Create app.py

app.py

@@ -0,0 +1,89 @@
+from fastapi import FastAPI, File, UploadFile, HTTPException
+from fastapi.responses import JSONResponse
+import os
+import shutil
+from langchain_google_genai import ChatGoogleGenerativeAI, GoogleGenerativeAIEmbeddings
+from langchain_community.vectorstores import FAISS
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain_community.document_loaders import PyPDFLoader
+from langchain.chains import RetrievalQA
+
+app = FastAPI(title="RAG Chatbot API")
+
+# Ensure directories exist
+os.makedirs("documents", exist_ok=True)
+os.makedirs("vectorstore", exist_ok=True)
+
+# Initialize Gemini LLM
+llm = ChatGoogleGenerativeAI(
+    model="gemini-1.5-flash",
+    google_api_key=os.getenv("GOOGLE_API_KEY")
+)
+
+# Initialize embeddings
+embeddings = GoogleGenerativeAIEmbeddings(
+    model="models/embedding-001",
+    google_api_key=os.getenv("GOOGLE_API_KEY")
+)
+
+# Path for vector store
+VECTOR_STORE_PATH = "vectorstore/index"
+
+def process_pdf(pdf_path):
+    """Process and index a PDF document."""
+    try:
+        loader = PyPDFLoader(pdf_path)
+        documents = loader.load()
+        text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)
+        texts = text_splitter.split_documents(documents)
+        if os.path.exists(VECTOR_STORE_PATH):
+            vector_store = FAISS.load_local(VECTOR_STORE_PATH, embeddings, allow_dangerous_deserialization=True)
+            vector_store.add_documents(texts)
+        else:
+            vector_store = FAISS.from_documents(texts, embeddings)
+        vector_store.save_local(VECTOR_STORE_PATH)
+        return {"status": "Document processed and indexed successfully"}
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=f"Error processing document: {str(e)}")
+
+def answer_query(query):
+    """Answer a query using the RAG pipeline."""
+    if not os.path.exists(VECTOR_STORE_PATH):
+        return {"error": "No documents indexed yet. Please upload a document first."}
+    try:
+        vector_store = FAISS.load_local(VECTOR_STORE_PATH, embeddings, allow_dangerous_deserialization=True)
+        qa_chain = RetrievalQA.from_chain_type(
+            llm=llm,
+            chain_type="stuff",
+            retriever=vector_store.as_retriever(search_kwargs={"k": 3}),
+            return_source_documents=True
+        )
+        result = qa_chain({"query": query})
+        return {
+            "answer": result["result"],
+            "source_documents": [doc.page_content[:200] for doc in result["source_documents"]]
+        }
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=f"Error answering query: {str(e)}")
+
+@app.post("/upload-document")
+async def upload_document(file: UploadFile = File(...)):
+    """API to upload and process a PDF document."""
+    if not file.filename.endswith(".pdf"):
+        raise HTTPException(status_code=400, detail="Only PDF files are allowed")
+    file_path = f"documents/{file.filename}"
+    with open(file_path, "wb") as buffer:
+        shutil.copyfileobj(file.file, buffer)
+    result = process_pdf(file_path)
+    return JSONResponse(content=result, status_code=200)
+
+@app.post("/ask-question")
+async def ask_question(query: str):
+    """API to answer a query based on indexed documents."""
+    result = answer_query(query)
+    return JSONResponse(content=result, status_code=200)
+
+@app.get("/health")
+async def health_check():
+    """Health check endpoint."""
+    return {"status": "API is running"}