Closed jayvdb closed 4 years ago
Oddly, test_add_channel_invalid_encoding
seems to have fixed itself.
test_parse_invalid_uri
looks very easy to fix, but requires some knowledge of what should occur in this case.
test_parse_invalid_uri
looks very easy to fix, but requires some knowledge of what should occur in this case.
I think treating the actual observed behaviour as de-facto correct is fine.
It seems a bit odd, assuming the test passed at some point.
In effect, somehow, the test result is inverted.
It seems a bit odd, assuming the test passed at some point.
I seem to remember that Python's urlparse module has been updated to be more standard-compliant. So that would explain the change in behaviour.
Fixed in pywebsocket3: https://github.com/GoogleChromeLabs/pywebsocket3
Version 1 of pywebsocket is no longer maintained so closing the issue here.
Two failures
See https://travis-ci.org/jayvdb/pywebsocket/jobs/512359637 for full log