Update app.py

app.py

@@ -1,154 +1,186 @@
-import os
-from flask import Flask, render_template, request, jsonify
-import google.generativeai as genai
-from PIL import Image
-from dotenv import load_dotenv
-import time
-
-# Load environment variables
-load_dotenv()
-
-# Configure Gemini API with key from environment variable
-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)
-except Exception as e:
-    print(f"Error configuring Gemini API: {str(e)}")
-
-# Initialize Flask app
-app = Flask(__name__)
-
-
-def extract_text_with_gemini(image_path):
-    """Extract text from image using Gemini Vision model"""
-    max_retries = 3
-    retry_delay = 2
-
-    for attempt in range(max_retries):
-        try:
-            # Initialize Gemini Pro Vision model
-            model = genai.GenerativeModel('gemini-2.0-flash')
-
-            # Load the image
-            with Image.open(image_path) as img:
-                # Create prompt for text extraction
-                prompt = "Extract all the text from this image. Return only the extracted text, nothing else."
-
-                # Generate response with image
-                response = model.generate_content([prompt, img])
-
-                # Validate response
-                if not response or not hasattr(response, 'text') or not response.text:
-                    raise ValueError("Received empty response from Gemini API")
-
-                return response.text.strip()
-        except Exception as e:
-            print(f"Attempt {attempt + 1} failed: {str(e)}")
-            if attempt < max_retries - 1:
-                time.sleep(retry_delay)
-                continue
-            return "Could not extract text from the image. Please try with a clearer image."
-
-
-def translate_text(text):
-    """Translate text from English to Hindi using Gemini"""
-    max_retries = 3
-    retry_delay = 2
-
-    # Check if there's text to translate
-    if not text or text.strip() == "":
-        return "No text to translate."
-
-    for attempt in range(max_retries):
-        try:
-            # Initialize Gemini model
-            model = genai.GenerativeModel('gemini-2.0-flash')
-
-            # Create prompt for translation
-            prompt = f"""
-            Translate the following English text to Hindi. 
-            Keep proper names, titles, and organization names unchanged.
-            Text to translate: {text}
-            """
-
-            # Generate response
-            response = model.generate_content(prompt)
-
-            # Validate response
-            if not response or not hasattr(response, 'text') or not response.text:
-                raise ValueError("Received empty response from Gemini API")
-
-            return response.text.strip()
-        except Exception as e:
-            print(f"Translation attempt {attempt + 1} failed: {str(e)}")
-            if attempt < max_retries - 1:
-                time.sleep(retry_delay)
-                continue
-            return "Translation failed. Please try again later."
-
-
-@app.route('/')
-def home():
-    return render_template('index.html')
-
-
-@app.route('/upload', methods=['POST'])
-def upload_file():
-    if 'file' not in request.files:
-        return jsonify({'error': 'No file uploaded'}), 400
-
-    file = request.files['file']
-    if file.filename == '':
-        return jsonify({'error': 'No file selected'}), 400
-
-    # Check file extension
-    allowed_extensions = {'png', 'jpg', 'jpeg', 'gif', 'bmp'}
-    if '.' not in file.filename or file.filename.rsplit('.', 1)[1].lower() not in allowed_extensions:
-        return jsonify({'error': 'Invalid file format. Please upload an image (PNG, JPG, JPEG, GIF, BMP).'}), 400
-
-    temp_path = None
-    try:
-        # Create temp directory if it doesn't exist
-        temp_dir = "temp"
-        if not os.path.exists(temp_dir):
-            os.makedirs(temp_dir)
-
-        # Save the uploaded file temporarily with a unique name
-        temp_path = os.path.join(temp_dir, f"temp_image_{int(time.time())}.png")
-        file.save(temp_path)
-
-        # Extract text using Gemini
-        extracted_text = extract_text_with_gemini(temp_path)
-
-        # Translate text
-        translated_text = translate_text(extracted_text)
-
-        return jsonify({
-            'original_text': extracted_text,
-            'translated_text': translated_text
-        })
-    except Exception as e:
-        print(f"Error processing image: {str(e)}")
-        return jsonify({
-            'error': 'An error occurred while processing your image. Please try again.'
-        }), 500
-    finally:
-        # Clean up temporary file if it exists
-        try:
-            if temp_path and os.path.exists(temp_path):
-                os.remove(temp_path)
-        except Exception as e:
-            print(f"Failed to remove temporary file: {str(e)}")
-
-
-if __name__ == '__main__':
-    # Ensure the template folder exists
-    if not os.path.exists('templates'):
-        os.makedirs('templates')
-        print("Created 'templates' directory. Please place your HTML files here.")
-
-    # For Hugging Face Spaces, we need to listen on 0.0.0.0 and port 7860
+import os
+from flask import Flask, render_template, request, jsonify
+import google.generativeai as genai
+from PIL import Image
+from dotenv import load_dotenv
+import time
+import traceback
+import sys
+
+# Load environment variables
+load_dotenv()
+
+# Configure Gemini API with key from environment variable
+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())
+
+# Initialize Flask app
+app = Flask(__name__)
+
+
+def extract_text_with_gemini(image_path):
+    """Extract text from image using Gemini Vision model"""
+    max_retries = 3
+    retry_delay = 2
+
+    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
+
+            # Load the image
+            with Image.open(image_path) as img:
+                print(f"Image loaded from {image_path} (Size: {img.size}, Format: {img.format})")
+                
+                # Create prompt for text extraction
+                prompt = "Extract all the text from this image. Return only the extracted text, nothing else."
+
+                # Generate response with image
+                print("Sending request to Gemini API for text extraction...")
+                response = model.generate_content([prompt, img])
+
+                # Validate response
+                if not response or not hasattr(response, 'text') or not response.text:
+                    raise ValueError("Received empty response from Gemini API")
+
+                extracted_text = response.text.strip()
+                print(f"Successfully extracted text (length: {len(extracted_text)})")
+                return extracted_text
+        except Exception as e:
+            print(f"Attempt {attempt + 1} failed: {str(e)}")
+            print(traceback.format_exc())
+            if attempt < max_retries - 1:
+                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."
+
+
+def translate_text(text):
+    """Translate text from English to Hindi using Gemini"""
+    max_retries = 3
+    retry_delay = 2
+
+    # Check if there's text to translate
+    if not text or text.strip() == "":
+        return "No text to translate."
+
+    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
+
+            # Create prompt for translation
+            prompt = f"""
+            Translate the following English text to Hindi. 
+            Keep proper names, titles, and organization names unchanged.
+            Text to translate: {text}
+            """
+
+            # Generate response
+            print("Sending request to Gemini API for translation...")
+            response = model.generate_content(prompt)
+
+            # Validate response
+            if not response or not hasattr(response, 'text') or not response.text:
+                raise ValueError("Received empty response from Gemini API")
+
+            translated_text = response.text.strip()
+            print(f"Successfully translated text (length: {len(translated_text)})")
+            return translated_text
+        except Exception as e:
+            print(f"Translation attempt {attempt + 1} failed: {str(e)}")
+            print(traceback.format_exc())
+            if attempt < max_retries - 1:
+                print(f"Retrying in {retry_delay} seconds...")
+                time.sleep(retry_delay)
+                continue
+            return "Translation failed. Please try again later."
+
+
+@app.route('/')
+def home():
+    return render_template('index.html')
+
+
+@app.route('/upload', methods=['POST'])
+def upload_file():
+    print("Received upload request")
+    if 'file' not in request.files:
+        print("No file part in the request")
+        return jsonify({'error': 'No file uploaded'}), 400
+
+    file = request.files['file']
+    if file.filename == '':
+        print("No file selected")
+        return jsonify({'error': 'No file selected'}), 400
+
+    # Check file extension
+    allowed_extensions = {'png', 'jpg', 'jpeg', 'gif', 'bmp'}
+    if '.' not in file.filename or file.filename.rsplit('.', 1)[1].lower() not in allowed_extensions:
+        print(f"Invalid file format: {file.filename}")
+        return jsonify({'error': 'Invalid file format. Please upload an image (PNG, JPG, JPEG, GIF, BMP).'}), 400
+
+    temp_path = None
+    try:
+        # Create temp directory if it doesn't exist
+        temp_dir = "temp"
+        if not os.path.exists(temp_dir):
+            os.makedirs(temp_dir)
+            print(f"Created temp directory: {temp_dir}")
+
+        # Save the uploaded file temporarily with a unique name
+        temp_path = os.path.join(temp_dir, f"temp_image_{int(time.time())}.png")
+        file.save(temp_path)
+        print(f"Saved uploaded file to {temp_path}")
+
+        # Extract text using Gemini
+        print("Starting text extraction...")
+        extracted_text = extract_text_with_gemini(temp_path)
+        print(f"Text extraction result: {extracted_text[:100]}...")
+
+        # Translate text
+        print("Starting text translation...")
+        translated_text = translate_text(extracted_text)
+        print(f"Translation result: {translated_text[:100]}...")
+
+        return jsonify({
+            'original_text': extracted_text,
+            'translated_text': translated_text
+        })
+    except Exception as e:
+        error_msg = f"Error processing image: {str(e)}"
+        print(error_msg)
+        print(traceback.format_exc())
+        return jsonify({
+            'error': 'An error occurred while processing your image. Please try again.'
+        }), 500
+    finally:
+        # Clean up temporary file if it exists
+        try:
+            if temp_path and os.path.exists(temp_path):
+                os.remove(temp_path)
+                print(f"Removed temporary file: {temp_path}")
+        except Exception as e:
+            print(f"Failed to remove temporary file: {str(e)}")
+
+
+if __name__ == '__main__':
+    # Ensure the template folder exists
+    if not os.path.exists('templates'):
+        os.makedirs('templates')
+        print("Created 'templates' directory. Please place your HTML files here.")
+
+    # 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)))