JSON in Python

Python ↔ JSON Data Type Conversion

The following table shows how Python data types are converted to their JSON equivalents during serialization and deserialization:

remember this table.

---

This repository contains small examples demonstrating how to work with JSON in Python:

• json_and_pythonObj.py: How to serialize (encode) Python built-in types to JSON and deserialize (decode) JSON into Python built-ins using json.dumps, json.dump, json.loads, and json.load.
• json_and_customObj.py: How to serialize custom Python objects by providing a default function or by subclassing json.JSONEncoder, and how to decode JSON back into custom objects using object_hook.
• example_json_format.json: A sample JSON file showing a nested object with arrays and booleans.
• person.json: A sample JSON file used by json_and_pythonObj.py.

Concepts

1. Serialization (Encoding)

   • Converting Python objects to JSON format.
   • Use json.dumps(obj) to get a JSON string.
   • Use json.dump(obj, file) to write JSON directly to a file.
   • Both functions accept formatting parameters like indent=4 for pretty printing.

2. Deserialization (Decoding)

   • Converting JSON strings/files back into Python objects.
   • Use json.loads(json_string) to parse a JSON string into Python built-ins (usually dict, list, str, int, float, bool, None).
   • Use json.load(file) to parse JSON from a file.

3. Built-in types vs custom objects

   • The json module can serialize built in Python types directly (dicts, lists, strings, numbers, booleans, None).
   • Custom classes (instances of user-defined classes) are not serializable by default. Attempting to json.dumps() them raises a TypeError.

4. Making custom objects serializable

   • Option A: Provide a default callable to json.dumps that returns a JSON serializable representation (eg. a dict).

   • Option B: Subclass json.JSONEncoder and override default(self, obj) to return a serializable representation.

5. Decoding custom objects

   • json.loads(..., object_hook=...) accepts a callable that transforms dicts into custom objects. If the object_hook returns an instance, that instance will replace the dict.

6. Pitfall: shadowing the standard library

   • One thing I ran into while writing these examples: if you name a script after a standard library module (for example json.py, os.py, sys.py, requests.py), Python will import your local file first and shadow the real module. That can lead to confusing errors like AttributeError: module 'json' has no attribute 'dumps', RecursionError, or other import failures, I hit this while developing the examples, so watch out for it.

   • How to fix

     • Don’t name your script json.py (or any other stdlib/module name). Use something like json_example.py instead.
     • Remove cached bytecode: delete the __pycache__ folder and any *.pyc files.
     • Restart your Python interpreter or editor/IDE so the old module is not still loaded.
     • Check the current working directory for accidentally named files, Python resolves imports from there first.

Files and How to Run

Open a terminal (PowerShell on Windows) in the repository root (c:\JSON_in_python).

Run examples:

python .\json_and_pythonObj.py
python .\json_and_customObj.py

Expected behavior:

• json_and_pythonObj.py prints a pretty JSON string for person and writes/reads person.json.
• json_and_customObj.py prints the JSON representation of a Person instance and shows decoding back into a Person object using object_hook.

License

This is my example code for learning purposes. Use freely.