPILL) [Selected tests] Mismatch numberOfFailures calculated (0) vs reported (1)
{87381} 20190409.170742 [33m[ DEBUG ][0m (BLUEPILL) [XX.xctest] Mismatch numberOfFailures calculated (0) vs reported (1)
{87381} 20190409.170742 [33m[ DEBUG ][0m (BLUEPILL) [XXTests] Mismatch numberOfFailures calculated (0) vs reported (1)
{87381} 20190409.170742 [33m[ DEBUG ][0m (BLUEPILL) [XXTests] Mismatch totalTime calculated (0.000000) vs reported (4.644000)
that can happen if your +setUp test method throws an exception. For example I see:
Test Suite 'rdar46239726Tests' failed at 2019-04-09 23:06:30.857.
Executed 0 tests, with 1 failure (1 unexpected) in 0.000 (0.004) seconds
Test Suite 'rdar46239726Tests.xctest' failed at 2019-04-09 23:06:30.858.
Executed 0 tests, with 1 failure (1 unexpected) in 0.000 (0.005) seconds
Test Suite 'Selected tests' failed at 2019-04-09 23:06:30.859.
Executed 0 tests, with 1 failure (1 unexpected) in 0.000 (0.006) seconds
Much appreciated @khandpur! Do you have any suggestions for how to make bluepill rerun tests that hit this? It looks like a lot of the cases that hit this issue are happening due exceptions in setup
.
I think this is a complicated change. AFAICT, Bluepill doesn't have a good place to put failures in test suites (not test cases). I'm not even sure where that failure would fit in the JUnit report, or how the retry mechanism would work... let me think about this and see if I can come up with something 😅
Good afternoon!
I'm using EarlGrey with Bluepill, and noticed that there was a subset of failures that are never retried or reported.
Here's the stats:
XCTest says:
Any suggestions @bayareabear or @ob ?