Skip to main content

A lightweight library to convert between TOON (Token-Oriented Object Notation) and popular data formats (JSON, YAML, XML, CSV).

Project description

🚀 TOON Converter (Python)

A lightweight library to convert between TOON (Token-Oriented Object Notation) and popular data formats (JSON, YAML, XML, CSV).

Reduce your LLM token costs by up to 40% using the TOON format!

📦 Installation

pip install toon-parse

🚀 Quick Start

Basic Usage (Synchronous)

from toon_parse import ToonConverter

# 1. Python Object to TOON
data = {"name": "Alice", "age": 30, "active": True}
toon_string = ToonConverter.from_json(data)
print(toon_string)
# Output:
# name: "Alice"
# age: 30
# active: true

# 2. TOON to Python Object
json_output = ToonConverter.to_json(toon_string)
print(json_output)
# Output: {'name': 'Alice', 'age': 30, 'active': True}

Mixed Text Support

The library can automatically extract and convert JSON, XML, and CSV data embedded within normal text. This is perfect for processing LLM outputs.

from toon_parse import ToonConverter

# Text with embedded JSON
mixed_text = """
Here is the user profile you requested:
{
    "id": 101,
    "name": "Bob",
    "roles": ["admin", "editor"]
}
Please verify this information.
"""

# Automatically finds JSON, converts it to TOON, and preserves surrounding text
result = ToonConverter.from_json(mixed_text)
print(result)

# Output:
# Here is the user profile you requested:
# id: 101
# name: "Bob"
# roles[2]: "admin", "editor"
# Please verify this information.

🔐 Secure Conversion Middleware (New!)

The ToonConverter can act as a secure middleware for processing encrypted data streams (e.g., from microservices). It handles the full Decrypt -> Convert -> Encrypt pipeline internally.

Supported Algorithms

  • Fernet: High security (AES-128). Requires cryptography.
  • XOR: Lightweight obfuscation.
  • Base64: Encoding only.

Conversion Modes

  1. "middleware": Encrypted Input → Encrypted Output (Decrypt → Convert → Re-encrypt)
  2. "ingestion": Encrypted Input → Plain Output (Decrypt → Convert)
  3. "export": Plain Input → Encrypted Output (Convert → Encrypt)
  4. "no_encryption": Standard conversion (default)

Example Workflow

from toon_parse import ToonConverter, Encryptor
from cryptography.fernet import Fernet

# Setup
key = Fernet.generate_key()
enc = Encryptor(key=key, algorithm='fernet')
converter = ToonConverter(encryptor=enc)

# --- Mode 1: Middleware (Encrypted -> Encrypted) ---
raw_data = '{"user": "Alice", "role": "admin"}'
encrypted_input = enc.encrypt(raw_data)  # Simulate upstream encrypted data

# Converter decrypts, converts to TOON, and re-encrypts
encrypted_toon = converter.from_json(
    encrypted_input, 
    conversion_mode="middleware"
)
print(f"Secure Result: {encrypted_toon}")

# --- Mode 2: Ingestion (Encrypted -> Plain) ---
plain_toon = converter.from_json(
    encrypted_input,
    conversion_mode="ingestion"
)
print(f"Decrypted TOON: {plain_toon}")

# --- Mode 3: Export (Plain -> Encrypted) ---
my_data = {"status": "ok"}
secure_packet = converter.from_json(
    my_data,
    conversion_mode="export"
)
print(f"Encrypted Output: {secure_packet}")

⚡ Async Usage

For non-blocking operations in async applications (e.g., FastAPI), use AsyncToonConverter.

import asyncio
from toon_parse import AsyncToonConverter, Encryptor

async def main():
    # 1. Standard Async Usage
    converter = AsyncToonConverter()
    text = 'Data: <user><name>Alice</name></user>'
    toon = await converter.from_xml(text)
    print(toon)

    # 2. Async with Secure Middleware
    enc = Encryptor(algorithm='base64')
    secure_converter = AsyncToonConverter(encryptor=enc)
    
    # Decrypt -> Convert -> Encrypt (Middleware Mode)
    encrypted_msg = "eyJrZXkiOiAidmFsIn0=" # Base64 for {"key": "val"}
    
    # Use conversion_mode to specify pipeline behavior
    result = await secure_converter.from_json(
        encrypted_msg, 
        conversion_mode="middleware"
    )
    print(result)

asyncio.run(main())

📚 Features & Support

Feature JSON XML CSV YAML TOON
Python Dict/List Input N/A N/A N/A N/A
Pure String Input
Mixed Text Support
Async Support
Encryption Support
  • Mixed Text: Finds occurrences of data formats in text (JSON, XML, CSV) and converts them in-place.
  • Encryption: Supports Fernet, XOR, and Base64 middleware conversions.

⚙️ Static vs Instance Usage

Conversion Methods (from_json, to_json, etc.)

All conversion methods support both static and instance calling patterns:

from toon_parse import ToonConverter

# ✅ Static Usage (No Encryption)
toon = ToonConverter.from_json({"key": "value"})

# ✅ Instance Usage (Encryption Supported)
converter = ToonConverter(encryptor=enc)
toon = converter.from_json({"key": "value"}, conversion_mode="export")

Important:

  • Static calls (ToonConverter.from_json(...)) work but cannot use encryption features.
  • Instance calls are required to use conversion_mode and encryption middleware.

The same applies to async methods.

Validate Method

The validate() method is strictly static and does not support encryption:

# ✅ Correct Usage
result = ToonConverter.validate('key: "value"')

# ❌ Will NOT work with encryption
converter = ToonConverter(encryptor=enc)
result = converter.validate(encrypted_data)  # No decryption happens!

Why? Validation returns a dictionary (not a string), which cannot be encrypted. If you need to validate encrypted data, decrypt it first manually:

decrypted = enc.decrypt(encrypted_toon)
result = ToonConverter.validate(decrypted)

The same applies to AsyncToonConverter.validate().

🛠 API Reference

Core Converters

ToonConverter (Synchronous)

Constructor: ToonConverter(encryptor: Encryptor = None)

All conversion methods accept an optional conversion_mode argument:

  • conversion_mode: "no_encryption" (default), "middleware", "ingestion", "export".

  • from_json(data, conversion_mode=...): Converts dict/list/string to TOON.

  • to_json(toon_string, return_json=True, conversion_mode=...): Converts TOON to Python/JSON.

  • from_xml, from_csv, from_yaml, to_xml, to_csv, to_yaml: Equivalent methods.

  • validate(toon_string): Static Method Only. Validates TOON syntax. Does not support encryption.

AsyncToonConverter (Asynchronous)

Constructor: AsyncToonConverter(encryptor: Encryptor = None)

  • Mirrors all ToonConverter methods as async functions (e.g., await conv.from_json(...)).
  • Supports the same conversion_mode parameters for encryption pipelines.
  • validate(toon_string): Static Method Only. Async validation. No encryption support.

Encryption

Encryptor

Constructor: Encryptor(key=None, algorithm='fernet')

  • algorithm: 'fernet' (default), 'xor', 'base64'.
  • key: Required for Fernet/XOR.
  • encrypt(data), decrypt(data): Helper methods.

Utility Functions

from toon_parse import extract_json_from_string, extract_xml_from_string, extract_csv_from_string
# Direct access to extraction logic without conversion

📄 License

MIT License

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

toon_parse-2.0.0.tar.gz (26.8 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

toon_parse-2.0.0-py3-none-any.whl (20.7 kB view details)

Uploaded Python 3

File details

Details for the file toon_parse-2.0.0.tar.gz.

File metadata

  • Download URL: toon_parse-2.0.0.tar.gz
  • Upload date:
  • Size: 26.8 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.3

File hashes

Hashes for toon_parse-2.0.0.tar.gz
Algorithm Hash digest
SHA256 03c8b2681369d2b16926d167a762b9a6f868597b8808aa2221ac7b8e174c6ee7
MD5 5e045d3fdaef6f682256dbaf03bb2493
BLAKE2b-256 1761660cd6b3efd3e8572f7be1c2a12a5bb9aba3ffe566d42e09cd23f7f11432

See more details on using hashes here.

File details

Details for the file toon_parse-2.0.0-py3-none-any.whl.

File metadata

  • Download URL: toon_parse-2.0.0-py3-none-any.whl
  • Upload date:
  • Size: 20.7 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.3

File hashes

Hashes for toon_parse-2.0.0-py3-none-any.whl
Algorithm Hash digest
SHA256 9ea04ce13139e07f0cdf7b8c4f8ce8de30063e707d7c0c002b90359824d6cc64
MD5 51ae9991ee5e9b7c1d0c6dbbf3504782
BLAKE2b-256 6a8e1d7a6eb382f5c6552409eb6b715e7c5317389aadb94e423e43d49c876ff8

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page