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)

Hi, I got the following error when trying to use use the code:

`Exception: Java Exception

The above exception was the direct cause of the following exception:

java.lang.ClassNotFoundException Traceback (most recent call last)
File ~/anaconda3/envs/py39/lib/python3.9/site-packages/jpype/imports.py:195, in _JImportLoader.find_spec(self, name, path, target)
193 try:
194 # Use forname because it give better diagnostics
--> 195 cls = _jpype._java_lang_Class.forName(name, True, _jpype.JPypeClassLoader)
197 # This code only is hit if an error was not thrown

java.lang.ClassNotFoundException: java.lang.ClassNotFoundException: org.apache

The above exception was the direct cause of the following exception:

ImportError Traceback (most recent call last)
Input In [39], in <cell line: 14>()
12 from java.awt.image import BufferedImage
13 from java.io import File
---> 14 from org.apache.pdfbox.pdmodel import PDDocument
15 from org.apache.pdfbox.text import PDFTextStripper

File ~/anaconda3/envs/py39/lib/python3.9/site-packages/jpype/imports.py:203, in _JImportLoader.find_spec(self, name, path, target)
201 # Not found is acceptable
202 except Exception as ex:
--> 203 raise ImportError("Failed to import '%s'" % name) from ex
205 # Import the java module
206 return _ModuleSpec(name, self)

ImportError: Failed to import 'org.apache'`

@sfinotti The script can't find the pdfbox jar file - you need to either copy the jar file to the indicated user cache directory (e.g., ~/Library/Caches/python-pdfbox on MacOS or ~/.cache/python-pdfbox on Linux) or update the first few lines to pass the appropriate path to jpype.addClassPath()

@lebedov Thank you for the prompt answer.
The jar file was already there:

finotti@H170:~/.cache/python-pdfbox$ ls -l total 32 -rwxrwxr-x 1 finotti finotti 28694 jun 27 10:26 pdfbox-app-2.0.26.jar finotti@H170:~/.cache/python-pdfbox$

Maybe the jar version ??

@sfinotti The file size looks suspiciously small - check that its SHA256 checksum satisfies what is listed on Apache's website using sha256sum. If it doesn't match, try downloading the jar file again (the script works fine with 2.0.26 on my system).

@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))
...