Update app.py

app.py

@@ -7,48 +7,57 @@ from langchain_community.embeddings import HuggingFaceEmbeddings
 from langchain_community.vectorstores import FAISS
 import google.generativeai as genai
 
-# Folder where FAISS index will be saved
+# Path to save vector index
 INDEX_DIR = "rag_multi_pdf_index"
 
-# Build knowledge base from multiple PDFs
+# Step 1: Create knowledge base from PDFs
 def create_knowledge_base(pdf_files: List[gr.File]) -> str:
+    if not pdf_files:
+        return "❌ No PDFs uploaded."
+
     all_chunks = []
     splitter = RecursiveCharacterTextSplitter(chunk_size=500, chunk_overlap=100)
 
     for file in pdf_files:
         loader = PyPDFLoader(file.name)
-        docs = loader.load()
-        chunks = splitter.split_documents(docs)
-        all_chunks.extend(chunks)
+        try:
+            docs = loader.load()
+            chunks = splitter.split_documents(docs)
+            all_chunks.extend(chunks)
+        except Exception as e:
+            return f"❌ Error reading {file.name}: {str(e)}"
+
+    if not all_chunks:
+        return "❌ No content extracted from PDFs."
 
     embeddings = HuggingFaceEmbeddings(model_name="all-MiniLM-L6-v2")
     vectorstore = FAISS.from_documents(all_chunks, embeddings)
     vectorstore.save_local(INDEX_DIR)
 
-    return f"✅ Knowledge base created from {len(pdf_files)} PDFs."
+    return f"✅ Knowledge base created with {len(all_chunks)} chunks from {len(pdf_files)} PDFs."
 
-# Load existing vector index
+# Step 2: Load vectorstore
 def load_vectorstore() -> FAISS:
     embeddings = HuggingFaceEmbeddings(model_name="all-MiniLM-L6-v2")
     return FAISS.load_local(INDEX_DIR, embeddings)
 
-# Ask a question using Gemini + vector search
+# Step 3: Ask question via Gemini using retrieved context
 def chat_with_rag(api_key: str, question: str) -> str:
     if not api_key or not api_key.startswith("AI"):
-        return "❌ Invalid Gemini API Key. Please enter a valid key starting with 'AI'."
+        return "❌ Invalid Gemini API Key. It should start with 'AI'."
 
     try:
         genai.configure(api_key=api_key)
         model = genai.GenerativeModel("gemini-pro")
     except Exception as e:
-        return f"❌ Failed to configure Gemini: {str(e)}"
+        return f"❌ Gemini configuration error: {str(e)}"
 
     try:
         vs = load_vectorstore()
         top_docs = vs.similarity_search(question, k=3)
         context = "\n\n".join([doc.page_content for doc in top_docs])
     except Exception as e:
-        return f"❌ Error retrieving context: {str(e)}"
+        return f"❌ Error loading vectorstore or retrieving context: {str(e)}"
 
     prompt = f"""Use the following context to answer the question:\n\n{context}\n\nQuestion: {question}"""
 
@@ -56,26 +65,26 @@ def chat_with_rag(api_key: str, question: str) -> str:
         response = model.generate_content(prompt)
         return response.text
     except Exception as e:
-        return f"❌ Gemini Error: {str(e)}"
+        return f"❌ Gemini error: {str(e)}"
+
+# Step 4: Gradio UI
+with gr.Blocks(title="📚 RAG Q&A with Gemini") as demo:
+    gr.Markdown("## 📄 Upload multiple PDFs → 🧠 Build Knowledge Base → 🤖 Ask Questions with Gemini")
 
-# Gradio UI
-with gr.Blocks(title="📚 RAG Chat with Gemini (Multi-PDF)") as demo:
-    gr.Markdown("## 🧠 Upload Multiple PDFs & Chat Using Gemini")
+    api_key = gr.Textbox(label="🔐 Gemini API Key", placeholder="Enter your Gemini API Key", type="password")
 
-    api_key = gr.Textbox(label="🔐 Gemini API Key", placeholder="Paste your API key here", type="password")
-    
-    with gr.Row():
-        pdfs = gr.File(label="📂 Upload PDFs", file_types=[".pdf"], file_count="multiple")
-        create_btn = gr.Button("📄 Create Knowledge Base")
-        kb_status = gr.Textbox(label="📦 Knowledge Base Status", interactive=False)
+    pdfs = gr.File(label="📂 Upload PDFs", file_types=[".pdf"], file_count="multiple")
+    create_btn = gr.Button("📄 Create Knowledge Base")
+    kb_status = gr.Textbox(label="📦 Knowledge Base Status", interactive=False)
 
     create_btn.click(fn=create_knowledge_base, inputs=[pdfs], outputs=[kb_status])
 
-    question = gr.Textbox(label="❓ Your Question", placeholder="Ask something based on your PDFs")
-    answer = gr.Textbox(label="💬 Gemini Answer", lines=10)
+    question = gr.Textbox(label="❓ Ask a Question")
+    answer = gr.Textbox(label="💬 Gemini Answer", lines=10, interactive=False)
     ask_btn = gr.Button("🚀 Ask")
 
     ask_btn.click(fn=chat_with_rag, inputs=[api_key, question], outputs=[answer])
 
+# Step 5: Launch app
 if __name__ == "__main__":
     demo.launch()