smiles-validator 0.1.3

Lightweight Pydantic v2 validator for RDKit SMILES: sanitizes, canonicalizes & caches results

Smiles-Validator

Lightweight Pydantic v2 validator for RDKit SMILES strings that validates, sanitizes, and optionally returns canonical or original input.

Install

pip install smiles-validator

Requires Python ≥ 3.10 · Pydantic v2 · RDKit 2024.9+

Usage

from pydantic import BaseModel
from typing import Annotated
from smiles_validator import SmilesText, SmilesValidator

class Model(BaseModel):
    # Default: Canonical SMILES
    canonical: SmilesText
    # Keep original input
    original: Annotated[str, SmilesValidator(keep_original=True)]

m = Model(
    canonical="C1=CC=CC=C1",
    original="C1=CC=CC=C1"
)

print(m.canonical)  # => "c1ccccc1"
print(m.original)   # => "C1=CC=CC=C1"

Features

• 🚀 Fast SMILES validation using RDKit
• 🔄 Canonicalization with optional original input preservation
• 📦 Pydantic v2 integration
• 📊 Comprehensive test coverage
• 🧪 CI/CD with multiple Python versions

API

SmilesValidator Class

SmilesValidator(keep_original: bool = False)

Parameters:

• keep_original: If True, returns the exact input string after validation. If False (default), returns the canonical SMILES.

SmilesText Type

SmilesText is an alias for Annotated[str, SmilesValidator()] that automatically uses the default validator settings (canonical output).

Development

# Install dependencies
uv sync --locked --all-extras --dev

# Run tests
just test

# Run linting
just check

# Run formatting
just format

License

MIT License · GitHub

Acknowledgements

• Built with RDKit for chemical structure handling
• Uses Pydantic v2 for validation integration