import string
from io import BytesIO
from typing import AsyncGenerator

from r2r.base.abstractions.document import DataType
from r2r.base.parsers.base_parser import AsyncParser


class PDFParser(AsyncParser[DataType]):
    """A parser for PDF data."""

    def __init__(self):
        try:
            from pypdf import PdfReader

            self.PdfReader = PdfReader
        except ImportError:
            raise ValueError(
                "Error, `pypdf` is required to run `PyPDFParser`. Please install it using `pip install pypdf`."
            )

    async def ingest(self, data: DataType) -> AsyncGenerator[str, None]:
        """Ingest PDF data and yield text from each page."""
        if isinstance(data, str):
            raise ValueError("PDF data must be in bytes format.")

        pdf = self.PdfReader(BytesIO(data))
        for page in pdf.pages:
            page_text = page.extract_text()
            if page_text is not None:
                page_text = "".join(
                    filter(lambda x: x in string.printable, page_text)
                )
                yield page_text