test_hf_only.py

#!/usr/bin/env python3
"""
Test script to validate Hugging Face only loading.
"""

import os
import warnings
warnings.filterwarnings('ignore')

def test_hf_only_loading():
    """Test that the loader only works with Hugging Face."""
    print("🤗 TESTING HUGGING FACE ONLY LOADING")
    print("=" * 50)
    
    from data_loader import AgriculturalDataLoader
    
    # Check if HF token is available
    hf_token = os.environ.get("HF_TOKEN")
    if not hf_token:
        print("⚠️  No HF_TOKEN found in environment variables")
        print("💡 Set HF_TOKEN to test Hugging Face loading")
        print("🔧 For this test, we'll try without token (may fail)")
    
    try:
        # Create loader (HF only)
        loader = AgriculturalDataLoader(
            dataset_id="HackathonCRA/2024",
            hf_token=hf_token
        )
        
        print(f"🤗 Attempting to load from dataset: {loader.dataset_id}")
        
        # Load data
        df = loader.load_all_files()
        
        print(f"✅ Success! Loaded {len(df):,} records from Hugging Face")
        print(f"📊 Years: {sorted(df['year'].unique())}")
        print(f"🌱 Crops: {df['crop_type'].nunique()}")
        print(f"📍 Plots: {df['plot_name'].nunique()}")
        print(f"💊 Herbicide applications: {df['is_herbicide'].sum()}")
        
        return True
        
    except Exception as e:
        print(f"❌ Failed to load from Hugging Face: {e}")
        print("💡 This is expected if the dataset doesn't exist yet")
        print("🔧 Make sure to upload your dataset to HF Hub first")
        return False

def test_no_local_fallback():
    """Test that there's no local fallback."""
    print("\n🚫 TESTING NO LOCAL FALLBACK")
    print("=" * 50)
    
    from data_loader import AgriculturalDataLoader
    
    try:
        # Create loader with non-existent dataset
        loader = AgriculturalDataLoader(
            dataset_id="nonexistent/dataset"
        )
        
        # This should fail without falling back to local
        df = loader.load_all_files()
        
        print(f"❌ Unexpected success - loaded {len(df)} records")
        print("⚠️  This suggests local fallback is still active")
        return False
        
    except Exception as e:
        print(f"✅ Expected failure: {e}")
        print("✅ Confirmed: No local fallback, HF only")
        return True

def test_simple_usage():
    """Test simple usage pattern."""
    print("\n📝 SIMPLE USAGE EXAMPLE")
    print("=" * 50)
    
    print("💡 Recommended usage pattern:")
    print()
    
    usage_code = '''
from data_loader import AgriculturalDataLoader

# Simple HF-only loader
loader = AgriculturalDataLoader(dataset_id="HackathonCRA/2024")

# Load data (will use HF_TOKEN from environment)
df = loader.load_all_files()

# Analyze data
print(f"Loaded {len(df)} records from Hugging Face")
'''
    
    print(usage_code)
    
    try:
        from data_loader import AgriculturalDataLoader
        loader = AgriculturalDataLoader(dataset_id="HackathonCRA/2024")
        print("✅ Loader created successfully")
        print(f"🎯 Target dataset: {loader.dataset_id}")
        print(f"🔑 Using token: {'Yes' if loader.hf_token else 'No (from env)'}")
        
        return True
        
    except Exception as e:
        print(f"❌ Failed to create loader: {e}")
        return False

def main():
    """Run all tests."""
    print("🚜 HUGGING FACE ONLY - VALIDATION TESTS")
    print("=" * 60)
    print()
    
    results = []
    
    # Test 1: HF loading
    results.append(("HF Only Loading", test_hf_only_loading()))
    
    # Test 2: No local fallback
    results.append(("No Local Fallback", test_no_local_fallback()))
    
    # Test 3: Simple usage
    results.append(("Simple Usage", test_simple_usage()))
    
    # Summary
    print("\n📋 TEST SUMMARY")
    print("=" * 30)
    
    passed = 0
    for test_name, result in results:
        status = "✅ PASS" if result else "❌ FAIL"
        print(f"{test_name:<20} {status}")
        if result:
            passed += 1
    
    print(f"\n🎯 Results: {passed}/{len(results)} tests passed")
    
    if passed >= 2:  # Allow HF loading to fail if dataset doesn't exist
        print("🎉 Validation successful! Loader is HF-only.")
    else:
        print("⚠️  Validation issues detected.")
    
    print("\n🚀 DEPLOYMENT CHECKLIST:")
    print("✅ Remove local file dependencies")
    print("✅ HF-only data loading")
    print("✅ No fallback mechanisms")
    print("🔲 Upload dataset to HF Hub")
    print("🔲 Set HF_TOKEN in production")
    print("🔲 Test with real HF dataset")

if __name__ == "__main__":
    main()