How to use pdfbox's PDFTextStripper class in Python.

jpype_pdf_text_stripper.py

#!/usr/bin/env python3

"""
How to use pdfbox's PDFTextStripper class in Python.
"""

import pathlib
import pkg_resources
import re
import urllib.request

import appdirs
import jpype
import jpype.imports
import numpy as np

# Replace with path to pdfbox jar file:
a = appdirs.AppDirs('python-pdfbox')
cache_dir = pathlib.Path(a.user_cache_dir)
file_list = list(cache_dir.glob('pdfbox-app-*.jar'))
def f(s):
    v = re.search('pdfbox-app-([\w\.\-]+)\.jar', s.name).group(1)
    return pkg_resources.parse_version(v)
jpype.addClassPath(sorted(file_list, key=f)[-1])

if not jpype.isJVMStarted():
    jpype.startJVM(jpype.getDefaultJVMPath(), '-Djava.awt.headless=true', convertStrings=False)
from java.awt.image import BufferedImage
from java.io import File
from org.apache.pdfbox.pdmodel import PDDocument
from org.apache.pdfbox.text import PDFTextStripper

def extract_text(in_file):
    """
    Extract text of PDF file.

    Parameters
    ----------
    in_file : str
        Path to input PDF file.

    Returns
    -------
    text : str
        Extracted text.
    """

    doc = PDDocument.load(File(in_file))
    pdf_text_stripper = PDFTextStripper()
    text = pdf_text_stripper.getText(doc)
    return str(text)

if __name__ == '__main__':
    import os
    import tempfile
    import urllib

    # Download sample multipage PDF:
    data = urllib.request.urlopen('https://researchtorevenue.files.wordpress.com/2015/04/1r41ai10801601_fong.pdf').read()
    fd, name = tempfile.mkstemp()
    f = open(name, 'wb')
    f.write(data)
    f.close()
    result = extract_text(name)
    os.unlink(name)

@lebedov Thank you again. You nailed it !!! The file was corrupted I downloaded it again and it's working now. Thank you !!!

Nice example, almost looks easier than using the CLI!

For PDFBox 3.0 they've split the loading API from PDDocument, so use:

from org.apache.pdfbox import Loader

...
doc = Loader.loadPDF(File(in_file))
...