Skip to main content
Version: 3.x

Running checks and autofix

Running checks​

Run all enabled checks:

rev-dep config run

List every issue instead of the default truncated output:

rev-dep config run --list-all-issues

Can also run specific workspaces:

rev-dep config run --workspaces packages/app,packages/shared

Also lint the config for dead glob patterns while running (adds almost no time - it reuses the run's graph):

rev-dep config run --lint-config

This fails the run on any lint error and prints only the counts. See Linting the config for details.

Output formats​

Information about the supported output formats can be found in Output formats.

Exit code​

  • Exit code 0 when all enabled checks pass
  • Exit code 1 when violations exist

Using AutoFix​

Some checks can automatically apply fixes when both of these are true:

  • the check supports autofix
  • you run rev-dep config run --fix

Fixable detectors​

  • importConventions
  • unusedExportsDetection
  • orphanFilesDetection

Note that autofix: true must be set in the config for the detector be fixable.

Exit code nuance with --fix and --recheck​

Using --fix flag adjust the source code, but do not perform validation after the fixes are applied.

This means it will still exit with code 1.

To perform a re-check after applying fixes, you can use the --recheck flag in combination with --fix:

rev-dep config run --fix --recheck

Additional pass will be performed to check if all issues are resolved.

Autofixing orphan files issues​

Note that autofixing orphan files issues removes the orphan file.

Removing one orphan file can cause another file to become orphan.

Running rev-dep config run --fix several times might be needed to remove all orphan files.

Updating the duplicated-code baseline​

--fix never touches duplicated code - there is no edit a tool can safely make for you. The other thing config run can write is the snapshot: the committed file that records which duplications are already acknowledged.

rev-dep config run --update-snapshot

This rewrites every duplicatedCodeDetection that has a snapshotPath, so whatever this run found becomes the new baseline. Detections without one are unaffected.

Use it twice: once when adopting the check on a codebase that already has duplication, and again after you resolve some of it, so the baseline follows the code. See duplicated code.