Added a comment (#2)

- Added a comment (dd04eeacaea2330910f5fff63845c62d6b43abf4)


Co-authored-by: Tan Zi Xuan <[email protected]>

app.py

@@ -1,218 +1,219 @@
-"""
-Streamlit web app for Singtel Bill Scanner
-This creates a user-friendly interface for the bill scanner
-"""
-
-import streamlit as st
-from PIL import Image
-import io
-import base64
-
-# Only import heavy libraries when needed
-@st.cache_resource
-def load_ocr_model():
-    """Load the OCR model (cached for performance)"""
-    from transformers import pipeline
-    return pipeline("image-to-text", model="microsoft/trocr-base-handwritten")
-
-def process_bill_image(image, pipe):
-    """Process the uploaded bill image"""
-    try:
-        # Process with TrOCR
-        result = pipe(image)
-        extracted_text = result[0]['generated_text']
-        
-        # Simple parsing
-        import re
-        
-        # Extract key information
-        parsed_data = {
-            'raw_text': extracted_text,
-            'total_amount': None,
-            'due_date': None,
-            'account_number': None,
-            'services': []
-        }
-        
-        # Look for total amount
-        amount_patterns = [
-            r'Total[:\s]*\$?([0-9,]+\.?[0-9]*)',
-            r'Amount Due[:\s]*\$?([0-9,]+\.?[0-9]*)',
-            r'\$([0-9,]+\.?[0-9]*)',
-        ]
-        
-        for pattern in amount_patterns:
-            match = re.search(pattern, extracted_text, re.IGNORECASE)
-            if match:
-                try:
-                    parsed_data['total_amount'] = float(match.group(1).replace(',', ''))
-                    break
-                except ValueError:
-                    continue
-        
-        # Look for due date
-        date_patterns = [
-            r'Due[:\s]*(\d{1,2}[/-]\d{1,2}[/-]\d{2,4})',
-            r'(\d{1,2}[/-]\d{1,2}[/-]\d{2,4})',
-        ]
-        
-        for pattern in date_patterns:
-            match = re.search(pattern, extracted_text, re.IGNORECASE)
-            if match:
-                parsed_data['due_date'] = match.group(1)
-                break
-        
-        # Look for account number
-        account_patterns = [
-            r'Account[:\s]*([0-9A-Z-]+)',
-            r'A/C[:\s]*([0-9A-Z-]+)',
-        ]
-        
-        for pattern in account_patterns:
-            match = re.search(pattern, extracted_text, re.IGNORECASE)
-            if match:
-                parsed_data['account_number'] = match.group(1).strip()
-                break
-        
-        return parsed_data
-        
-    except Exception as e:
-        st.error(f"Error processing image: {e}")
-        return None
-
-def main():
-    st.set_page_config(
-        page_title="Singtel Bill Scanner",
-        page_icon="📱",
-        layout="wide"
-    )
-    
-    st.title("📱 Singtel Bill Scanner")
-    st.markdown("### AI-Powered OCR for Singtel Bills")
-    
-    st.markdown("""
-    Upload an image of your Singtel bill and extract key information automatically using AI!
-    
-    **Features:**
-    - 🔍 Extract text from handwritten and printed bills
-    - 💰 Identify total amounts and charges
-    - 📅 Find due dates
-    - 🔢 Extract account numbers
-    """)
-    
-    # Sidebar with instructions
-    with st.sidebar:
-        st.markdown("### 📋 Instructions")
-        st.markdown("""
-        1. **Take a clear photo** of your Singtel bill
-        2. **Upload the image** using the file uploader
-        3. **Wait for processing** (may take a few seconds)
-        4. **Review extracted information**
-        
-        **Tips for better results:**
-        - Use good lighting
-        - Keep the image straight
-        - Ensure text is clearly visible
-        - Avoid shadows and glare
-        """)
-        
-        st.markdown("### 🔧 Technical Details")
-        st.markdown("""
-        - **Model**: Microsoft TrOCR
-        - **Accuracy**: High for clear images
-        - **Processing**: ~3-5 seconds
-        - **Privacy**: Images not stored
-        """)
-    
-    # Main content area
-    col1, col2 = st.columns([1, 1])
-    
-    with col1:
-        st.markdown("### 📤 Upload Bill Image")
-        
-        uploaded_file = st.file_uploader(
-            "Choose a bill image...",
-            type=['png', 'jpg', 'jpeg'],
-            help="Upload a clear image of your Singtel bill"
-        )
-        
-        if uploaded_file is not None:
-            # Display the uploaded image
-            image = Image.open(uploaded_file)
-            st.image(image, caption="Uploaded Bill", use_column_width=True)
-            
-            # Process button
-            if st.button("🔍 Extract Information", type="primary"):
-                with st.spinner("Processing image with AI..."):
-                    # Load model
-                    pipe = load_ocr_model()
-                    
-                    # Process image
-                    result = process_bill_image(image, pipe)
-                    
-                    if result:
-                        st.session_state['processing_result'] = result
-                        st.success("✅ Processing completed!")
-    
-    with col2:
-        st.markdown("### 📊 Extracted Information")
-        
-        if 'processing_result' in st.session_state:
-            result = st.session_state['processing_result']
-            
-            # Display parsed information
-            st.markdown("#### 💰 Bill Summary")
-            
-            col_a, col_b = st.columns(2)
-            
-            with col_a:
-                if result['total_amount']:
-                    st.metric("Total Amount", f"${result['total_amount']:.2f}")
-                else:
-                    st.metric("Total Amount", "Not detected")
-            
-            with col_b:
-                if result['due_date']:
-                    st.metric("Due Date", result['due_date'])
-                else:
-                    st.metric("Due Date", "Not detected")
-            
-            if result['account_number']:
-                st.markdown(f"**Account Number:** {result['account_number']}")
-            else:
-                st.markdown("**Account Number:** Not detected")
-            
-            # Raw extracted text
-            st.markdown("#### 📝 Raw Extracted Text")
-            st.text_area(
-                "Full text extracted from image:",
-                value=result['raw_text'],
-                height=150,
-                disabled=True
-            )
-            
-            # Download option
-            st.markdown("#### 💾 Export Data")
-            import json
-            json_data = json.dumps(result, indent=2)
-            st.download_button(
-                label="📄 Download as JSON",
-                data=json_data,
-                file_name="bill_data.json",
-                mime="application/json"
-            )
-            
-        else:
-            st.info("👆 Upload an image and click 'Extract Information' to see results here")
-    
-    # Footer
-    st.markdown("---")
-    st.markdown("""
-    <div style='text-align: center'>
-        <p>Built with ❤️ using Streamlit and Hugging Face Transformers</p>
-        <p>🤖 Powered by Microsoft TrOCR | 🔒 Your images are processed locally and not stored</p>
-    </div>
-    """, unsafe_allow_html=True)
-
-if __name__ == "__main__":
-    main()
+"""
+Streamlit web app for Singtel Bill Scanner
+This creates a user-friendly interface for the bill scanner
+"""
+
+import streamlit as st
+from PIL import Image
+import io
+import base64
+
+# Only import heavy libraries when needed
+@st.cache_resource
+def load_ocr_model():
+    """Load the OCR model (cached for performance)"""
+    from transformers import pipeline
+    return pipeline("image-to-text", model="microsoft/trocr-base-handwritten")
+
+def process_bill_image(image, pipe):
+    """Process the uploaded bill image"""
+    try:
+        # Process with TrOCR
+        result = pipe(image)
+        extracted_text = result[0]['generated_text']
+        
+        # Simple parsing
+        import re
+        #added something
+        
+        # Extract key information
+        parsed_data = {
+            'raw_text': extracted_text,
+            'total_amount': None,
+            'due_date': None,
+            'account_number': None,
+            'services': []
+        }
+        
+        # Look for total amount
+        amount_patterns = [
+            r'Total[:\s]*\$?([0-9,]+\.?[0-9]*)',
+            r'Amount Due[:\s]*\$?([0-9,]+\.?[0-9]*)',
+            r'\$([0-9,]+\.?[0-9]*)',
+        ]
+        
+        for pattern in amount_patterns:
+            match = re.search(pattern, extracted_text, re.IGNORECASE)
+            if match:
+                try:
+                    parsed_data['total_amount'] = float(match.group(1).replace(',', ''))
+                    break
+                except ValueError:
+                    continue
+        
+        # Look for due date
+        date_patterns = [
+            r'Due[:\s]*(\d{1,2}[/-]\d{1,2}[/-]\d{2,4})',
+            r'(\d{1,2}[/-]\d{1,2}[/-]\d{2,4})',
+        ]
+        
+        for pattern in date_patterns:
+            match = re.search(pattern, extracted_text, re.IGNORECASE)
+            if match:
+                parsed_data['due_date'] = match.group(1)
+                break
+        
+        # Look for account number
+        account_patterns = [
+            r'Account[:\s]*([0-9A-Z-]+)',
+            r'A/C[:\s]*([0-9A-Z-]+)',
+        ]
+        
+        for pattern in account_patterns:
+            match = re.search(pattern, extracted_text, re.IGNORECASE)
+            if match:
+                parsed_data['account_number'] = match.group(1).strip()
+                break
+        
+        return parsed_data
+        
+    except Exception as e:
+        st.error(f"Error processing image: {e}")
+        return None
+
+def main():
+    st.set_page_config(
+        page_title="Singtel Bill Scanner",
+        page_icon="📱",
+        layout="wide"
+    )
+    
+    st.title("📱 Singtel Bill Scanner")
+    st.markdown("### AI-Powered OCR for Singtel Bills")
+    
+    st.markdown("""
+    Upload an image of your Singtel bill and extract key information automatically using AI!
+    
+    **Features:**
+    - 🔍 Extract text from handwritten and printed bills
+    - 💰 Identify total amounts and charges
+    - 📅 Find due dates
+    - 🔢 Extract account numbers
+    """)
+    
+    # Sidebar with instructions
+    with st.sidebar:
+        st.markdown("### 📋 Instructions")
+        st.markdown("""
+        1. **Take a clear photo** of your Singtel bill
+        2. **Upload the image** using the file uploader
+        3. **Wait for processing** (may take a few seconds)
+        4. **Review extracted information**
+        
+        **Tips for better results:**
+        - Use good lighting
+        - Keep the image straight
+        - Ensure text is clearly visible
+        - Avoid shadows and glare
+        """)
+        
+        st.markdown("### 🔧 Technical Details")
+        st.markdown("""
+        - **Model**: Microsoft TrOCR
+        - **Accuracy**: High for clear images
+        - **Processing**: ~3-5 seconds
+        - **Privacy**: Images not stored
+        """)
+    
+    # Main content area
+    col1, col2 = st.columns([1, 1])
+    
+    with col1:
+        st.markdown("### 📤 Upload Bill Image")
+        
+        uploaded_file = st.file_uploader(
+            "Choose a bill image...",
+            type=['png', 'jpg', 'jpeg'],
+            help="Upload a clear image of your Singtel bill"
+        )
+        
+        if uploaded_file is not None:
+            # Display the uploaded image
+            image = Image.open(uploaded_file)
+            st.image(image, caption="Uploaded Bill", use_column_width=True)
+            
+            # Process button
+            if st.button("🔍 Extract Information", type="primary"):
+                with st.spinner("Processing image with AI..."):
+                    # Load model
+                    pipe = load_ocr_model()
+                    
+                    # Process image
+                    result = process_bill_image(image, pipe)
+                    
+                    if result:
+                        st.session_state['processing_result'] = result
+                        st.success("✅ Processing completed!")
+    
+    with col2:
+        st.markdown("### 📊 Extracted Information")
+        
+        if 'processing_result' in st.session_state:
+            result = st.session_state['processing_result']
+            
+            # Display parsed information
+            st.markdown("#### 💰 Bill Summary")
+            
+            col_a, col_b = st.columns(2)
+            
+            with col_a:
+                if result['total_amount']:
+                    st.metric("Total Amount", f"${result['total_amount']:.2f}")
+                else:
+                    st.metric("Total Amount", "Not detected")
+            
+            with col_b:
+                if result['due_date']:
+                    st.metric("Due Date", result['due_date'])
+                else:
+                    st.metric("Due Date", "Not detected")
+            
+            if result['account_number']:
+                st.markdown(f"**Account Number:** {result['account_number']}")
+            else:
+                st.markdown("**Account Number:** Not detected")
+            
+            # Raw extracted text
+            st.markdown("#### 📝 Raw Extracted Text")
+            st.text_area(
+                "Full text extracted from image:",
+                value=result['raw_text'],
+                height=150,
+                disabled=True
+            )
+            
+            # Download option
+            st.markdown("#### 💾 Export Data")
+            import json
+            json_data = json.dumps(result, indent=2)
+            st.download_button(
+                label="📄 Download as JSON",
+                data=json_data,
+                file_name="bill_data.json",
+                mime="application/json"
+            )
+            
+        else:
+            st.info("👆 Upload an image and click 'Extract Information' to see results here")
+    
+    # Footer
+    st.markdown("---")
+    st.markdown("""
+    <div style='text-align: center'>
+        <p>Built with ❤️ using Streamlit and Hugging Face Transformers</p>
+        <p>🤖 Powered by Microsoft TrOCR | 🔒 Your images are processed locally and not stored</p>
+    </div>
+    """, unsafe_allow_html=True)
+
+if __name__ == "__main__":
+    main()