Update src/streamlit_app.py

src/streamlit_app.py

@@ -2,53 +2,70 @@ import streamlit as st
 import os
 from rag_engine import ProjectRAGEngine
 
-# Page configuration to ensure wide display for citations [cite: 31]
-st.set_page_config(page_title="IIR Project Analyzer", layout="wide")
+# -------------------------------
+# Page configuration
+# -------------------------------
+st.set_page_config(
+    page_title="IIR Project Analyzer",
+    layout="wide"
+)
+
 st.title("📂 Industrial Project Report Analyzer")
 
-# 1. API Key Security Logic [cite: 21]
-# Checks Hugging Face Secrets first, then falls back to user input
-api_key = os.environ.get("OPENAI_API_KEY")
-if not api_key:
-    api_key = st.sidebar.text_input("Enter OpenAI API Key", type="password")
+# -------------------------------
+# Environment validation
+# -------------------------------
+openai_key = os.getenv("OPENAI_API_KEY")
+openrouter_key = os.getenv("OPENROUTER_API_KEY")
 
-if not api_key:
-    st.warning("Please provide an OpenAI API Key in the sidebar to proceed.")
+if not openai_key or not openrouter_key:
+    st.error(
+        "Missing API keys.\n\n"
+        "Please set the following environment variables:\n"
+        "- OPENAI_API_KEY (for embeddings)\n"
+        "- OPENROUTER_API_KEY (for LLM)"
+    )
     st.stop()
 
-# 2. Session Persistence [cite: 8, 44]
+# -------------------------------
+# Session persistence
+# -------------------------------
 if "engine" not in st.session_state:
-    st.session_state.engine = ProjectRAGEngine(api_key)
+    st.session_state.engine = ProjectRAGEngine()
 
-# 3. Document Management Interface [cite: 6, 29]
+# -------------------------------
+# File upload
+# -------------------------------
 uploaded_files = st.sidebar.file_uploader(
-    "Upload Project PDFs", 
-    type="pdf", 
+    "Upload Project PDFs",
+    type="pdf",
     accept_multiple_files=True
 )
 
 if uploaded_files:
-    # Check if files are already processed to prevent redundant API calls [cite: 8]
-    if "processed_files" not in st.session_state or len(st.session_state.processed_files) != len(uploaded_files):
+    if (
+        "processed_files" not in st.session_state
+        or set(st.session_state.processed_files) != {f.name for f in uploaded_files}
+    ):
         with st.spinner("Analyzing project reports..."):
-            # Ensure the temp directory has write permissions to avoid 403/Permission errors
             temp_dir = "temp"
-            if not os.path.exists(temp_dir): 
-                os.makedirs(temp_dir, mode=0o777)
-            
+            os.makedirs(temp_dir, exist_ok=True)
+
             paths = []
             for f in uploaded_files:
                 path = os.path.join(temp_dir, f.name)
                 with open(path, "wb") as out:
                     out.write(f.getbuffer())
                 paths.append(path)
-            
-            # Simultaneous querying setup [cite: 7]
+
             st.session_state.engine.process_documents(paths)
             st.session_state.processed_files = [f.name for f in uploaded_files]
+
         st.success("Reports indexed. Ready for queries.")
 
-# 4. Question Answering Interface [cite: 30-31]
+# -------------------------------
+# Chat interface
+# -------------------------------
 query = st.chat_input("Ex: 'Compare the budgets of these projects'")
 
 if query:
@@ -56,15 +73,18 @@ if query:
         st.write(query)
 
     with st.chat_message("assistant"):
-        # Handle cases with no documents [cite: 25]
         answer, sources = st.session_state.engine.get_answer(query)
-        st.markdown(f"### Response\n{answer}") # cite: 16
-        
-        # Mandatory Source Attribution Display [cite: 11, 17, 31]
+
+        st.markdown("### Response")
+        st.write(answer)
+
         if sources:
-            with st.expander("View Source Attribution & Quotes"): # cite: 18
+            with st.expander("📌 Source Attribution & Quotes"):
                 for idx, s in enumerate(sources):
-                    doc_name = os.path.basename(s['metadata']['source'])
-                    page_num = s['metadata']['page'] + 1 # PDF index starts at 0
-                    st.markdown(f"**Source {idx+1}:** {doc_name} (Page {page_num})")
-                    st.caption(f"Verbatim Quote: \"{s['content'][:300]}...\"")
+                    doc_name = os.path.basename(s["metadata"]["source"])
+                    page_num = s["metadata"]["page"] + 1
+
+                    st.markdown(
+                        f"**Source {idx + 1}:** {doc_name} (Page {page_num})"
+                    )
+                    st.caption(f"\"{s['content'][:300]}...\"")