jsonify-objects

Recursively convert almost any Python object to a JSON-serializable structure.

Installation

pip install jsonify-objects

Usage

dumps

Drop-in wrapper around json.dumps that handles non-serializable types.

from jsonify_objects import dumps

dumps({"a": 1, "b": True})
# '{\n  "a": "1",\n  "b": "true"\n}'

jsonify_objects

Returns the converted structure without serializing to a string.

from jsonify_objects import jsonify_objects

jsonify_objects({"a": 1, "b": True})
# {"a": "1", "b": "true"}

Behavior

Primitives

Primitive values inside plain dicts and lists are converted to strings.

dumps({"x": 42, "flag": True, "color": Color.RED})
# {"x": "42", "flag": "true", "color": "red"}

Dataclasses and objects

By default, non-serializable objects fall back to repr.

dumps({"user": UserDataclass("Alice", 30)})
# {"user": "UserDataclass(name='Alice', age=30)"}

Pass serialize_objects=True to expand objects and dataclasses into dicts. Numeric fields on expanded objects preserve their native JSON types.

dumps({"user": UserDataclass("Alice", 30)}, serialize_objects=True)
# {"user": {"name": "Alice", "age": 30}}

Nested objects are expanded recursively.

dumps({"person": Person("Bob", Address("NYC"))}, serialize_objects=True)
# {"person": {"name": "Bob", "address": {"city": "NYC"}}}

Objects without a __dict__ (e.g. slots-only classes) fall back to repr.

indent

dumps accepts an indent parameter (default 2) passed through to json.dumps.

dumps({"a": 1}, indent=4)