Name objects after inline images are considered binary data

[yaseerapure]

I'm trying to read a pdf using PyPdf but it gave me this error, although my pdf file is not corrupted. but when i replace the version from 5.3.0 to 5.1.0. the error got resolved
PdfReadError: Unexpected end of stream

Environment

Ubuntu 20.0

Code + PDF

This is a minimal, complete example that shows the issue:

from langchain_community.document_loaders import PyPDFLoader, DirectoryLoader
from langchain.text_splitter import RecursiveCharacterTextSplitter

DATA_PATH = 'data/'

def load_pdf_files(data):
    loader=DirectoryLoader(data,glob='*.pdf',loader_cls=PyPDFLoader)
    documnets=loader.load()
    return documnets

documents=load_pdf_files(data=DATA_PATH)
print("length of documents",len(documents))

This is the pdf file I'm using
https://www.academia.edu/32752835/The_GALE_ENCYCLOPEDIA_of_MEDICINE_SECOND_EDITION

[stefan6419846]

Thanks for the report. This error is caused by our refined handling of inline images. We currently detect

BI
/CS/G
/W 57
/H 55
/BPC 8
/I true
/F/Fl
/DP<</Predictor 15
/Columns 57>>
....
EI Q
/R10 gs
/R12 cs

(part of the content stream of page 2) as the inline image being followed by binary data due to the name /R10 having more than three characters.

I did some quick testing and it seems like I found a solution to resolve this. PR will follow when I have found some more time to do proper testing for it as well in the next days.