Consider go1.13 errors package

[ahmetb]

Go has introduced new error wrapping primitives. Especially the errors.Is check can be useful to krew in certain few locations that we rely on error type (ENOENT etc).

https://golang.org/doc/go1.13#error_wrapping

[corneliusweig]

Before doing this, we should think about what implementation details due to the errors API we want to make public (the blog post has more details on this topic).

Also, https://github.com/fatih/errwrap might be helpful.

[erain]

/assign I can help with this.

[erain]

For reference, the blog post @corneliusweig mentioned is https://blog.golang.org/go1.13-errors

[ahmetb]

Thanks. It would be good to see if we get Stack Traces like the current err pkg when run with -v 5 etc

[corneliusweig]

Thanks @erain. I got confused and thought that the link from @.ahmetb was the same 🤦‍♂️

[ferhatelmas]

Also, fatih/errwrap might be helpful.

I offered help to integrate errwrap into golangci-lint. It have it locally, will push upstream accordingly.

[corneliusweig]

@ferhatelmas would that make sense? My understanding is that it should be a deliberate choice to wrap or not to wrap (see blog.golang.org/go1.13-errors). A linter would simply report any such occurrence.

[ferhatelmas]

I think yes because even if it's decided to wrap, people are error-prone and might forget putting %w and might use %s or %q.

Also, it helps in transition with --fix flag of golangci-lint to automatically refactor codebase.

[corneliusweig]

But if I understand correctly, the decision to wrap or not is done on a case-by-case basis, right?

[ferhatelmas]

Yes. Here is a reply to it from the famous blog post.

When adding additional context to an error, either with fmt.Errorf or by implementing a custom type, you need to decide whether the new error should wrap the original. There is no single answer to this question; it depends on the context in which the new error is created.

That's why it should be opt-in linter, false by default.

[ahmetb]

Here's the criteria I expect from this new errors mechanism:

• ability to wrap errors (to add more context/reason) without having to implement a custom type
• preserves syntax while printing the wrapped error messages (reason: reason: root cause)
• preserves ability to print errors normally (as above) or with stack trace when -v≥1 specified (currently achieved wit %+v)
• ability to assess if one of the errors in the wrapped chain is of a certain error instance (e.g. io.EOF) or can be evaluated with methods like os.IsNotExist()

[ahmetb]

Promising new library: https://github.com/rotisserie/eris not sure if compatible with go1.13 errors but seems to have the same capabilities and nearly identical interface.

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

[erain]

/unassign @erain

I am really sorry but I don't have the bandwidth for working on this now.

[corneliusweig]

ping @robertvolkmann. You expressed some interest in this a while back, I think.

[fejta-bot]

Stale issues rot after 30d of inactivity.
Mark the issue as fresh with /remove-lifecycle rotten.
Rotten issues close after an additional 30d of inactivity.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle rotten

[fejta-bot]

Rotten issues close after 30d of inactivity.
Reopen the issue with /reopen.
Mark the issue as fresh with /remove-lifecycle rotten.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/close

[k8s-ci-robot]

@fejta-bot: Closing this issue.

In response to this:

Rotten issues close after 30d of inactivity.
Reopen the issue with /reopen.
Mark the issue as fresh with /remove-lifecycle rotten.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/close

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.