src/loader.py

import os

from langchain_community.document_loaders import PyPDFLoader


def load_all_pdfs(folder_path="data"):
    """Load all PDF files from a folder into LangChain documents."""
    documents = []

    if not os.path.isdir(folder_path):
        return documents

    for file_name in sorted(os.listdir(folder_path)):
        if file_name.lower().endswith(".pdf"):
            loader = PyPDFLoader(os.path.join(folder_path, file_name))
            documents.extend(loader.load())

    return documents