You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Is your feature request related to a problem? Please describe.
Right now, we have DISABLE_ERRORS_LINTERS as a convenient way to provide a list of linters whose findings will not cause MegaLinter to fail a build.
The vision I have is that MegaLinter could run and fail builds if scanners in REPOSITORY_.* produce findings but allow MegaLinter to pass builds that have findings in the other linters.
So, the security-centric REPOSITORY_.* scanners findings' will be errors while everything else has errors disabled. It's the inverse of DISABLE_ERRORS_LINTERS.
Describe the solution you'd like
Example:
DISABLE_ERRORS: trueENABLE_ERRORS_LINTERS:
- REPOSITORY-SEMGREP
- REPOSITORY-GITLEAKS
- [.. rest of the security scanners go here]
(the first line may or may not be implied by the presence of ENABLE_ERRORS_LINTERS)
Describe alternatives you've considered
DISABLE_ERRORS: falseDISABLE_ERRORS_LINTERS:
- MARKDOWN_MARKDOWNLINT
- MARKDOWN_REMARK_LINT
- MARKDOWN_MARKDOWN_LINK_CHECK
- MARKDOWN_MARKDOWN_TABLE_FORMATTER
- [.. every other linter MegaLinter supports except the security scanners]
Additional context
This could mirror the behavior of other variables in the activation / deactivation functionality, but that may be more involved than would be required at an early stage.
I looked through the code and the issues on GitHub but I didn't see what I was looking for, but that's not to say there isn't a fancy, super-excellent way to get from here to there without changing a line of code.
Just an idea. Thank you!
The text was updated successfully, but these errors were encountered:
@nvuillam Sure. I'm thinking that when it's invoked, we handle it as the user specified DISABLE_ERRORS_LINTERS and, instead of setting it to the value the user provided, set it to the difference of what the user provided:
I haven't tested that to see if it would work -- I don't even know if the types would align -- but I think that ought to be sufficient to convey the idea.
@wesley-dean-flexion like for ENABLE_LINTERS and DISABLE_LINTERS, if ENABLE variable is set, then its value is applied , and DISABLE variable is ignored :)
Is your feature request related to a problem? Please describe.
Right now, we have
DISABLE_ERRORS_LINTERS
as a convenient way to provide a list of linters whose findings will not cause MegaLinter to fail a build.The vision I have is that MegaLinter could run and fail builds if scanners in
REPOSITORY_.*
produce findings but allow MegaLinter to pass builds that have findings in the other linters.So, the security-centric
REPOSITORY_.*
scanners findings' will be errors while everything else has errors disabled. It's the inverse ofDISABLE_ERRORS_LINTERS
.Describe the solution you'd like
Example:
(the first line may or may not be implied by the presence of
ENABLE_ERRORS_LINTERS
)Describe alternatives you've considered
Additional context
This could mirror the behavior of other variables in the activation / deactivation functionality, but that may be more involved than would be required at an early stage.
I looked through the code and the issues on GitHub but I didn't see what I was looking for, but that's not to say there isn't a fancy, super-excellent way to get from here to there without changing a line of code.
Just an idea. Thank you!
The text was updated successfully, but these errors were encountered: