New regex matching for years and fixed UI issues

src/document_processor.py

@@ -126,9 +126,20 @@ class DocumentProcessor:
                             metadata["authors"] = possible_authors
 
                     # crude heuristic: find year (e.g., 2023, 2024)
-                    year_match = re.search(r"\b(19|20)\d{2}\b", first_page)
-                    if year_match:
-                        metadata["year"] = year_match.group(0)
+                    year_patterns = [
+                        r"\b(19|20)\d{2}\b",  # Basic year
+                        r"©\s*(19|20)\d{2}",  # Copyright year
+                        r"\((19|20)\d{2}\)",  # Year in parentheses
+                        r"(19|20)\d{2}[,.)]",  # Year followed by comma/period
+                    ]
+
+                    for pattern in year_patterns:
+                        year_match = re.search(pattern, first_page)
+                        if year_match:
+                            year_text = re.search(r"(19|20)\d{2}", year_match.group(0))
+                            if year_text:
+                                metadata["year"] = year_text.group(0)
+                                break
 
                 except Exception:
                     pass

streamlit_app.py

@@ -39,7 +39,7 @@ with st.sidebar:
         "What is paternalism in bioethics?",
         "What are the ethical issues with genetic testing?",
         "How should AI bias in healthcare be addressed?",
-        "What are arguments raised by antinatalists?",
+        "Is antinatalism rational?",
         "What does it mean for women’s autonomy to be respected?"
     ]
 
@@ -119,32 +119,6 @@ if question and st.session_state.query_count < 30:
             # Display the final answer
             response_placeholder.markdown(answer)
 
-
-
-        """
-        with debug_col:
-            st.markdown("### 🔍 Debug Info")
-
-            # Show search results info
-            if 'bot' in st.session_state:
-                # Get search results for debug display
-                search_results = st.session_state.bot.vector_store.search(question, k=3)
-                with st.expander("📊 Search Results", expanded=True):
-                    for i, r in enumerate(search_results):
-                        st.write(f"**Result {i + 1}** (Score: {r.get('similarity_score', 0):.3f})")
-                        st.write(f"Source: {r['metadata'].get('filename', 'Unknown')}")
-                        st.write(f"Preview: {r['content'][:200]}...")
-                        st.write("---")
-
-            # Show response metadata
-            st.metric("Response Time", f"{response_time:.2f}s")
-            st.metric("Model", "GPT-4o-mini")
-
-            # Show conversation history count
-            if hasattr(st.session_state.bot, 'history'):
-                st.metric("Conversation Turn", len(st.session_state.bot.history))
-        """
-
         # Show source information
         with st.expander("📚 About the Sources"):
             st.markdown("""