added standard naming exclusion but...

[turbomam]

StandardNamingConfig.init() got an unexpected keyword argument 'exclude'

[github-actions[bot]]

PR Preview Action v1.6.1 :---: |

:rocket: View preview at
https://GenomicsStandardsConsortium.github.io/mixs/pr-preview/pr-1010/

|

Built to branch gh-pages at 2025-07-01 15:05 UTC.
Preview will be ready when the GitHub Pages deployment is complete.

[mslarae13]

LinkML doesn't yet provide the ability to exclude in the linter. This PR adds that, but can't be merged until LinkML is updated.

[turbomam]

prior to this PR, .linkmllint.yaml

extends: recommended
rules:
  recommended:
    level: error
    exclude_type:
      - enum_definition
  standard_naming:
    level: error
    exclude_type:
      - permissible_value

PR adds

    exclude:
      - HACCP_term
      - IFSAC_category
      - air_PM_concen
      - ferm_pH
      - microb_start_taxID
      - soil_pH
      - spikein_AMR
      - water_pH

proposed final version

extends: recommended
rules:
  standard_naming:
    level: warning
    exclude_type:
      - enum_definition
      - permissible_value
    exclude:
      - HACCP_term
      - IFSAC_category
      - air_PM_concen
      - ferm_pH
      - microb_start_taxID
      - soil_pH
      - spikein_AMR
      - water_pH
  no_empty_title:
    level: warning
    exclude_type:
      - enum_definition
  recommended:
    level: warning
    exclude_type:
      - enum_definition
      - permissible_value

Key Differences

1. Severity: current uses error level, final version uses warning level
2. Scope:
   • current allows enum names to be checked for standard naming (only excludes permissible values)
   • final version excludes both enum names AND permissible values from standard naming
3. Coverage: current only configures 2 rules, final version configures 3 rules (adds no_empty_title)
4. Individual exclusions: final version uses the new exclude feature to ignore specific problematic slot names
5. Philosophy: current is more strict (still checks enum names), final version is more permissive (ignores all enum-related naming issues)