app.py

# app.py
import os
import warnings
from dotenv import load_dotenv
import gradio as gr
from qdrant_search import QdrantSearch
from langchain_groq import ChatGroq
from nomic_embeddings import EmbeddingsModel

# Load environment variables from .env file
load_dotenv()

# Suppress FutureWarnings
warnings.filterwarnings("ignore", category=FutureWarning)

# Disable tokenizers parallelism to avoid potential issues
os.environ["TOKENIZERS_PARALLELISM"] = "FALSE"

# Initialize global variables
collection_names = ["docs_v1_2", "docs_v2_2", "docs_v3_2"]
limit = 5

# Initialize the language model
llm = ChatGroq(model="mixtral-8x7b-32768")

# Initialize the embeddings model
embeddings = EmbeddingsModel()

# Initialize Qdrant search with necessary credentials
search = QdrantSearch(
    qdrant_url=os.environ["QDRANT_CLOUD_URL"],
    api_key=os.environ["QDRANT_API_KEY"],
    embeddings=embeddings
)

def chat_endpoint(question: str):
    """
    Handles the chat functionality by processing the user's question,
    retrieving relevant documents, generating an answer, and returning sources.

    Args:
        question (str): The user's question.

    Returns:
        Tuple[str, str]: The generated answer and the sources used.
    """
    query = question.strip()
    if not query:
        return "❌ **Error:** Query cannot be empty.", "No sources available."

    # Step 1: Retrieve relevant documents from Qdrant
    retrieved_docs = search.query_multiple_collections(query, collection_names, limit)

    if not retrieved_docs:
        return "⚠️ **No relevant documents found** for your query.", "No sources available."

    # Step 2: Prepare the context from retrieved documents
    context = "\n\n".join([doc['text'] for doc in retrieved_docs])

    # Step 3: Construct the prompt with context and question
    prompt = (
        "You are LangAssist, a knowledgeable assistant for the LangChain Python Library. "
        "Given the following context from the documentation, provide a helpful answer to the user's question.\n\n"
        "### Context:\n{context}\n\n"
        "### Question:\n{question}\n\n"
        "### Answer:"
    ).format(context=context, question=query)

    # Step 4: Generate an answer using the language model
    try:
        answer = llm.invoke(prompt)
    except Exception as e:
        return f"⚠️ **Error generating answer:** {str(e)}", "No sources available."

    # Prepare sources
    sources_md = "\n\n".join([
        f"**Source:** {src['source']}\n**Excerpt:** {src['text']}"
        for src in retrieved_docs
    ])

    return answer.content.strip(), sources_md

# Create Gradio Interface
interface = gr.Interface(
    fn=chat_endpoint,
    inputs=gr.Textbox(
        lines=2,
        placeholder="Type your question here...",
        label="Your Question"
    ),
    outputs=[
        gr.Markdown(label="Answer"),
        gr.Markdown(label="Sources")
    ],
    title="🗨️ LangAssist Chat",
    description="Ask questions about the LangChain Python Library and get answers based on the latest documentation."
)

# If running locally, uncomment the following lines:
# if __name__ == "__main__":
#     interface.launch()