Using Adobe Readers Export as text function in python

Question

I want to convert lots of PDFs into text files. The formatting is very important and only Adobe Reader seems to get it right (PDFMiner or PyPDF2 do not.)

Is there a way to automate the "export as text" function from Adobe Reader?

Answer 1

The following code will do what you want for one file. I recommend organizing the script into a few little functions and then calling the functions in a loop to process many files. You'll need to install the keyboard library using pip, or some other tool.

import pathlib as pl
import os
import keyboard
import time
import io


KILL_KEY = 'esc'
read_path  = pl.Path("C:/Users/Sam/Downloads/WS-1401-IP.pdf")
####################################################################


write_path = pl.Path(str(read_path.parent/read_path.stem) + ".txt")
overwrite_file = os.path.exists(write_path)

# alt      -- activate keyboard shortcuts
# `F`      -- open file menu
# `v`      -- select "save as text" option
# keyboard.write(write_path)
# `alt+s`  -- save button
# `ctrl+w` -- close file


os.startfile(read_path)
time.sleep(1)
keyboard.press_and_release('alt')
time.sleep(1)
keyboard.press_and_release('f') # -- open file menu
time.sleep(1)
keyboard.press_and_release('v') # -- select "save as text" option
time.sleep(1)
keyboard.write(str(write_path))
time.sleep(1)
keyboard.press_and_release('alt+s')
time.sleep(2)
if overwrite_file:
    keyboard.press_and_release('y')

# wait for program to finish saving
waited_too_long = True
for _ in range(5):
    time.sleep(1)
    if os.path.exists(write_path):
        waited_too_long = False
        break

if waited_too_long:
    with io.StringIO() as ss:
        print(
            "program probably saved to somewhere other than",
            write_path,
            file = ss
        )
        msg = ss.getvalue()
    raise ValueError(msg)

keyboard.press_and_release('ctrl+w') # close the file