Update src/utils.py

src/utils.py

@@ -13,6 +13,73 @@ DEFAULT_SESSION_STATE = {
 }
 
 
+def configure_page() -> None:
+    """
+    Configures the Streamlit page.
+    """
+    st.set_page_config(page_title="myRAG", 
+                       layout="wide", 
+                       page_icon=":rocket:")
+
+
+def apply_style():
+    st.markdown("""
+    <style>
+    @import url('https://fonts.googleapis.com/css2?family=Work+Sans:wght@400;700&display=swap');
+    html, body, .stApp,
+    .css-1v3fvcr, .css-ffhzg2, .css-1d391kg,
+    div[data-testid="stMarkdownContainer"],
+    div[data-testid="stText"],
+    div[data-testid="stTextInput"],
+    div[data-testid="stSelectbox"],
+    div[data-testid="stCheckbox"],
+    div[data-testid="stSlider"],
+    label, input, textarea, button, select,
+    .stButton, .stTextInput > div, .stMarkdown, .stCaption,
+    .streamlit-expanderHeader, .st-expander > div,
+    h1, h2, h3, h4, h5, h6,
+    .stMarkdown h1, .stMarkdown h2, .stMarkdown h3 {
+        font-family: 'Work Sans', sans-serif !important;
+    }
+    /* Ensure bold text uses the correct font weight */
+    strong, b, .stMarkdown strong, .stMarkdown b {
+        font-family: 'Work Sans', sans-serif !important;
+        font-weight: 700 !important;
+    }
+    </style>
+    """, unsafe_allow_html=True)
+
+
+def breaks(n=1):
+    """
+    Creates a line break.
+    """
+    if n == 1:
+        st.markdown("<br>",unsafe_allow_html=True)
+    elif n == 2:
+        st.markdown("<br><br>",unsafe_allow_html=True)
+    elif n == 3:
+        st.markdown("<br><br><br>",unsafe_allow_html=True)
+    else:
+        st.markdown("<br><br><br><br>",unsafe_allow_html=True)
+
+
+
+import streamlit as st
+import os
+import base64
+from PIL import Image
+import io
+
+
+DEFAULT_SESSION_STATE = {
+    # PDF Upload
+    'uploaded_files_name': [],
+    'collections_files_name': [],
+    'uploaded_files_raw': [],
+}
+
+
 def configure_page() -> None:
     """
     Configures the Streamlit page.
@@ -66,137 +133,43 @@ def breaks(n=1):
 
 def load_background_image():
     """
-    Debug version to diagnose image loading issues.
+    Loads and displays a background image with an overlaid title.
+    Uses PIL for safer image processing that passes HF security scans.
     """
-    import os
-    import base64
-    from PIL import Image
-    import io
-    
     possible_paths = [
         "../images/image6.jpg",      # Local development (from src/ folder)
         "images/image6.jpg",         # Docker container (from /app)
         "./images/image6.jpg",       # Current directory
     ]
     
-    # Debug: Show current working directory and file structure
-    st.write("**Debug Information:**")
-    st.write(f"Current working directory: `{os.getcwd()}`")
-    
-    # Show what files exist in current directory
-    try:
-        current_files = os.listdir('.')
-        st.write(f"Files in current directory: {current_files}")
-    except:
-        st.write("Cannot list current directory")
-    
-    # Check if images directory exists
-    if os.path.exists('images'):
-        try:
-            image_files = os.listdir('images')
-            st.write(f"Files in images directory: {image_files}")
-        except:
-            st.write("Cannot list images directory")
-    else:
-        st.write("Images directory does not exist")
-    
-    # Test each possible path
-    st.write("**Testing each path:**")
     image_path = None
-    for i, path in enumerate(possible_paths):
-        exists = os.path.exists(path)
-        st.write(f"{i+1}. `{path}` - {'EXISTS' if exists else 'NOT FOUND'}")
-        if exists and not image_path:
+    for path in possible_paths:
+        if os.path.exists(path):
             image_path = path
+            break
     
     if not image_path:
         st.error("Could not find image6.jpg in any expected location")
-        st.write("**Suggestion:** Make sure image6.jpg is in the 'images' folder")
-        return
-    
-    # Debug: File details
-    st.write(f"**Using image:** `{image_path}`")
-    
-    try:
-        file_size = os.path.getsize(image_path)
-        st.write(f"File size: {file_size:,} bytes")
-        
-        # Check file signature
-        with open(image_path, 'rb') as f:
-            first_bytes = f.read(10)
-            st.write(f"First 10 bytes: {first_bytes}")
-            
-        # Check if it's a valid JPEG (should start with FF D8)
-        if first_bytes[:2] == b'\xff\xd8':
-            st.write("Valid JPEG header detected")
-        else:
-            st.write("Not a standard JPEG file (might still work)")
-            
-    except Exception as e:
-        st.error(f"Error reading file details: {e}")
         return
     
-    # Test 1: Try Streamlit's native image display
-    st.write("**Test 1: Streamlit native image display**")
-    try:
-        st.image(image_path, caption="Direct Streamlit image test", width=300)
-        st.write("Streamlit can display the image")
-    except Exception as e:
-        st.error(f"Streamlit image display failed: {e}")
-    
-    # Test 2: Try PIL image opening
-    st.write("**Test 2: PIL image processing**")
     try:
+        # Use PIL for safe image processing (avoids HF pickle scanner issues)
         img = Image.open(image_path)
-        st.write(f"PIL opened successfully")
-        st.write(f"Format: {img.format}")
-        st.write(f"Size: {img.size}")
-        st.write(f"Mode: {img.mode}")
         
-        # Test 3: Try converting to PNG
-        st.write("**Test 3: Converting to PNG in memory**")
+        # Convert to PNG format in memory for consistency
         img_buffer = io.BytesIO()
         img.save(img_buffer, format='PNG')
         img_bytes = img_buffer.getvalue()
-        st.write(f"PNG conversion successful, size: {len(img_bytes):,} bytes")
         
-        # Test 4: Try base64 encoding
-        st.write("**Test 4: Base64 encoding**")
+        # Encode to base64
         base64_image = base64.b64encode(img_bytes).decode()
-        st.write(f"Base64 encoding successful, length: {len(base64_image):,} characters")
-        st.write(f"First 50 chars: `{base64_image[:50]}...`")
-        
-        # Test 5: Try simple CSS background
-        st.write("**Test 5: Simple CSS background test**")
-        st.markdown(
-            f"""
-            <div style="
-                background-image: url('data:image/png;base64,{base64_image}');
-                background-size: cover;
-                background-position: center;
-                height: 200px;
-                width: 100%;
-                border: 3px solid red;
-                color: white;
-                display: flex;
-                align-items: center;
-                justify-content: center;
-                font-size: 24px;
-                font-weight: bold;
-                text-shadow: 2px 2px 4px rgba(0,0,0,0.8);
-            ">
-                SIMPLE CSS TEST
-            </div>
-            """,
-            unsafe_allow_html=True
-        )
         
-        # Test 6: Try your original complex CSS
-        st.write("**Test 6: Your original CSS styling**")
+        # Inject CSS for the background and title overlay
         st.markdown(
             f"""
             <style>
-            .debug-bg-container {{
+            /* Background container with image */
+            .bg-container {{
                 position: relative;
                 background-image: url("data:image/png;base64,{base64_image}");
                 background-size: cover;
@@ -207,10 +180,10 @@ def load_background_image():
                 margin: 0 auto;
                 filter: contrast(110%) brightness(210%);
                 border-radius: 100px;
-                overflow: hidden;
-                border: 2px solid blue;
+                overflow: hidden;  
             }}
-            .debug-bg-container::after {{
+            /* Overlay for dimming effect */
+            .bg-container::after {{
                 content: '';
                 position: absolute;
                 top: 0;
@@ -220,69 +193,63 @@ def load_background_image():
                 background-color: rgba(20, 10, 20, 0.44);
                 z-index: 1;
             }}
-            .debug-overlay-title {{
+            /* Overlay title styling */
+            .overlay-title {{
                 position: absolute;
                 top: 50%;
                 left: 50%;
                 transform: translate(-50%, -50%);
-                color: white;
+                color: black;
                 font-size: 50px;
                 font-weight: bold;
-                text-shadow: 2px 2px 4px rgba(0, 0, 0, 1);
+                text-shadow: 1px 1px 3px rgba(255, 255, 255, .0);
                 text-align: center;
                 z-index: 2;
             }}
             </style>
-            <div class="debug-bg-container">
-                <div class="debug-overlay-title">Mistral-RAG</div>
+            """,
+            unsafe_allow_html=True
+        )
+
+        # Create the background container with an overlaid title
+        st.markdown(
+            """
+            <div class="bg-container">
+                <div class="overlay-title">Mistral-RAG</div>
             </div>
             """,
             unsafe_allow_html=True
         )
         
-        st.write("**All tests completed!** If you can see the images above, the problem might be with your original CSS classes or styling conflicts.")
-        
-    except Exception as pil_error:
-        st.error(f"PIL processing failed: {pil_error}")
-        
-        st.write("**Test: Direct file base64 encoding (bypass PIL)**")
-        try:
-            with open(image_path, 'rb') as f:
-                img_data = f.read()
-                img_base64 = base64.b64encode(img_data).decode()
-                
-            st.write(f"Direct base64 encoding successful")
-            st.markdown(
-                f"""
-                <div style="
-                    background-image: url('data:image/jpeg;base64,{img_base64}');
-                    background-size: cover;
-                    background-position: center;
-                    height: 200px;
-                    width: 100%;
-                    border: 3px solid green;
-                    color: white;
-                    display: flex;
-                    align-items: center;
-                    justify-content: center;
-                    font-size: 24px;
-                    font-weight: bold;
-                ">
-                    DIRECT FILE ENCODING TEST
-                </div>
-                """,
-                unsafe_allow_html=True
-            )
-            
-        except Exception as direct_error:
-            st.error(f"Direct file encoding also failed: {direct_error}")
+    except Exception as e:
+        st.error(f"Error loading background image: {e}")
+
+
+def initialise_session_state():
+    """
+    Initializes the session state variables if not already set.
+    """
+    for key, default_val in DEFAULT_SESSION_STATE.items():
+        if key not in st.session_state:
+            st.session_state[key] = default_val
+
+
+def file_uploader():
+    uploaded_files = st.file_uploader(
+        "",
+        type=["txt", "pdf"], 
+        accept_multiple_files=True)  
     
-    st.write("---")
-    st.write("**Next Steps:**")
-    st.write("1. If any of the tests above show images, your image file is fine")
-    st.write("2. If no images show, try uploading a different image file")
-    st.write("3. If simple CSS works but complex CSS doesn't, there might be a styling conflict")
-    st.write("4. Check browser console for any CSS errors")
+    if uploaded_files:  # Check if list is not empty
+        for file in uploaded_files:  # Process each file
+            if file.name not in st.session_state.uploaded_files_name:
+                # Append to session state lists safely
+                st.session_state.uploaded_files_name.append(file.name)
+                st.session_state.uploaded_files_raw.append(file)
+                st.success(f"Added new file: {file.name}")
+   
+    else:
+        st.info("Please upload a PDF file to proceed.")
 
 
 def initialise_session_state():