Best practices validator

[rly]

Is your feature request related to a problem? Please describe.
In LinkML there are sometimes multiple equivalent ways of specifying a schema, e.g., due to syntactic sugar. While I think LinkML should allow users to express schema however they want within the language, I think it would be nice to have an additional layer of standardization of schema that follows an opinionated set of best practices, and a validator that checks for that.

For example, the Simple Implicit Scalar Pattern for modeling quantities and units of measurement is marked as an anti-pattern and could be flagged in a best practices validator.

This is inspired by some of our work on linkml arrays: #1887 (comment)
Here:

array:
  minimum_number_dimensions: 3
  maximum_number_dimensions: False
  dimensions:
    - alias: time

is equivalent to:

array:
  minimum_number_dimensions: 3
  maximum_number_dimensions: False
  dimensions:
    - alias: time
    - {}
    - {}

but the second example more explicitly shows that the second and third required dimensions are left unspecified.

array:
  minimum_number_dimensions: 3
  maximum_number_dimensions: 3

is equivalent to:

array:
  exact_number_dimensions: 3

but I prefer the second example because even though min/max is more explicit, it suggests a range. Perhaps that is a controversial example, but I am sure there are other examples of good patterns / best practices to follow.

Related:

• checking for semantic versioning of schema Provide semver conventions for schema developers, plus tools to check conformance #1262

How important is this feature?
• Low - it's an enhancement but not crucial for work

When will use cases depending on this become relevant?
• Long-term - 6 months - 1 year

[cmungall]

Should this be folded into https://linkml.io/linkml/schemas/linter.html?

[rly]

Oh! I totally missed that. I think folding it into the linter would work great.