Skip abandoned repos

[fthomas]

This implements the idea proposed in #3554 to stop creating wasted PRs by skipping repos whose last commit is older than a configured threshold. The PR contains these changes:

• a new GitAlg.getCommitDate which returns the commit date as Timestamp of a given SHA-1
• a new RepoCache.commitDate: Timestamp field that contains the commit date that corresponds to RepoCache.sha1
• a new RepoConfig.lastCommitMaxAge: Option[FiniteDuration] field that contains the threshold for deciding if a repo should be skipped. A repo is skipped, if the last commit is greater than this FiniteDuration.
• a default lastCommitMaxAge is added to the default Scala Steward config. Other Scala Steward instances and every repo can override this with a greater value. It is not possible to unset it or override it with a smaller value.
• RepoCache.checkCache is modified such that after retrieving a fresh cache, the commit date is compared to lastCommitMaxAge and an error is raised eventually

Notes:

• The addition of a non optional field to RepoCache means that reading an old cache will log an error like this:

2025-01-20 21:16:44,985 ERROR Failed to parse or decode JSON from workspace/store/repo_cache/v1/github/scala-steward-org/test-repo-1/repo_cache.json
io.circe.DecodingFailure$DecodingFailureImpl: DecodingFailure at .commitDate: Missing required field

Fortunately this is only a log message so that Scala Steward will recompute the RepoCache and proceed as normal then. Adding this field is therefore effectively invalidating all RepoCaches out there.

• A skipped repo looks like this in the log:

2025-01-21 08:35:25,033 INFO  ──────────── Steward scala-steward-org/test-repo-1:scala-cli-test ────────────
2025-01-21 08:35:25,033 INFO  Check cache of scala-steward-org/test-repo-1:scala-cli-test
2025-01-21 08:35:25,738 ERROR Steward scala-steward-org/test-repo-1:scala-cli-test failed
org.scalasteward.core.repocache.RepoCacheAlg$$anon$1: Skipping because last commit is older than 2d
2025-01-21 08:35:25,738 INFO  ──────────── Total time: Steward scala-steward-org/test-repo-1:scala-cli-test: 705ms ────────────

lastCommitMaxAge was set to 2 days in this case.

Closes: #3554

[codecov]

Codecov Report

Attention: Patch coverage is 88.88889% with 3 lines in your changes missing coverage. Please review.

Project coverage is 89.84%. Comparing base (f5a5a09) to head (fb06636).

Files with missing lines	Patch %	Lines
...org/scalasteward/core/repocache/RepoCacheAlg.scala	75.00%	3 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #3559      +/-   ##
==========================================
+ Coverage   89.81%   89.84%   +0.02%     
==========================================
  Files         171      171              
  Lines        4988     5011      +23     
  Branches      495      494       -1     
==========================================
+ Hits         4480     4502      +22     
- Misses        508      509       +1     

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

[mzuehlke]

Only a small change to add such a valuable feature 👍
Not sure where, but would be good to add a note about this feature.
Maybe a short above the value in default.scala-steward.conf ?

And 18 month of inactivity ? I think I would have choosen at most 12 month. But I have no clue if this would be too short.

[mzuehlke]

Just checked, the public instance check uses 24 month: https://github.com/VirtusLab/scala-steward-repos/blob/main/scripts/test-repos.scala#L32

[exoego]

This is awesome.

I feel 540 days is very tolerant but fine.
It is common not to touch repo in maintenance-mode for months in work😌

Dependabot: 90 days
https://docs.github.com/en/code-security/dependabot/working-with-dependabot/managing-pull-requests-for-dependency-updates