Update app.py

app.py

@@ -6,6 +6,7 @@ from dotenv import load_dotenv
 import time
 import traceback
 import sys
+import json
 
 # Load environment variables
 load_dotenv()
@@ -15,16 +16,33 @@ api_key = os.getenv("GEMINI_API_KEY", "AIzaSyB0IOx76FydAk4wabMz1juzzHF5oBiHW64")
 if api_key == "AIzaSyB0IOx76FydAk4wabMz1juzzHF5oBiHW64":
     print("WARNING: Using hardcoded API key. Set GEMINI_API_KEY environment variable instead.")
 
-try:
-    genai.configure(api_key=api_key)
-    print("Successfully configured Gemini API with provided key")
-except Exception as e:
-    print(f"Error configuring Gemini API: {str(e)}")
-    print(traceback.format_exc())
+# Function to test API connectivity
+def test_gemini_api():
+    try:
+        genai.configure(api_key=api_key)
+        
+        # Test with a simple text prompt
+        model = genai.GenerativeModel('gemini-1.0-pro')
+        response = model.generate_content("Hello, please respond with 'API is working'")
+        
+        if not response or not hasattr(response, 'text') or not response.text:
+            print("WARNING: Received empty response during API test")
+            return False
+            
+        print(f"API Test Response: {response.text.strip()}")
+        return True
+    except Exception as e:
+        print(f"ERROR: Failed to connect to Gemini API: {str(e)}")
+        print(traceback.format_exc())
+        return False
 
 # Initialize Flask app
 app = Flask(__name__)
 
+# Configure error responses
+@app.errorhandler(500)
+def server_error(e):
+    return jsonify(error="Internal server error: " + str(e)), 500
 
 def extract_text_with_gemini(image_path):
     """Extract text from image using Gemini Vision model"""
@@ -34,13 +52,36 @@ def extract_text_with_gemini(image_path):
     for attempt in range(max_retries):
         try:
             print(f"Attempt {attempt + 1} to extract text using Gemini...")
-            # Initialize Gemini Pro Vision model
-            model = genai.GenerativeModel('gemini-1.0-pro-vision')  # Fallback to a stable model
+            
+            # Try different model versions in case one is not available
+            model_options = ['gemini-1.0-pro-vision', 'gemini-pro-vision']
+            
+            for model_name in model_options:
+                try:
+                    print(f"Trying model: {model_name}")
+                    model = genai.GenerativeModel(model_name)
+                    break
+                except Exception as model_error:
+                    print(f"Error with model {model_name}: {str(model_error)}")
+                    if model_name == model_options[-1]:  # Last model option
+                        raise
+                    continue
 
             # Load the image
             with Image.open(image_path) as img:
                 print(f"Image loaded from {image_path} (Size: {img.size}, Format: {img.format})")
                 
+                # Resize image if too large (API may have size limits)
+                max_dimension = 1024
+                if img.width > max_dimension or img.height > max_dimension:
+                    print(f"Resizing large image from {img.width}x{img.height}")
+                    ratio = min(max_dimension / img.width, max_dimension / img.height)
+                    new_width = int(img.width * ratio)
+                    new_height = int(img.height * ratio)
+                    img = img.resize((new_width, new_height))
+                    print(f"Resized to {new_width}x{new_height}")
+                    img.save(image_path)  # Save resized image
+                
                 # Create prompt for text extraction
                 prompt = "Extract all the text from this image. Return only the extracted text, nothing else."
 
@@ -62,7 +103,7 @@ def extract_text_with_gemini(image_path):
                 print(f"Retrying in {retry_delay} seconds...")
                 time.sleep(retry_delay)
                 continue
-            return "Could not extract text from the image. Please try with a clearer image."
+            return f"Could not extract text from the image: {str(e)}"
 
 
 def translate_text(text):
@@ -73,12 +114,27 @@ def translate_text(text):
     # Check if there's text to translate
     if not text or text.strip() == "":
         return "No text to translate."
+    
+    # If the text indicates an error occurred during extraction, don't try to translate
+    if text.startswith("Could not extract text from the image"):
+        return "Cannot translate due to OCR failure."
 
     for attempt in range(max_retries):
         try:
             print(f"Attempt {attempt + 1} to translate text using Gemini...")
-            # Initialize Gemini model
-            model = genai.GenerativeModel('gemini-1.0-pro')  # Fallback to a stable model
+            # Try different model versions
+            model_options = ['gemini-1.0-pro', 'gemini-pro']
+            
+            for model_name in model_options:
+                try:
+                    print(f"Trying model: {model_name}")
+                    model = genai.GenerativeModel(model_name)
+                    break
+                except Exception as model_error:
+                    print(f"Error with model {model_name}: {str(model_error)}")
+                    if model_name == model_options[-1]:  # Last model option
+                        raise
+                    continue
 
             # Create prompt for translation
             prompt = f"""
@@ -105,7 +161,7 @@ def translate_text(text):
                 print(f"Retrying in {retry_delay} seconds...")
                 time.sleep(retry_delay)
                 continue
-            return "Translation failed. Please try again later."
+            return f"Translation failed: {str(e)}"
 
 
 @app.route('/')
@@ -113,6 +169,17 @@ def home():
     return render_template('index.html')
 
 
+@app.route('/api-status', methods=['GET'])
+def api_status():
+    """Endpoint to check API status"""
+    api_working = test_gemini_api()
+    return jsonify({
+        'status': 'ok' if api_working else 'error',
+        'api_working': api_working,
+        'message': 'Gemini API is working correctly' if api_working else 'Gemini API is not working'
+    })
+
+
 @app.route('/upload', methods=['POST'])
 def upload_file():
     print("Received upload request")
@@ -133,6 +200,10 @@ def upload_file():
 
     temp_path = None
     try:
+        # Check API status before proceeding
+        if not test_gemini_api():
+            return jsonify({'error': 'Gemini API is not working. Please check your API key and try again.'}), 500
+
         # Create temp directory if it doesn't exist
         temp_dir = "temp"
         if not os.path.exists(temp_dir):
@@ -163,7 +234,7 @@ def upload_file():
         print(error_msg)
         print(traceback.format_exc())
         return jsonify({
-            'error': 'An error occurred while processing your image. Please try again.'
+            'error': error_msg
         }), 500
     finally:
         # Clean up temporary file if it exists
@@ -181,6 +252,13 @@ if __name__ == '__main__':
         os.makedirs('templates')
         print("Created 'templates' directory. Please place your HTML files here.")
 
+    # Test API connectivity at startup
+    api_working = test_gemini_api()
+    if api_working:
+        print("✅ Gemini API connection successful!")
+    else:
+        print("❌ WARNING: Gemini API connection failed. The application may not work correctly!")
+
     # For Hugging Face Spaces, we need to listen on 0.0.0.0 and port 7860
     print(f"Starting Flask app on port {os.environ.get('PORT', 7860)}")
     app.run(host='0.0.0.0', port=int(os.environ.get('PORT', 7860)))