Closed cosenal closed 5 months ago
All modified and coverable lines are covered by tests :white_check_mark:
Project coverage is 98.22%. Comparing base (
ccb9ff6
) to head (c3a9474
).
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
I think you should add the details from How do you detect a 'release' PR?
in your PR description to this section of the docs.
I think you should add the details from
How do you detect a 'release' PR?
in your PR description to this section of the docs.
@purva-thakre Yes, good idea
@purva-thakre comment addressed, PTAL
You can see this in action now! https://github.com/unitaryfund/mitiq/actions/runs/8899188557/job/24437953232?pr=2338 ๐ well.. linkcheck fails, but the change from this PR is working :)
Fixes #2283 as agreed in Mitiq community call on April 26, 2024.
Description
Linkcheck was disabled in #2287, because it was making every pull request (PR) build fail with false errors. Some websites (e.g., doi.org) have bot-preventive measures for which they return 4xx errors when a request is made. In this PR we ignore those websites, on the assumption that we trust their links are not going to break.
Some false errors may still occurr, so we decided to do a linkcheck pass only on 'release' PRs.
How do you detect a 'release' PR?
We check whether the word 'release' is in the PR title (case-insensitive). This will have some false positives (just like this PR itself ๐ ), but it should have no false negatives, see this search on past release PRs. On false positives we just run make linkcheck superfluously, not a big deal.
Bonus
I fixed some linkcheck warning by explicitly replacing links that have been redirected since the time they were added to our docs.
Testing
The new step was tested locally with
where
act_event.json
is a file with content