Classify and Extract Compliance Documents with Unstructured and spaCy

                      
                      import os
import spacy
from typing import List, Dict, Any

# Load spaCy model
nlp = spacy.load('en_core_web_md')

# Configuration
class Config:
    DATA_DIRECTORY: str = os.getenv('DATA_DIRECTORY', './data')
    OUTPUT_DIRECTORY: str = os.getenv('OUTPUT_DIRECTORY', './output')

# Class to handle document classification
class DocumentClassifier:
    def __init__(self, config: Config) -> None:
        self.config = config

    def classify_documents(self, documents: List[str]) -> Dict[str, Any]:
        results = {}
        for doc in documents:
            try:
                processed_doc = nlp(doc)
                # Simple classification based on entities
                results[doc] = [ent.label_ for ent in processed_doc.ents]
            except Exception as e:
                results[doc] = {'error': str(e)}
        return results

if __name__ == '__main__':
    config = Config()
    classifier = DocumentClassifier(config)
    # Example documents
    documents = [
        "This is a financial report from 2022.",
        "Compliance audit results for Q1 2023."
    ]
    classifications = classifier.classify_documents(documents)
    print(classifications)