from llama_index.core.schema import Document
from llama_index.core.tools import FunctionTool
from llama_index.retrievers.bm25 import BM25Retriever
from llama_index.core.node_parser import SentenceSplitter

import datasets


# Load the dataset
guest_dataset = datasets.load_dataset("agents-course/unit3-invitees", split="train")

# Convert dataset entries into Document objects
docs = [
    Document(
        text="\n".join([
            f"Name: {guest['name']}",
            f"Relation: {guest['relation']}",
            f"Description: {guest['description']}",
            f"Email: {guest['email']}"
        ]),
        metadata={"name": guest["name"]}
    )
    for guest in guest_dataset
]




# initialize node parser
splitter = SentenceSplitter(chunk_size=512)
nodes = splitter.get_nodes_from_documents(docs)
bm25_retriever = BM25Retriever.from_defaults(nodes=nodes) 

def get_guest_info_retreiver(query: str) -> str:
    """Fetches guest information based on the query."""
    # Retrieve the most relevant document
    results = bm25_retriever.retrieve(query)
    
    if results:
        # Return the text of the most relevant document
        return "\n\n".join([doc.text for doc in results[:3]])
    else:
        return "No relevant guest information found."
    
guest_info_tool = FunctionTool.from_defaults(get_guest_info_retreiver)