Release Delphi Epidata 4.1.25

[github-actions]

Releasing Delphi Epidata 4.1.25.

• Dependencies:
  • chore: sync main->dev #1490
• Docs:
  • docs: very minor dev doc update #1491
  • Covid act now source links and context around deactivation #1492
  • Put hhs docs under inactive signals #1493
  • Add back .md suffix to ind-combo #1494
• Client:
  • One-time version check #1456
• Other:
  • refactor: use delphi_utils.logger instead of copied file #1488
  • fix: wrong blame commit #1495

[melange396]

CI failure appears to be due to changes in #1456 , but (bizarrely!) that PR passed the tests itself...

[minhkhul]

Per @melange396 @dshemetov suggestions, moving version check in #1456 outside of that class definition.

[dshemetov]

CI is taking a while to start. GH issue?

[dshemetov]

Yep: https://www.githubstatus.com/incidents/m70hk23gx3nx

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarCloud

[melange396]

when i applied equivalent changes from #1497 in this branch (the last two commits above), the test no longer fails... but this still does not explain why the test was previously failing here yet did not fail on the first commit of #1497 : https://github.com/cmu-delphi/delphi-epidata/actions/runs/10009669257/job/27669065083?pr=1497

[melange396]

i think i figured out the root of the problem; explanation copied from slack thread:

The only real difference between the dev branch and the branch created for a release PR are the version numbers that get bumped. The version number in the dev branch exactly matches the version of the client package in pypi, so when the client is imported, [the current implementation of] ._version_check() requests info from PyPI and the if fails and the method exits. This is a success and as intended. When the release action bumps the internal version number, that if will instead pass, and then it uses Epidata.log() to print a message -- but since log() is also a static method being used before the class definition is finalized and it is NOT called using the .__func__() "trick", it fails. Interestingly, the failing test output in GH actions ultimately blames the client file at line 67 for not having Epidata defined. Here is that line in the current dev branch: link ... But the first real problem actually happens when an Exception is raised 6 lines above, when it tries to print the message that the version is out of date! That Exception gets lost because the same problem happens again inside the except: handler block! So a good solution is still is to move the version check outside of the class definition body.

[melange396]

Lets close this PR and recreate the release after we merge #1497