pdfloader.py

from enum import Enum
from typing import List
from langchain_community.document_loaders import PyMuPDFLoader
from langchain_core.documents import Document
import asyncio

class PDFLoaderWrapper():
    class LoaderType(str, Enum):
        PYMUPDF = "pymupdf"

    def __init__(self, file_path: str | List[str] , loader_type: LoaderType = LoaderType.PYMUPDF):
        self.file_path = file_path if isinstance(file_path, list) else [file_path]
        self.loader_type = loader_type

    async def aload(self) -> List[Document]:
        all_docs = []
        for file_path in self.file_path:
            if self.loader_type == self.LoaderType.PYMUPDF:
                try:
                    loader = PyMuPDFLoader(file_path)
                    docs = await loader.aload()
                    all_docs.extend(docs)
                except Exception as e:
                    print(f"Error loading file {file_path}: {e}")
                    continue
        return all_docs