Running doctests using doctest-plus

[prady0t]

Description

Running doctests on /pybamm using pytest's doctest-plus plugin. We decided to move ahead with doctestplus as xdoctest was throwing random internal errors. See discussion here.

Fixes #4031
Also addresses #3617

[prady0t]

Still working on latexify error.

[agriyakhetarpal]

I think it makes sense to skip the failing doctest. The functionality is being tested in the unit tests, and testing through a doctest whether a PNG or the LaTeX form of an equation can be printed out or not is quite inconvenient. Please see https://github.com/scientific-python/pytest-doctestplus#skipping-tests for instructions.

[prady0t]

I think it makes sense to skip the failing doctest. The functionality is being tested in the unit tests, and testing through a doctest whether a PNG or the LaTeX form of an equation can be printed out or not is quite inconvenient. Please see https://github.com/scientific-python/pytest-doctestplus#skipping-tests for instructions.

Makes sense. Skipping these tests.

[agriyakhetarpal]

The doctests failed because of the now redundant .. doctest:: directive that Sphinx won't recognise without sphinx.ext.doctest. The Windows build is unrelated, I triggered a re-run for it.

[agriyakhetarpal]

Looks good to me! I added a few suggestions plus a fix for the failing doctest. Thanks, @prady0t!

[agriyakhetarpal]

Looks like there's another one of the .. doctest:: directives below the one that we removed – could you fix that, @prady0t? This should be good to merge after that.

[prady0t]

Done!

[prady0t]

Test failure due to non-matching hashes happens quite frequently. Why does it happen? Is there a way to fix it?

Downloading types_python_dateutil-2.9.0.20240316-py3-none-any.whl (9.7 kB)
ERROR: THESE PACKAGES DO NOT MATCH THE HASHES FROM THE REQUIREMENTS FILE. If you have updated the package versions, please update the hashes. Otherwise, examine the package contents carefully; someone may have tampered with them.
    unknown package:
        Expected sha256 d37f1d8cab6fca11d8ba44d6404fdbb9b796d4e6f447d12317e4ce0660310072
             Got        f6974e35bc2c42ad0002e80e04815ae8f65a012154c0de8b4daf00c9417b8bb4

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 99.56%. Comparing base (b4f8380) to head (c20d315).
Report is 218 commits behind head on develop.

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #4117      +/-   ##
===========================================
+ Coverage    99.55%   99.56%   +0.01%     
===========================================
  Files          288      288              
  Lines        21790    21790              
===========================================
+ Hits         21693    21696       +3     
+ Misses          97       94       -3     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[agriyakhetarpal]

Test failure due to non-matching hashes happens quite frequently. Why does it happen? Is there a way to fix it?

This is a bug with pip's hash-checking functionality. I assume there are three intermittent factors that can lead to the error:

1. pip either computes the SHA-256 checksum of the cached types_python_dateutil package incorrectly
2. pip either receives an incorrect SHA-256 checksum and that for the cached types_python_dateutil is correct
3. The GitHub Actions cache mechanisms are corrupted or that they corrupt this file (a bit unlikely).

Either way, a re-run typically fixes the issue, but it's strange that it happens with types_python_dateutil everytime I see it.

[agriyakhetarpal]

Thanks for your work, @prady0t!

[prady0t]

Test failure due to non-matching hashes happens quite frequently. Why does it happen? Is there a way to fix it?

This is a bug with pip's hash-checking functionality. I assume there are three intermittent factors that can lead to the error:

1. pip either computes the SHA-256 checksum of the cached types_python_dateutil package incorrectly
2. pip either receives an incorrect SHA-256 checksum and that for the cached types_python_dateutil is correct
3. The GitHub Actions cache mechanisms are corrupted or that they corrupt this file (a bit unlikely).

Either way, a re-run typically fixes the issue, but it's strange that it happens with types_python_dateutil everytime I see it.

I see 🤔 So we can't do anything about it for the moment.