chore(deps): update github.com/urfave/cli to 3.8.0

[cuixq]

Unblocks #2674

For github.com/urfave/cli/v3 version 3.8.0, it seems an empty flag is no longer treated as an error but instead an empty flag will be set, see urfave/cli#2297.

Thus this PR updated the expected exit codes and error messages for test cases involving empty flags.

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 79.72%. Comparing base (8ddcc85) to head (8a31a0b).

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2726      +/-   ##
==========================================
- Coverage   79.77%   79.72%   -0.05%     
==========================================
  Files         117      117              
  Lines        8016     8016              
==========================================
- Hits         6395     6391       -4     
- Misses       1246     1248       +2     
- Partials      375      377       +2     

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[G-Rath]

I don't think this is correct - the value being empty should not cause it to fallback to the default

what I would expect to happen is plugins ends up with an empty string, which we should consider an error

[another-rex]

There's a separate test with --no-default that does what you describe, and stays that way.

[G-Rath]

relating to my inline comment, could you check if this changes what happens with --config=? if it means we're using the default instead of erroring (one way or another), then frankly I'd say the upstream is wrong in their behaviour because it feels dangerous to fallback to the default.

overall this change has some weird potential implications, so I think we should dig a bit deeper to confirm what the actual result is

[another-rex]

I'm not sure that's what happens though? An empty plugins string with out experimental-no-default-plugins should just use the default plugins right?

[another-rex]

Pushed similar changes to #2674 directly.