test_api.py

from gradio_client import Client, handle_file
from PIL import Image
import base64
import io
import os

# def test_base64_api(client_url="http://localhost:7860"):
#     """Test the base64 API endpoint"""
#     print("🧪 Testing Base64 API Endpoint...")
    
#     try:
#         # Initialize the client
#         client = Client(client_url)
        
#         # Check if temp_drawing.png exists, if not create a simple test image
#         if not os.path.exists("temp_drawing.png"):
#             print("Creating a test image...")
#             # Create a simple test image (white background with black circle)
#             img = Image.new('RGB', (400, 400), 'white')
#             from PIL import ImageDraw
#             draw = ImageDraw.Draw(img)
#             draw.ellipse([150, 150, 250, 250], fill='black')
#             img.save("temp_drawing.png")
        
#         # Load the image
#         image = Image.open("temp_drawing.png")
#         print(f"✓ Loaded image: {image.size}")
        
#         # Convert image to base64
#         buffered = io.BytesIO()
#         image.save(buffered, format="PNG")
#         img_str = base64.b64encode(buffered.getvalue()).decode()
#         print(f"✓ Converted to base64 (length: {len(img_str)})")
        
#         # Call the API endpoint with correct parameter names
#         result = client.predict(
#             image_data=img_str,
#             top_k=5,
#             api_name="/classify_image_api"
#         )
        
#         print("✓ API Response:")
#         print(f"  Success: {result.get('success', False)}")
#         if result.get('success'):
#             print("  Predictions:")
#             for i, pred in enumerate(result.get('predictions', []), 1):
#                 print(f"    {i}. {pred['category']} - {pred['confidence']:.3f}")
#         else:
#             print(f"  Error: {result.get('error', 'Unknown error')}")
            
#     except Exception as e:
#         print(f"❌ Base64 API test failed: {str(e)}")


# def test_main_interface(client_url="http://localhost:7860"):
#     """Test the main drawing interface"""
#     print("\n🧪 Testing Main Drawing Interface...")
    
#     try:
#         # Initialize the client
#         client = Client(client_url)
        
#         # Check if temp_drawing.png exists
#         if not os.path.exists("temp_drawing.png"):
#             print("Creating a test image...")
#             # Create a simple test image (white background with black circle)
#             img = Image.new('RGB', (400, 400), 'white')
#             from PIL import ImageDraw
#             draw = ImageDraw.Draw(img)
#             draw.ellipse([150, 150, 250, 250], fill='black')
#             img.save("temp_drawing.png")
        
#         # Load and convert image to the format expected by Sketchpad
#         image = Image.open("temp_drawing.png")
        
#         # Create a mock sketchpad data structure
#         sketchpad_data = {
#             "background": image,
#             "layers": [image],
#             "composite": image
#         }
        
#         # Test the main gradio_classify function
#         result = client.predict(
#             canvas=sketchpad_data,
#             api_name="/gradio_classify"
#         )
        
#         print("✓ Main Interface Response:")
#         print(result)
            
#     except Exception as e:
#         print(f"❌ Main interface test failed: {str(e)}")
#         print("💡 Try using the web interface directly instead")


# def test_huggingface_space():
#     """Test the deployed Hugging Face space"""
#     print("\n🧪 Testing Hugging Face Space...")
    
#     try:
#         # Test your deployed space
#         client = Client("souvikg544/quickdraw-classifier")
        
#         # Check if temp_drawing.png exists
#         if not os.path.exists("temp_drawing.png"):
#             print("Creating a test image...")
#             img = Image.new('RGB', (400, 400), 'white')
#             from PIL import ImageDraw
#             draw = ImageDraw.Draw(img)
#             draw.ellipse([150, 150, 250, 250], fill='black')
#             img.save("temp_drawing.png")
        
#         # Load and convert image to the format expected by Sketchpad
#         image = Image.open("temp_drawing.png")
        
#         # Create a mock sketchpad data structure
#         sketchpad_data = {
#             "background": image,
#             "layers": [image], 
#             "composite": image
#         }
        
#         # Test with sketchpad format
#         result = client.predict(
#             canvas=sketchpad_data,
#             api_name="/gradio_classify"
#         )
        
#         print("✓ Hugging Face Space Response:")
#         print(result)
        
#     except Exception as e:
#         print(f"❌ Hugging Face space test failed: {str(e)}")


# def create_test_drawings():
#     """Create various test drawings for testing"""
#     print("\n🎨 Creating test drawings...")
    
#     from PIL import ImageDraw
    
#     # Test drawing 1: Circle (might be classified as sun, moon, etc.)
#     img1 = Image.new('RGB', (400, 400), 'white')
#     draw1 = ImageDraw.Draw(img1)
#     draw1.ellipse([150, 150, 250, 250], fill='black')
#     img1.save("test_circle.png")
#     print("✓ Created test_circle.png")
    
#     # Test drawing 2: Rectangle (might be classified as house, car, etc.)
#     img2 = Image.new('RGB', (400, 400), 'white')
#     draw2 = ImageDraw.Draw(img2)
#     draw2.rectangle([100, 150, 300, 250], fill='black')
#     img2.save("test_rectangle.png")
#     print("✓ Created test_rectangle.png")
    
#     # Test drawing 3: Triangle (might be classified as tree, mountain, etc.)
#     img3 = Image.new('RGB', (400, 400), 'white')
#     draw3 = ImageDraw.Draw(img3)
#     draw3.polygon([(200, 100), (150, 200), (250, 200)], fill='black')
#     img3.save("test_triangle.png")
#     print("✓ Created test_triangle.png")


# def test_multiple_drawings():
#     """Test classification with multiple different drawings"""
#     print("\n🧪 Testing Multiple Drawings...")
    
#     test_files = ["test_circle.png", "test_rectangle.png", "test_triangle.png"]
    
#     try:
#         client = Client("http://localhost:7860")
        
#         for test_file in test_files:
#             if os.path.exists(test_file):
#                 print(f"\n--- Testing {test_file} ---")
                
#                 # Load image and create sketchpad format
#                 image = Image.open(test_file)
#                 sketchpad_data = {
#                     "background": image,
#                     "layers": [image],
#                     "composite": image
#                 }
                
#                 result = client.predict(
#                     canvas=sketchpad_data,
#                     api_name="/gradio_classify"
#                 )
#                 print(result)
#             else:
#                 print(f"❌ {test_file} not found")
                
#     except Exception as e:
#         print(f"❌ Multiple drawings test failed: {str(e)}")


# def test_base64_with_file():
#     """Test base64 API with actual file conversion"""
#     print("\n🧪 Testing Base64 API with File Conversion...")
    
#     try:
#         client = Client("http://localhost:7860")
        
#         # Use one of our test files
#         test_file = "test_circle.png"
#         if os.path.exists(test_file):
#             # Load and convert to base64
#             with open(test_file, "rb") as f:
#                 img_bytes = f.read()
#                 img_b64 = base64.b64encode(img_bytes).decode()
            
#             print(f"✓ Converted {test_file} to base64")
            
#             # Test the base64 API with correct parameter names
#             result = client.predict(
#                 image_data=img_b64,
#                 top_k=5,
#                 api_name="/classify_image_api"
#             )
            
#             print("✓ Base64 API Response:")
#             print(result)
#         else:
#             print("❌ No test file found for base64 testing")
            
#     except Exception as e:
#         print(f"❌ Base64 file test failed: {str(e)}")


# if __name__ == "__main__":
#     print("🚀 QuickDraw Classifier API Test Suite")
#     print("=" * 50)
    
#     # Create test drawings first
#     create_test_drawings()
    
#     # Test local server (make sure your app is running on localhost:7860)
#     print("\n📍 Testing Local Server...")
#     test_base64_api("http://localhost:7860")
#     test_main_interface("http://localhost:7860")
#     test_multiple_drawings()
    
#     # Test Hugging Face deployment
#     test_huggingface_space()
    
#     # Test base64 API with file conversion
#     test_base64_with_file()
    
#     print("\n✅ Test suite completed!")
#     print("\n💡 Tips:")
#     print("- Make sure your local server is running: python app.py")
#     print("- For HF Space testing, ensure your space is public and running")
#     print("- Check the generated test images: test_circle.png, test_rectangle.png, test_triangle.png")


from gradio_client import Client

image = Image.open("temp_drawing.png")
buffered = io.BytesIO()
image.save(buffered, format="PNG")
img_str = base64.b64encode(buffered.getvalue()).decode()

# client = Client("souvikg544/quickdraw-classifier")
client = Client("https://6820c804655322f295.gradio.live/")
result = client.predict(
		image_data=img_str,
		top_k=5,
        word="broom",
		api_name="/classify_image_api"
)
print(result)