app.py

import os
import gradio as gr
from langchain.chat_models import ChatOpenAI
from langchain.document_loaders import WikipediaLoader
from langchain.text_splitter import CharacterTextSplitter
from langchain.embeddings import HuggingFaceEmbeddings
from langchain.vectorstores import FAISS
from langchain.chains import ConversationalRetrievalChain
from langchain.callbacks.base import BaseCallbackHandler
from langchain.memory import ConversationBufferMemory
import traceback

# --- Memory Cache ---
class MemoryCache:
    def __init__(self):
        self.cache = {}

    def get(self, query: str):
        if query in self.cache:
            print(f"Cache hit: {query}")
        return self.cache.get(query)

    def set(self, query: str, response: str):
        print(f"Saving to cache: {query}")
        self.cache[query] = response

# --- Callback Logger ---
class LoggingCallbackHandler(BaseCallbackHandler):
    def __init__(self):
        self.logs = []

    def on_chain_start(self, serialized, inputs, **kwargs):
        self.logs.append(f"Chain start. Inputs: {inputs}")
        print(f"Chain start. Inputs: {inputs}")

    def on_chain_end(self, outputs, **kwargs):
        self.logs.append(f"Chain end. Outputs: {outputs}")
        print(f"Chain end. Outputs: {outputs}")

    def on_retriever_start(self, *args, **kwargs):
        self.logs.append("Retrieval start.")
        print("Retrieval start.")

    def on_retriever_end(self, *args, **kwargs):
        self.logs.append("Retrieval end.")
        print("Retrieval end.")

    def on_llm_start(self, *args, **kwargs):
        self.logs.append("LLM start.")
        print("LLM start.")

    def on_llm_end(self, result, *args, **kwargs):
        try:
            final_text = result.generations[0][0].text
            self.logs.append(f"LLM end. Text: {final_text}")
            print(f"LLM end. Text: {final_text}")
        except Exception as e:
            self.logs.append(f"LLM error: {e}")
            print(f"LLM error: {e}")

    def get_logs(self):
        return "\n".join(self.logs)

    def clear_logs(self):
        self.logs = []

# --- Q&A System ---
class GenAIQASystem:
    def __init__(self):
        self.cache = MemoryCache()
        self.callback_handler = LoggingCallbackHandler()
        self.content = None
        self.qa_chain = None
        self.memory = ConversationBufferMemory(memory_key="chat_history", return_messages=True)
        self.initialized = False

    def initialize(self, api_key=None):
        try:
            if api_key:
                os.environ["OPENAI_API_KEY"] = api_key

            if "OPENAI_API_KEY" not in os.environ:
                return False, "OpenAI API key is not set"

            if self.initialized:
                return True, "System already initialized"

            print("Loading Wikipedia page content for Generative artificial intelligence")
            loader = WikipediaLoader(query="Generative artificial intelligence")
            docs = loader.load()
            if not docs:
                return False, "Wikipedia content not loaded. Check query or connection."
            self.content = docs[0].page_content
            print("Page loaded")

            text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=100)
            texts = text_splitter.split_text(self.content)

            embeddings = HuggingFaceEmbeddings(model_name="all-MiniLM-L6-v2")

            if os.path.exists("faiss_index"):
                print("Loading FAISS index from disk...")
                vectorstore = FAISS.load_local("faiss_index", embeddings)
            else:
                print("Creating new FAISS index...")
                vectorstore = FAISS.from_texts(texts, embeddings)
                vectorstore.save_local("faiss_index")

            llm = ChatOpenAI(
                model_name="gpt-3.5-turbo",
                temperature=0,
                callbacks=[self.callback_handler]
            )

            self.qa_chain = ConversationalRetrievalChain.from_llm(
                llm=llm,
                retriever=vectorstore.as_retriever(),
                memory=self.memory,
                callbacks=[self.callback_handler]
            )

            self.initialized = True
            return True, "System initialized successfully"

        except Exception as e:
            print(traceback.format_exc())
            return False, f"Error initializing system: {str(e)}"

    def process_query(self, query):
        if not self.initialized:
            return "System not initialized. Please set your OpenAI API key first."

        cached_answer = self.cache.get(query)
        if cached_answer:
            return f"[Cache] Answer:\n{cached_answer}"

        self.callback_handler.clear_logs()
        print("\n[Retrieval] Processing query...")
        result = self.qa_chain({"question": query})
        answer = result.get("answer", "No answer found")
        self.cache.set(query, answer)

        return answer

    def get_logs(self):
        return self.callback_handler.get_logs()

# --- Create System Instance ---
qa_system = GenAIQASystem()

# --- Gradio Interface ---
def set_api_key(api_key):
    success, message = qa_system.initialize(api_key)
    return message

def respond(message, history):
    response = qa_system.process_query(message)
    history.append((message, response))
    return history

def view_logs():
    return qa_system.get_logs()

# --- Gradio UI ---
with gr.Blocks(title="Generative AI Q/A System") as demo:
    gr.Markdown("# Generative AI Q/A System")
    gr.Markdown("Ask questions about Generative AI using this LangChain-based Q/A system.")

    with gr.Tab("Chat"):
        chat_interface = gr.ChatInterface(fn=respond)

    with gr.Tab("System Logs"):
        logs_output = gr.Textbox(label="System Logs", lines=20)
        view_logs_button = gr.Button("View Logs")
        view_logs_button.click(view_logs, [], logs_output)

    with gr.Tab("Settings"):
        api_key_input = gr.Textbox(type="password", label="OpenAI API Key")
        api_submit = gr.Button("Set API Key")
        api_status = gr.Textbox(label="Status")
        api_submit.click(set_api_key, [api_key_input], [api_status])

    gr.Markdown("## About")
    gr.Markdown("""
    This Q/A system uses LangChain and OpenAI to answer questions based on the Wikipedia page about Generative AI.

    Features:
    - Caching mechanism to avoid repeating work
    - Callback logging to track processing
    - Persistent vector database (FAISS)

    Created by Anjali Haryani (Modified for Hugging Face deployment)
    """)

if __name__ == "__main__":
    demo.launch()