find a way to work with PDF data without having to store it locally on the machine?

[will-afs]

No description provided.

[will-afs]

from pdfminer.pdfparser import PDFParser
from pdfminer.pdfdocument import PDFDocument
import urllib.request
from io import BytesIO

def extract_pdf_metadata(pdf_file:str) -> dict:
    """Returns the metadata of a PDF

    Parameters:
    pdf_path (str) : the path of the PDF file of which metadata should be extracted

    Returns:
    dict: metadata of the PDF, presented as a JSON structured as follows : 
        {
            metadata:{
                        "Author": "AURORE",
                        "CreationDate": "D:20200325185329+01'00'",
                        "Creator": "Microsoft Office Word 2007",
                        "ModDate": "D:20210311153835+01'00'",
                        "Producer": "Microsoft Office Word 2007",
                        "Title": "DOSSIER COUP DE POUCE 2020"
                    }
            content:"Lorem ipsum dolor sit amet, ..."
        }
    """
    pdf_parser = PDFParser(pdf_file)
    doc = PDFDocument(pdf_parser)
    metadata = doc.info[0]
    for (key, value) in doc.info[0].items():
        # Need to decode each value from bytestrings toward strings
        metadata[key] = value.decode("utf-8", errors='ignore')
    return metadata
        

if __name__ == '__main__':
    response = urllib.request.urlopen('http://arxiv.org/pdf/cs/9308101v1')
    pdf_txt = response.read()
    fileObj = BytesIO()
    fileObj.write(pdf_txt)
    metadata = extract_pdf_metadata(fileObj)
    print(metadata)

[will-afs]

Closed with commit #cebc9a24d06933fd9b969901ba1fd8b38ad51861