canonmap 0.1.16

CanonMap - A Python library for entity canonicalization and mapping

CanonMap

A Python library for data mapping and canonicalization.

Installation

pip install canonmap

Dependencies

CanonMap requires the following key dependencies:

• Python 3.8 or higher
• spaCy and its English language model

The spaCy English language model (en_core_web_sm) will be automatically downloaded when you first use the library. No manual installation is required.

Quick Start

from canonmap import CanonMap

# Initialize the library
canon = CanonMap()

# Generate artifacts from a CSV file
artifacts = canon.generate_artifacts(
    csv_path="path/to/your/data.csv",
    entity_fields=["name", "email"],
    use_other_fields_as_metadata=True
)

# Save artifacts to files
zip_path = canon.save_artifacts(
    artifacts=artifacts,
    output_path="output",
    name="my_data"
)

print(f"Artifacts saved to: {zip_path}")

Detailed Example

Here's a complete example showing how to use the library in a real-world scenario:

from canonmap import CanonMap
import pandas as pd
from pathlib import Path

def process_customer_data(input_csv: str, output_dir: str):
    # Initialize CanonMap
    canon = CanonMap()
    
    # Define the entity fields we want to extract
    entity_fields = [
        "customer_name",
        "email",
        "phone_number",
        "company"
    ]
    
    # Generate artifacts from the CSV
    artifacts = canon.generate_artifacts(
        csv_path=input_csv,
        entity_fields=entity_fields,
        use_other_fields_as_metadata=True,  # Include other columns as metadata
        num_rows=None  # Process all rows
    )
    
    # Create output directory if it doesn't exist
    output_path = Path(output_dir)
    output_path.mkdir(parents=True, exist_ok=True)
    
    # Save the artifacts
    zip_path = canon.save_artifacts(
        artifacts=artifacts,
        output_path=str(output_path),
        name="customer_data",
        save_metadata=True,
        save_schema=True
    )
    
    # You can also work with the artifacts directly
    metadata = artifacts["metadata"]
    schema = artifacts["schema"]
    
    # Example: Print some statistics
    print(f"Processed {metadata.get('row_count', 0)} rows")
    print(f"Found {len(schema.get('entities', []))} entities")
    
    return zip_path

# Usage
if __name__ == "__main__":
    zip_file = process_customer_data(
        input_csv="customers.csv",
        output_dir="processed_data"
    )
    print(f"Processing complete. Results saved to: {zip_file}")

Features

• Process CSV files and generate metadata and schema
• Extract and canonicalize entity fields
• Map data to standardized formats
• Save artifacts as JSON files or ZIP archives
• Configurable processing options

Requirements

• Python 3.8+
• See setup.py for full list of dependencies

License

MIT License