chore: Use attr.define instead of attr.s

[kacpermuda]

Problem

When facets are inheriting from BaseFacet V2, that is using @attr.define, but the child facet definition is made using @attr.s, after copying the child facet the attributes are not present in the copy. This can be an issue, as in the TransformTransport, we are copying the event before transformation. I've looked through the codebase, and only some DBT and Airflow Facets are defined in this mixed way. It's probably an issue about the slots and get/set state auto generation in attrs.

Solution

There is no easier solution that simply not mixing the attr decorators. I've tried also copying the object in a generic way, that would handle that mix, but ended up in rabbit hole, creating more and more complex solution. I think the good approach here is to keep the facets that we own tidy and not mixed up and not try to solve all the possible combinations. I think also the TransformTransport is the only one using copy, so the impact is relatively low. Copying works fine for v1 classes and v2 classes, it's only the mixing on inheritance that's causing the problem, so there is no need to clean up all the classes or use attr.define everywhere for now.

EDIT: After review, we replaced @attr.s with @attr.define in the entire codebase and added a pre-commit to check that.

One-line summary:

chore: Use attr.define instead of attr.s

Checklist

• You've signed-off your work
• Your pull request title follows our guidelines
• Your changes are accompanied by tests (if relevant)
• Your change contains a small diff and is self-contained
• You've updated any relevant documentation (if relevant)
• Your comment includes a one-liner for the changelog about the specific purpose of the change (not required for changes to tests, docs, or CI config)
• You've versioned the core OpenLineage model or facets according to SchemaVer (if relevant)
• You've added a header to source files (if relevant)

---

SPDX-License-Identifier: Apache-2.0
Copyright 2018-2025 contributors to the OpenLineage project

[mobuchowski]

@kacpermuda should we just move to attr.define and forbid creating the short form - maybe by simple grep, maybe by some custom rule?

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 85.66%. Comparing base (e8bd029) to head (5fae7e7).

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #3776      +/-   ##
==========================================
+ Coverage   85.63%   85.66%   +0.02%     
==========================================
  Files          57       57              
  Lines        3689     3689              
==========================================
+ Hits         3159     3160       +1     
+ Misses        530      529       -1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[kacpermuda]

Yeah, I can probably make the switch in the entire codebase and then make a pre-commit that rejects the presence of @attr.s or @s. Sounds good?