Prevent hang in git cat-file if repository is not a valid repository and other fixes

[zeripath]

This PR contains multiple fixes. The most important of which is:

• Prevent hang in git cat-file if the repository is not a valid repository

  Unfortunately it appears that if git cat-file is run in an invalid
  repository it will hang until stdin is closed. This will result in
  deadlocked /pulls pages and dangling git cat-file calls if a broken
  repository is tried to be reviewed or pulls exists for a broken
  repository.

  Fix It takes a long time to open a big pull request page #14734
  Fix iowait and memory consumption are skyrocketing regularly #9271
  Fix Task running since 2 weeks #16113

Otherwise there are a few small other fixes included which this PR was initially intending to fix:

• Fix panic on partial compares due to missing PullRequestWorkInProgressPrefixes
• Fix links on pulls pages due to regression from Multiple Escaping Improvements #17551 - by making most /issues routes match /pulls too - Fix Various actions on pull requests are broken #17983
• Fix links on feeds pages due to another regression from Multiple Escaping Improvements #17551 but also fix issue with syncing tags - Fix Clicking on a feature branch name from a dashboard entry directs to the wrong url (404) #17943
• Add missing locale entries for oauth group claims
• Prevent NPEs if ColorFormat is called on nil users, repos or teams.

[KN4CK3R]

Issue/Pull links are still not working in the news feed:

[codecov-commenter]

Codecov Report

Merging #17991 (f5c33da) into main (6e7d28c) will decrease coverage by 0.02%.
The diff coverage is 24.67%.

@@            Coverage Diff             @@
##             main   #17991      +/-   ##
==========================================
- Coverage   45.28%   45.26%   -0.03%     
==========================================
  Files         820      820              
  Lines       90973    91046      +73     
==========================================
+ Hits        41196    41209      +13     
- Misses      43219    43274      +55     
- Partials     6558     6563       +5     

Impacted Files	Coverage Δ
models/action.go	49.62% <0.00%> (-1.96%)	⬇️
models/org_team.go	57.50% <0.00%> (-0.74%)	⬇️
models/repo/repo.go	66.02% <0.00%> (-1.12%)	⬇️
models/user/user.go	61.36% <0.00%> (-0.59%)	⬇️
modules/git/repo_base_nogogit.go	80.00% <0.00%> (-6.85%)	⬇️
modules/indexer/code/bleve.go	66.52% <0.00%> (-0.84%)	⬇️
modules/indexer/code/elastic_search.go	1.61% <0.00%> (-0.02%)	⬇️
services/pull/pull.go	42.10% <0.00%> (-0.09%)	⬇️
modules/git/repo_commit_nogogit.go	56.97% <33.33%> (-1.36%)	⬇️
models/unittest/testdb.go	18.18% <38.46%> (+7.22%)	⬆️
... and 11 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 6e7d28c...f5c33da. Read the comment docs.

[Gusted]

Weird behavior for cat-file to not fail (almost sounds like a bug). Let's hope that other commands won't show this behavior as well.

[ryanfitzsimon]

Hi @zeripath, could you please explain how this resolves #14734?

I can't see how these changes will help in cases of slow performance, apart from in specific cases where hangs are occurring (which isn't what was reported).
The original report described PR opening being slow, but it eventually succeeded after nearly 2 minutes.

I have a repository that suffers from the performance problems, taking minutes to open PRs with large change counts.
I've just tried running git cat-file --batch in the bare repo on the Gitea host. I also ran a git fsck. Both ran without issue.

[ryanfitzsimon]

Also, what is meant by "not a valid repository"? Directories that aren't part of a repository (no .git directory)? Or repositories that have been somehow corrupted?

The former doesn't seem to cause a hang for me:

gitserver:~$ git version
git version 2.31.1
gitserver:~$ mkdir not-a-repo
gitserver:~$ cd not-a-repo/
gitserver::~/not-a-repo$ git cat-file --batch
fatal: not a git repository (or any of the parent directories): .git

If it's specific to a certain type of corrupt repository, I'm not sure how to test it.

[lunny]

@ryanfitzsimon This PR should resolve part of the hang, so I have reopened #14734

[zeripath]

Ok. I had thought that we had finally found the reason for this problem but clearly you disagree. Could you update the issue and report back if things have changed on main at all?