Lint makes it easy to run linters from Go code. This allows lint checks to be part of a regular go build
+ go test
workflow. False positives are easily ignored and linters are automatically integrated into CI pipelines without any extra effort. Check the project website to learn more about how it can be useful.
Download using
go get -t github.com/surullabs/lint
Run the default linters by adding a new test at the top level of your repository
func TestLint(t *testing.T) {
// Run default linters
err := lint.Default.Check("./...")
// Ignore lint errors from auto-generated files
err = lint.Skip(err, lint.RegexpMatch(`_string\.go`, `\.pb\.go`))
if err != nil {
t.Fatal("lint failures: %v", err)
}
}
lint
runs linters using the excellent os/exec
package. It searches all Go binary directories for the needed binaries and when they don't exist it downloads them using go get
. Errors generated by running linters are split by newline and can be skipped as needed.
gofmt
- Rungofmt -d
and report any differences as errorsgovet
- Rungo tool vet -shadow
golint
- https://github.com/golang/lintgosimple
- Code simplificationgostaticcheck
- Verify function argumentserrcheck
- Find ignored errors
Gometalinter runs a number of linters concurrently. It also vendors each of these and uses the vendored versions automatically. A vendored version of gometalinter
is included and can be used in the following manner. Please note that not all linters used by gometalinter have been tested.
import (
"testing"
"github.com/surullabs/lint/gometalinter"
)
func TestLint(t *testing.T) {
// Run default linters
metalinter := gometalinter.Check{
Args: []string{
// Arguments to gometalinter. Do not include the package names here.
},
}
if err := metalinter.Check("./..."); err != nil {
t.Fatal("lint failures: %v", err)
}
}
varcheck
- Detect unused variables and constantsstructcheck
- Detect unused struct fieldsaligncheck
- Detect suboptimal struct alignmentdupl
- Detect duplicated code
There are a number of excellent linters available for Go and Lint makes it easy to run them from tests. While building our mobile calendar app TimeFerret, (which is built primarily in Go), including scripts that run linters as part of every repository grew tiresome very soon. Using lint
to create tests that ran on each commit made the codebase much more stable, since any unneeded false positives were easily skipped. The main advantages of using lint
over running tools manually is:
- Skip false positives explicitly in your tests - This makes it easy to run only needed checks.
- Enforce linter usage with no overhead - No special build scripts are needed to install linters on each developer machine as they are automatically downloaded.
- Simple CI integration - Since linters are run as part of tests, there are no extra steps needed to integrate them into your CI pipeline.
Please check if you can use the gometalinter package first. If not, adding a new linter is made dead simple by the github.com/surullabs/lint/checkers
package. The entire source for the golint
integration is
import "github.com/surullabs/lint/checkers"
type Check struct {
}
func (Check) Check(pkgs ...string) error {
return checkers.Lint("golint", "", github.com/golang/lint/golint", pkgs)
}
The github.com/surullabs/lint/testutil
package contains utilities for testing custom linters.
You can also take a look at this CL which adds varcheck
for an example of how to add a linter.
If you'd like to vendor the linter source, please use the same method as the gometalinter
package.
Lint is available under the Apache License. See the LICENSE file for details.
Pull requests are always welcome! Please ensure any changes you send have an accompanying test case.