from datasets import load_dataset
from langchain_core.documents import Document

# The Hugging Face dataset is similar to a database table,
# and we're converting each row into a Document object.
def load_guest_documents() -> list[Document]:
    dataset = load_dataset(
        "agents-course/unit3-invitees",
        split="train"
    )

    documents = []

    for index, guest in enumerate(dataset):
        document = Document(
            page_content="\n".join([
                f"Name: {guest['name']}",
                f"Relation: {guest['relation']}",
                f"Description: {guest['description']}",
                f"Email: {guest['email']}"
            ]),
            metadata={
                "guest_id": index,
                "name": guest["name"]
            }
        )

        documents.append(document)

    return documents