backend/test_api.py

"""
Test script for Pathora Colposcopy API endpoints
Demonstrates how to use both AI model and LLM endpoints
"""

import requests
import json
import base64
from pathlib import Path

# API Configuration
BASE_URL = "http://localhost:8000"  # Change to your deployment URL
API_KEY = "your_gemini_api_key_here"  # For local testing

def test_health_check():
    """Test the health check endpoint"""
    print("=" * 60)
    print("Testing Health Check Endpoint")
    print("=" * 60)
    
    response = requests.get(f"{BASE_URL}/health")
    print(f"Status Code: {response.status_code}")
    print(f"Response: {json.dumps(response.json(), indent=2)}")
    print()

def test_acetowhite_detection(image_path: str):
    """Test acetowhite contour detection"""
    print("=" * 60)
    print("Testing Acetowhite Contour Detection")
    print("=" * 60)
    
    with open(image_path, 'rb') as f:
        files = {'file': f}
        data = {'conf_threshold': 0.5}
        
        response = requests.post(
            f"{BASE_URL}/api/infer-aw-contour",
            files=files,
            data=data
        )
    
    print(f"Status Code: {response.status_code}")
    result = response.json()
    
    # Print without base64 image for readability
    print(f"Status: {result.get('status')}")
    print(f"Detections: {result.get('detections')}")
    print(f"Contours: {len(result.get('contours', []))}")
    print(f"Confidence Threshold: {result.get('confidence_threshold')}")
    
    # Save result image if available
    if result.get('result_image'):
        output_path = "test_output_aw.png"
        img_data = base64.b64decode(result['result_image'])
        with open(output_path, 'wb') as f:
            f.write(img_data)
        print(f"Result image saved to: {output_path}")
    print()

def test_cervix_detection(image_path: str):
    """Test cervix bounding box detection"""
    print("=" * 60)
    print("Testing Cervix Bounding Box Detection")
    print("=" * 60)
    
    with open(image_path, 'rb') as f:
        files = {'file': f}
        data = {'conf_threshold': 0.4}
        
        response = requests.post(
            f"{BASE_URL}/api/infer-cervix-bbox",
            files=files,
            data=data
        )
    
    print(f"Status Code: {response.status_code}")
    result = response.json()
    
    print(f"Status: {result.get('status')}")
    print(f"Detections: {result.get('detections')}")
    print(f"Bounding Boxes: {json.dumps(result.get('bounding_boxes', []), indent=2)}")
    
    # Save result image if available
    if result.get('result_image'):
        output_path = "test_output_cervix.png"
        img_data = base64.b64decode(result['result_image'])
        with open(output_path, 'wb') as f:
            f.write(img_data)
        print(f"Result image saved to: {output_path}")
    print()

def test_batch_inference(image_paths: list):
    """Test batch inference on multiple images"""
    print("=" * 60)
    print("Testing Batch Inference")
    print("=" * 60)
    
    files = [('files', open(img, 'rb')) for img in image_paths]
    data = {'conf_threshold': 0.5}
    
    response = requests.post(
        f"{BASE_URL}/api/batch-infer",
        files=files,
        data=data
    )
    
    # Close file handles
    for _, f in files:
        f.close()
    
    print(f"Status Code: {response.status_code}")
    result = response.json()
    
    print(f"Status: {result.get('status')}")
    print(f"Total Files: {result.get('total_files')}")
    
    for i, res in enumerate(result.get('results', [])):
        print(f"\nImage {i+1}: {res.get('filename')}")
        print(f"  Status: {res.get('status')}")
        print(f"  Detections: {res.get('detections')}")
    print()

def test_chat():
    """Test LLM chat endpoint"""
    print("=" * 60)
    print("Testing Chat Endpoint")
    print("=" * 60)
    
    payload = {
        "message": "What are the typical signs of a high-grade squamous intraepithelial lesion (HSIL) on colposcopy?",
        "history": []
    }
    
    response = requests.post(
        f"{BASE_URL}/api/chat",
        json=payload
    )
    
    print(f"Status Code: {response.status_code}")
    
    if response.status_code == 200:
        result = response.json()
        print(f"Status: {result.get('status')}")
        print(f"Model: {result.get('model')}")
        print(f"Response:\n{result.get('response')}")
    else:
        print(f"Error: {response.json()}")
    print()

def test_chat_with_history():
    """Test chat with conversation history"""
    print("=" * 60)
    print("Testing Chat with History")
    print("=" * 60)
    
    payload = {
        "message": "What about low-grade lesions?",
        "history": [
            {
                "role": "user",
                "text": "What are high-grade lesions?"
            },
            {
                "role": "bot",
                "text": "High-grade lesions (HSIL) show dense acetowhite epithelium, coarse punctation, and sharp borders."
            }
        ]
    }
    
    response = requests.post(
        f"{BASE_URL}/api/chat",
        json=payload
    )
    
    print(f"Status Code: {response.status_code}")
    
    if response.status_code == 200:
        result = response.json()
        print(f"Response:\n{result.get('response')}")
    else:
        print(f"Error: {response.json()}")
    print()

def test_report_generation():
    """Test report generation endpoint"""
    print("=" * 60)
    print("Testing Report Generation")
    print("=" * 60)
    
    payload = {
        "patient_data": {
            "age": 35,
            "gravida": 2,
            "para": 2,
            "lmp": "2024-02-01",
            "indication": "Abnormal Pap smear - ASCUS",
            "menstrual_status": "Regular"
        },
        "exam_findings": {
            "native": {
                "cervix_visible": True,
                "transformation_zone": "Type 1 (fully visible)",
                "ectropion": "Mild",
                "discharge": "None"
            },
            "acetic_acid": {
                "acetowhite_lesions": True,
                "location": "6-9 o'clock position",
                "density": "Dense white",
                "borders": "Sharp, well-defined",
                "size": "Moderate (covering 2 quadrants)"
            },
            "green_filter": {
                "vascular_patterns": "Coarse punctation",
                "mosaic": "Present",
                "atypical_vessels": "None"
            },
            "lugol": {
                "iodine_uptake": "Partial iodine negative area",
                "pattern": "Corresponds to acetowhite area"
            }
        }
    }
    
    response = requests.post(
        f"{BASE_URL}/api/generate-report",
        json=payload
    )
    
    print(f"Status Code: {response.status_code}")
    
    if response.status_code == 200:
        result = response.json()
        print(f"Status: {result.get('status')}")
        print(f"Model: {result.get('model')}")
        print(f"\nGenerated Report:\n{'-' * 60}")
        print(result.get('report'))
        print('-' * 60)
    else:
        print(f"Error: {response.json()}")
    print()

def main():
    """Run all tests"""
    print("\n" + "=" * 60)
    print("PATHORA COLPOSCOPY API TEST SUITE")
    print("=" * 60 + "\n")
    
    # Test health check
    test_health_check()
    
    # Test AI model endpoints (you'll need to provide actual image paths)
    # Uncomment and add your image paths:
    # test_acetowhite_detection("path/to/your/image.jpg")
    # test_cervix_detection("path/to/your/image.jpg")
    # test_batch_inference(["image1.jpg", "image2.jpg"])
    
    # Test LLM endpoints
    test_chat()
    test_chat_with_history()
    test_report_generation()
    
    print("\n" + "=" * 60)
    print("ALL TESTS COMPLETED")
    print("=" * 60 + "\n")

if __name__ == "__main__":
    # Check if requests is installed
    try:
        import requests
    except ImportError:
        print("Please install requests: pip install requests")
        exit(1)
    
    main()