41-Feat: sklearn friends

[Gnpd]

This pull request introduces support for custom scikit-learn-compatible estimators in the SklearnSerializer, refactors the internal organization of sklearn-related serializers, and adds comprehensive tests for the new custom estimator logic. The changes improve extensibility, maintainability, and test coverage, allowing users to serialize and deserialize third-party or custom estimators alongside built-in ones.

Support for custom estimators in serialization:

• Added a new mechanism to allow users to provide custom or third-party estimators to SklearnSerializer via a new custom_estimators parameter. These are merged into the serializer's internal registry for the instance, enabling seamless serialization/deserialization of external classes.
• Implemented helper functions is_valid_estimator, normalize_estimators, and load_custom_estimators in openmodels/serializers/sklearn/_custom_estimator.py to validate, normalize, and load custom estimators.
• Updated all usages of the estimator registry in SklearnSerializer to use the instance-specific registry (self._all_estimators) instead of the global one, ensuring custom estimators are recognized during (de)serialization.

Refactoring and reorganization:

• Moved sklearn_serializer.py into a dedicated sklearn subpackage and updated all imports accordingly, improving codebase organization and clarity.

Solves #41

• Updated documentation strings in SklearnSerializer to clarify the new extensibility features and provide developer guidance for third-party package maintainers.

Testing improvements:

• Added a new test suite in test/test_custom_estimator.py covering all branches and edge cases of the custom estimator loading logic, including validation, normalization, error handling, and conflict resolution.

Other updates:

• Minor docstring and version reporting improvements in SklearnSerializer.

Protocol and interface changes:

• Updated the ModelSerializer protocol to allow arbitrary constructor arguments, supporting the new extensibility pattern.

[Copilot]

Pull Request Overview

This pull request refactors the sklearn serializer into a subdirectory structure and adds support for custom estimators. The main changes include reorganizing the codebase to use openmodels.serializers.sklearn.sklearn_serializer instead of the flat openmodels.serializers.sklearn_serializer, implementing a new custom estimator loading mechanism, and updating the sklearn version compatibility.

• Refactored sklearn serializer into a subdirectory structure (sklearn/)
• Added support for custom estimators via a new _custom_estimator module and constructor parameter
• Updated tested sklearn version from "1.7.1" to "1.7.2"

Reviewed Changes

Copilot reviewed 11 out of 12 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
openmodels/serializers/sklearn/sklearn_serializer.py	Added __init__ method with custom_estimators parameter, updated documentation, changed from module-level ALL_ESTIMATORS to instance-level _all_estimators, updated TESTED_VERSIONS and producer_version logic
openmodels/serializers/sklearn/_custom_estimator.py	New module implementing custom estimator validation, normalization, and loading logic
test/test_custom_estimator.py	New comprehensive test suite for the _custom_estimator module
openmodels/serializers/init.py	Updated import path from flat to nested structure
openmodels/protocols.py	Added __init__ method to ModelSerializer protocol to allow serializers to accept constructor parameters
test/*.py	Updated all import statements to use new nested module structure
openmodels/test_helpers.py	Updated import statement to use new nested module structure

Comments suppressed due to low confidence (2)

openmodels/serializers/sklearn/sklearn_serializer.py:305

• The comment references version '1.7.1' but TESTED_VERSIONS was updated to include '1.7.2'. This docstring should be updated to reflect the current tested versions or remove the specific version reference to avoid future inconsistency.
  openmodels/serializers/sklearn/sklearn_serializer.py:225
• This class does not call ModelSerializer.init during initialization. (SklearnSerializer.init may be missing a call to a base class init)

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.