server/scripts/download_nltk_data.py

#!/usr/bin/env python3
"""
Download all required NLTK data for the customer agent application.
Run this script once to ensure all NLTK resources are available.
"""

import nltk
import sys

def download_nltk_data():
    """Download all required NLTK datasets"""
    resources = [
        'punkt_tab',
        'punkt',
        'stopwords',
        'wordnet',
        'averaged_perceptron_tagger',
        'omw-1.4'
    ]
    
    print("Downloading NLTK data...")
    print("-" * 50)
    
    for resource in resources:
        try:
            print(f"Downloading {resource}...", end=" ")
            nltk.download(resource, quiet=True)
            print("✓")
        except Exception as e:
            print(f"✗ (Error: {e})")
    
    print("-" * 50)
    print("NLTK data download complete!")
    
    # Verify downloads
    print("\nVerifying downloads...")
    verification_paths = [
        'tokenizers/punkt_tab',
        'tokenizers/punkt',
        'corpora/stopwords',
        'corpora/wordnet'
    ]
    
    all_ok = True
    for path in verification_paths:
        try:
            nltk.data.find(path)
            print(f"✓ {path}")
        except LookupError:
            print(f"✗ {path} - NOT FOUND")
            all_ok = False
    
    if all_ok:
        print("\n✓ All NLTK resources are properly installed!")
        return 0
    else:
        print("\n✗ Some resources are missing. Please check the errors above.")
        return 1

if __name__ == "__main__":
    sys.exit(download_nltk_data())