feat: add optional logger wherever possible

[ccoVeille]

This commit introduces an optional logger parameter to various structs.
This enhancement allows users to provide custom logging implementations.

Fixes #3558

• Feature Request: ability to provide a logger as an Options setting #3558

Note

This is a follow-up of #3560 that was merged in a branch (not master) and #3618 that was updated by @ndyakov to add changes on it.

The MR is now ready to be merged.

• feat: add optional logger wherever possible #3560
• feat: optional logger #3618

[jit-ci]

❌ The following Jit checks failed to run:

• secret-detection-trufflehog
• static-code-analysis-semgrep-pro

#jit_bypass_commit in this PR to bypass, Jit Admin privileges required.

More info in the Jit platform.

[ccoVeille]

I apparently failed to keep your GPG signature @ndyakov, it sounds logic as I used history rewritting to fix an issue in one of the commit I did earlier. Feel free to pull, sign your commits, and push again. My fork is now all set for maintainers edits

[ndyakov]

Overall looks good to me, can we just make sure this one is addressed (preferably add a simple test):
#3618 (comment)

[ccoVeille]

Overall looks good to me, can we just make sure this one is addressed (preferably add a simple test):
#3618 (comment)

Here the difference I applied to an old commit.

func (cl *LoggerWrapper) Errorf(ctx context.Context, format string, v ...any) { 
         if cl == nil || cl.logger == nil { 
                 legacyLoggerWithLevel.Errorf(ctx, format, v...) 
                 return 
         } 
-         cl.logger.ErrorContext(ctx, format, v...)
+         cl.logger.ErrorContext(cl.printfToStructured(ctx, format, v...))
 }

I can add a test, but adding a test about a logger is never fun.

Maybe it can be done in a separate PR. I'm suggesting that, but if you are OK to wait for me to add the tests a few more days, maybe a week. I can do it.

[ndyakov]

@ccoVeille No problem, I will prepare a beta without the logger, you can write tests and resolve conflicts in the next week and then we will include it.

[ccoVeille]

@ndyakov please review again.

I feel the system is better now.

The only thing I'm unsure is the fact the LoggerWrapper should be exported (like in the current PR) or moved to internal/logging

EDIT: this is needed if a LoggerWrapper is provided directly

[ccoVeille]

Apparently, I broke something.

It's too late here, but the code is close to what is needed.

[ccoVeille]

I feel like I have to move out the change about fixing the panic and sprintf issue

[ccoVeille]

I feel like I have to move out the change about fixing the panic and sprintf issue

Moved to

• fix: bad sprintf syntax #3632
• fix: panic on nil error in handoffWorkerManager closeConnFromRequest #3633

[ccoVeille]

I squashed everything in one commit even yours @ndyakov.

I was done to deal with conflicts when someone was merging something 🙄

[ccoVeille]

The PR is still in DRAFT mode because I have no clear status on what is the minimal Go version for go-redis

As stated in #3631, right now the go.mod mentions Go 1.18+, but the current code can only work with Go 1.23+

• chore: make sure to run the CI on latest Go versions #3631

Except that the code can be reviewed.

But if the minimal version is moved to something at least Go 1.21, my implementation could be simplified a lot as I would be able to delete everything about //go:build tags