serialization to json

[ymahlau]

resolves #6

Summary by CodeRabbit

Release Notes

• New Features

  • Added serialization functionality to convert dataclass instances to and from JSON-compatible formats.
  • Enhanced attribute-setting methods with new parameters for controlling callback execution and modifying private fields.

• Behavior Changes

  • Slot-based dataclass configuration is no longer supported.

[coderabbitai]

Caution

Review failed

Pull request was closed or merged during review

📝 Walkthrough

Walkthrough

The PR adds serialization capabilities (to_dict, from_dict, to_json, from_json) to convert dataclass instances and JAX/NumPy arrays to/from JSON, enhances DataClass.aset with callback and private field control, disables slots-based dataclass support, and includes comprehensive test coverage for all new functionality.

Changes

Cohort / File(s)	Summary
Serialization Implementation src/drinx/serialize.py	New module with recursive encode/decode logic for primitives, tuples, dicts, JAX/NumPy arrays (with dtype/shape metadata), and dataclass instances; includes validation for field access and module imports during deserialization.
Public API Export src/drinx/__init__.py	Exports serialization helpers (to_dict, from_dict, to_json, from_json) in __all__ for direct import from drinx.
DataClass Core Enhancements src/drinx/base.py	Adds allow_private, bypass_callbacks to aset; introduces create_new_ok, bypass_callbacks to aset_inplace; adds _copy_and_set static helper; replaces dataclasses.replace logic; forces slots=False, weakref_slot=False in subclass decorator.
Decorator & Transform Updates src/drinx/transform.py	Disables slots support by explicitly forcing slots=False, weakref_slot=False; adds AttributeError handling in flatten_with_keys with clearer error messages.
Visualization Minor Fix src/drinx/visualize.py	Simplifies dtype/shape prefix formatting by using single-line generator expression in f-string.
Test Coverage for Serialization tests/test_serialize.py	Comprehensive test suite (+760 lines) validating round-trip serialization: primitive types, tuple/dict tagging, JAX/NumPy arrays with dtype/shape, dataclass metadata, nested structures, error cases, and public API exposure.
Enhanced aset Tests tests/test_base.py	Adds +481 lines of test coverage for allow_private, bypass_callbacks, create_new_ok parameters; validates callback execution control, private field mutations, and regression cases for unset fields.
Slots Test Removal tests/test_dataclass.py	Removes 9 lines of @dataclass(slots=True) test coverage from TestSlots, retaining only default (non-slots) validation.

Sequence Diagram(s)

sequenceDiagram
    participant Client
    participant PublicAPI as Public API<br/>(to_dict/from_dict)
    participant Serializer as Serializer<br/>(_serialize)
    participant Validator as Validator<br/>(Module Import)
    participant Deserializer as Deserializer<br/>(_deserialize)
    participant Constructor as DataClass<br/>Constructor

    Client->>PublicAPI: to_dict(obj)
    PublicAPI->>Serializer: _serialize(obj)
    Serializer->>Serializer: Check type (primitive/array/dataclass)
    alt JAX/NumPy Array
        Serializer->>Serializer: Extract dtype, shape, data
        Serializer-->>PublicAPI: {__jax__/numpy, dtype, shape, data}
    else Dataclass Instance
        Serializer->>Serializer: Get __module__, __name__
        Serializer->>Serializer: Filter init=True fields
        Serializer->>Serializer: Recursively serialize values
        Serializer-->>PublicAPI: {__module__, __name__, __value__}
    else Primitive/Collection
        Serializer-->>PublicAPI: JSON-safe structure
    end
    PublicAPI-->>Client: dict

    Client->>PublicAPI: from_dict(data)
    PublicAPI->>Validator: Validate markers & structure
    Validator->>Validator: Import module by name
    Validator->>Validator: Locate dataclass type
    PublicAPI->>Deserializer: _deserialize(data)
    Deserializer->>Deserializer: Reconstruct nested objects
    Deserializer->>Constructor: Instantiate with fields
    Constructor-->>Deserializer: instance
    Deserializer-->>PublicAPI: reconstructed object
    PublicAPI-->>Client: obj

Loading

Estimated code review effort

🎯 4 (Complex) | ⏱️ ~45 minutes

Possibly related PRs

• aset for private attributes #11: Makes identical DataClass modifications (allow_private/bypass_callbacks additions to aset/aset_inplace, _copy_and_set helper, forced slots=False/weakref_slot=False), suggesting overlapping feature development.

Poem

🐰 Hop into serialization's glow,
DataClasses now have callbacks to flow!
Arrays and dicts, all tagged with care,
Slots be gone—simplicity everywhere! 🥕

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 10.87% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly identifies the main change: adding JSON serialization functionality to the drinx package.
Linked Issues check	✅ Passed	The PR fulfills all requirements from issue #6: adds serialization helper functions (to_dict, from_dict, to_json, from_json) and supports serialization of both static and non-static/private dataclass fields.
Out of Scope Changes check	✅ Passed	All changes align with the serialization objective. Supporting updates to base.py and transform.py are necessary for serialization functionality.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[codecov-commenter]

Codecov Report

❌ Patch coverage is 91.09589% with 13 lines in your changes missing coverage. Please review.
✅ Project coverage is 87.90%. Comparing base (fa3bc32) to head (211a070).
⚠️ Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
src/drinx/base.py	81.25%	5 Missing and 4 partials ⚠️
src/drinx/transform.py	75.00%	4 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #12      +/-   ##
==========================================
+ Coverage   86.47%   87.90%   +1.43%     
==========================================
  Files           6        7       +1     
  Lines         451      587     +136     
  Branches       96      137      +41     
==========================================
+ Hits          390      516     +126     
- Misses         39       46       +7     
- Partials       22       25       +3     

☔ 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.