x/tools/gopls: stylistic diagnostics emitted for generated code

[muirdm]

In particular the staticcheck "ST*" stylistic checks seem to be reporting diagnostics for generated code. I think the staticcheck standalone driver suppresses stylistic errors for generated files.

To reproduce, ensure gopls staticcheck integration is enabled and create a file foo.gen.go:

// Code generated by robot.
// DO NOT EDIT!

package foo

var REALLY_Good_Name = 1

I expect to see no diagnostic for "REALLY_Good_Name" but I get ST1003.

[gopherbot]

Thank you for filing a gopls issue! Please take a look at the Troubleshooting guide, and make sure that you have provided all of the relevant information here.

[dominikh]

The issue here seems to be that your "code generated by" comment doesn't match the agreed upon format, nor any of the explicit exceptions that staticcheck handles. See #13560 and #13560 (comment) in particular.

[muirdm]

I based my comment off of tools/internal/lsp/source/util.go:

// Matches cgo generated comment as well as the proposed standard:
//	https://golang.org/s/generatedcode
var generatedRx = regexp.MustCompile(`// .*DO NOT EDIT\.?`)

This updated repro still reports the stylistic warning:

// Code generated by robot. DO NOT EDIT
package foo
var FOO_Bar = 1

The generated file suppression is in the staticcheck driver, no in the checks themselves, right? I didn't see any code in gopls trying to skip analyses based on generatedness.

[muirdm]

Ah, nevermind. I see the fact now in ReportfFG.

If I add the period at the end // Code generated by robot. DO NOT EDIT. it seems to work. There is some funny caching going on that made transitioning from not-generated to generated not work.

[dominikh]

For people watching at home: AFAIK gopls has no notion of hiding diagnostics in generated files, but the checks in staticcheck use a special analysis (honnef.co/go/tools/facts.Generated) to detect generated files. As such, using staticcheck's analyses in gopls (or any other runner) still benefits from the detection of generated files. Each analysis itself gets to decide whether a diagnostic should be hidden from generated files or not, without the runner's involvement.

[stamblerre]

FWIW, gopls does detect generated files in some cases (for example, if you start to edit the file). We don't hide diagnostics because you would want to know if it doesn't compile, but I guess what you're saying is that we can also rely on staticcheck to make a reasonable decision about if a diagnostic is worth showing in a generated file?

[dominikh]

but I guess what you're saying is that we can also rely on staticcheck to make a reasonable decision about if a diagnostic is worth showing in a generated file?

That is correct.