CI: Improve unit tests to test the bugtool's main() as well

[bernhardkaindl]

This PR raises the code coverage from of the bugtool script itself
from 50% to 84% by running almost all of bugtool from a single test
module by testing the bugtool's main() function.

This allows for full testing of nearly any other fixes and changes:

Add tests/unit/test_main.py which also uses test functions which were
added for earlier tests to test the main function of bugtool to produce
the bugtool output archives for all types: zip, tar, and tar.bz2.

The main test checks the output of the bugtool application to ensure that
it matches the expected output. It compares the captured output with the
expected output and performs various assertions to validate the output.

It extracts the output files from the archive and checks that the xAPI db
and the inventory.xml.

Tested output:

• The start and end of the collected bugtool messages from captured.out
• The inventory.xml file is checked to be validated using its XML schema.
• The xapi-db.xml is checked to have secrets filtered using a dummy xapi db

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (1aae840) 56.52% compared to head (0783b2f) 84.90%.
Report is 2 commits behind head on master.

Additional details and impacted files

@@             Coverage Diff             @@
##           master      #44       +/-   ##
===========================================
+ Coverage   56.52%   84.90%   +28.37%     
===========================================
  Files           8       15        +7     
  Lines        1647     1928      +281     
===========================================
+ Hits          931     1637      +706     
+ Misses        716      291      -425     

Flag	Coverage Δ
python2.7	83.40% <93.79%> (?)
python3.10.13	83.74% <100.00%> (?)
unittest	83.74% <100.00%> (+27.21%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

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

[psafont]

While there's quite a bit of code being introduced, it's just in the tests and not on the deployed code.

This, along with the coverage report that ensure the new code is run, makes the PR low-risk

[bernhardkaindl]

@psafont Thank you for your review and approval!

While there's quite a bit of code being introduced, it's just in the tests and not on the deployed code.

This, along with the coverage report that ensure the new code is run, makes the PR low-risk

Exactly, fully agreed!

I just pushed the final minor fixes of the GitHub Workflow file .github/workflows/main.yml!

Could you review the changes? They are:

• For the pytest-coverage-comment steps:

  • Ignore errors on missing permissions to make the coverage comment.
  • This can happen when PRs are opened in private repos (we need to update the GH perms for the repo)

• For the Codecov coverage upload steps:

  • Update the tag for the Python3 version to be format("python%s", python-version) not just unittest.
    • This shows us the python3 version the coverage has been generated with on https://app.codecov.io/gh/xenserver/status-report (besides python2.7 for the coverage from Python 2.7)