Closed bryanwweber closed 6 years ago
@bryanwweber this looks fine to me, but the AppVeyor build failed for some reason...
Looks like it failed because one of the expected ResourceWarnings
didn't show up. We might need to loosen the criteria on that test... or perhaps figure out a way to fix Habanero so it doesn't do that (I think its Habanero, anyways). I restarted the build, let's see what happens
Well, I don't know why the tests are failing, but it's not actually a problem with the code itself, its a problem with the test. Do you want to merge, or do you want me to try to fix it?
I wonder if this is another package version issue?
We probably need to try and fix this, because I don't think we'll have an updated Windows package until we do.
Welp, I don't think it's a habanero version issue, because it looks like 0.6.0 in both the Windows and Linux test logs.
I wonder if it's an issue with pytest, not habanero...
Here's the first failing test:
____________________ TestGetReference.test_valid_reference ____________________
self = <pyked.tests.test_converters.TestGetReference object at 0x000000604E661B38>
def test_valid_reference(self):
"""Ensure valid reference reads properly.
"""
root = etree.Element('experiment')
ref = etree.SubElement(root, 'bibliographyLink')
ref.set('doi', '10.1016/j.ijhydene.2007.04.008')
ref.set('preferredKey', 'Chaumeix, N., Pichon, S., Lafosse, F., Paillard, C.-E., '
'International Journal of Hydrogen Energy, 2007, (32) 2216-2226, '
'Fig. 12., right, open diamond'
)
with pytest.warns(UserWarning) as w:
ref = get_reference(root)
> assert w[1].message.args[0] == ('Using DOI to obtain reference information, '
'rather than preferredKey.'
)
pyked\tests\test_converters.py:145:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
self = WarningsChecker(record=True), i = 1
def __getitem__(self, i):
"""Get a recorded warning by index."""
> return self._list[i]
E IndexError: list index out of range
C:\Miniconda36-x64\envs\py3.5\lib\site-packages\_pytest\recwarn.py:139: IndexError
That last bit is kinda weird, no?
Alright, I'll try to dig into this over the holiday.
Merging #99 into master will not change coverage. The diff coverage is
n/a
.
@@ Coverage Diff @@
## master #99 +/- ##
=====================================
Coverage 100% 100%
=====================================
Files 4 4
Lines 879 880 +1
Branches 212 212
=====================================
+ Hits 879 880 +1
Impacted Files | Coverage Δ | |
---|---|---|
pyked/tests/__init__.py | 100% <0%> (ø) |
:arrow_up: |
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update 6b214c0...4ee30e4. Read the comment docs.
The issue is with Habanero, and more specifically, that they use the Requests library which leads to a bunch of ResourceWarnings (see https://github.com/requests/requests/issues/1882). Apparently (based on these results), how all that gets handled is different on Windows vs. *nix. Anyhow, I've updated how we handle the warnings in the test suite so it should hopefully be cross platform now.
All passed, LGTM.
CHANGELOG.md
Changes proposed in this pull request:
__version__
attribute from the_version
module at the top-level of the package@pr-omethe-us/chemked