Update app.py

app.py

@@ -3,9 +3,9 @@ import streamlit as st
 from openai import OpenAI
 import os
 from dotenv import load_dotenv
+import pdf2image
 from PIL import Image
 import io
-import fitz  # PyMuPDF
 import tempfile
 
 # Load environment variables
@@ -14,43 +14,24 @@ load_dotenv()
 client = OpenAI(api_key=os.getenv('OPENAI_API_KEY'))
 
 def convert_pdf_to_images(pdf_file):
-    """Convert PDF to list of images using PyMuPDF"""
-    images = []
+    """Convert PDF to list of images"""
     with tempfile.NamedTemporaryFile(delete=False, suffix='.pdf') as tmp_file:
         tmp_file.write(pdf_file.getvalue())
         pdf_path = tmp_file.name
-        
-        pdf_document = fitz.open(pdf_path)
-        for page_number in range(pdf_document.page_count):
-            page = pdf_document[page_number]
-            pix = page.get_pixmap()
-            img_data = pix.tobytes("png")
-            image = Image.open(io.BytesIO(img_data))
-            images.append(image)
-            
-        pdf_document.close()
+    
+    images = pdf2image.convert_from_path(pdf_path)
     os.unlink(pdf_path)
     return images
 
 def format_response(text):
     """Format the analysis response with clean styling"""
-    formatted_text = """
-    <div style="
-        background-color: white;
-        padding: 20px;
-        border-radius: 5px;
-        font-family: Arial, sans-serif;
-        box-shadow: 0 2px 4px rgba(0,0,0,0.1);
-    ">
-    """
+    formatted_text = ""
     
     # Split into pages
     pages = text.split("Page")
     
     for page_num, page_content in enumerate(pages[1:], 1):  # Skip first empty split
-        # Add page header
-        formatted_text += f'<div style="margin-bottom: 30px;">'
-        formatted_text += f'<h3 style="color: #2c3e50; margin-bottom: 15px;">Page {page_num}</h3>'
+        formatted_text += f'\n### Page {page_num}\n'
         
         # Process each line
         lines = page_content.split('\n')
@@ -62,18 +43,8 @@ def format_response(text):
                 
                 if ':' in line:
                     label, value = line.split(':', 1)
-                    formatted_text += f'<div style="margin-bottom: 10px; display: flex;">'
-                    formatted_text += f'<span style="font-weight: bold; color: #2c3e50; min-width: 200px;">{label.strip()}</span>'
-                    formatted_text += f'<span style="color: #34495e; flex: 1;">{value.strip()}</span>'
-                    formatted_text += '</div>'
-        
-        formatted_text += '</div>'
+                    formatted_text += f'- **{label.strip()}**: {value.strip()}\n'
         
-        # Add separator between pages except for the last page
-        if page_num < len(pages) - 1:
-            formatted_text += '<hr style="border: 1px solid #eee; margin: 20px 0;">'
-    
-    formatted_text += '</div>'
     return formatted_text
 
 def analyze_image(image):
@@ -121,62 +92,34 @@ def analyze_image(image):
 
 def main():
     st.set_page_config(page_title="Document Analysis App", layout="wide")
-    
-    # Custom CSS to set light background and improve button styling
-    st.markdown("""
-        <style>
-        .stApp {
-            background-color: white;
-        }
-        .stButton>button {
-            width: 100%;
-            background-color: #2c3e50;
-            color: white;
-            border: none;
-            padding: 10px 20px;
-            border-radius: 5px;
-            margin-top: 20px;
-            margin-bottom: 20px;
-        }
-        .stButton>button:hover {
-            background-color: #34495e;
-        }
-        .uploadedFile {
-            margin-bottom: 20px;
-        }
-        </style>
-    """, unsafe_allow_html=True)
-    
-    col1, col2, col3 = st.columns([1,2,1])
-    with col2:
-        st.title("Document Analysis App")
-        
-        uploaded_file = st.file_uploader("Upload document (PDF/Image)", type=['pdf', 'png', 'jpg', 'jpeg'])
-        
-        if uploaded_file is not None:
-            if uploaded_file.type == "application/pdf":
-                # Handle PDF
-                with st.spinner("Processing PDF..."):
-                    images = convert_pdf_to_images(uploaded_file)
-                    
-                    if st.button("Extract Information"):
-                        with st.spinner("Analyzing document..."):
-                            all_results = []
-                            for i, image in enumerate(images, 1):
-                                result = analyze_image(image)
-                                all_results.append(f"Page {i} Information:\n{result}")
-                            
-                            combined_results = "\n\n".join(all_results)
-                            st.markdown(format_response(combined_results), unsafe_allow_html=True)
-                            
-            else:
-                # Handle single image
-                image = Image.open(uploaded_file)
-                
+
+    st.title("Document Analysis App")
+    uploaded_file = st.file_uploader("Upload document (PDF/Image)", type=['pdf', 'png', 'jpg', 'jpeg'])
+
+    if uploaded_file is not None:
+        if uploaded_file.type == "application/pdf":
+            # Handle PDF
+            with st.spinner("Processing PDF..."):
+                images = convert_pdf_to_images(uploaded_file)
+
                 if st.button("Extract Information"):
                     with st.spinner("Analyzing document..."):
-                        result = analyze_image(image)
-                        st.markdown(format_response(result), unsafe_allow_html=True)
+                        all_results = []
+                        for i, image in enumerate(images, 1):
+                            result = analyze_image(image)
+                            all_results.append(f"Page {i} Information:\n{result}")
+
+                        combined_results = "\n\n".join(all_results)
+                        st.markdown(format_response(combined_results))
+
+        else:
+            # Handle single image
+            image = Image.open(uploaded_file)
+
+            if st.button("Extract Information"):
+                with st.spinner("Analyzing document..."):
+                    result = analyze_image(image)
+                    st.markdown(format_response(result))
 
 if __name__ == "__main__":
-    main()
+    main()