Thanks Håvard. Placing @pytest.mark.skip will get the job so worse come
to worse we can deal with it that way. The reason why I'd prefer not to
do it that way is that that here we only want to control the tests being
run from debian/tests. The @pytest.mark.skip will disable them for all
testing (e.g. users running the tests locally), which we don't
necessarily want.
The thing I don't understand is that the skips currently in place (in
debian/tests/python3) were working fine until relatively recently. I
don't know why they've stopped functioning (I'm assuming some syntax
change for handling the skip structure that is currently used in
debian/tests/python3).
The configuration of debian/tests/python3 of obfuscated by use of
junitxml, so hard to check. I think the problem is testid is no longer
matching the skip entries, where previously it did match.
Perhaps activating the -k-slow option like you suggested might be
simpler for us.
On 2020-01-06 09:44, Håvard Flaget Aasen wrote:
I might be way off here, but if you want to skip test you can create a
patch with this above the class or function you want to skip.
--
@pytest.mark.skip(reason="no way of currently testing this")
--
More info here [1]
This will of course skip the tests at build time as well.
I also saw some tests that was marked '@pytest.mark.slow'. These can
also
be skipped using 'pytest -k-slow' though you might skip over more
tests than you want.
I must admit I'm on thin ice here, cause I don't even know how your
script runs..
[1]: https://docs.pytest.org/en/latest/skipping.html#skip
