Download Perusall readings as PDF

data.py

title = "The title of the article"
urls="""
<image URLs scraped from the page>
"""

download_perusall.py

# dependencies: imagemagick, img2pdf
from data import title, urls

folder = title.replace(' ','-')
import requests
import os
if not os.path.exists(folder):
    os.mkdir(folder)
i = 0
for u in urls.splitlines():
    if u:
        print('Downloading chunk', i, 'of', title)
        open('{}/{:0>2}.png'.format(folder, i), 'wb').write(requests.get(u.strip()).content)
        i += 1

pgno = 1
for j in range(0, i, 6):
    f = ' '.join(['{}/{:0>2}.png'.format(folder, k) for k in range(j, min(i, j+6))])
    print('Converting page', pgno)
    os.system('convert -append %s %s/page_%s.png' % (f, folder, pgno))
    pgno += 1

print('Converting to pdf')
pages = ' '.join(['{}/page_{}.png'.format(folder, k) for k in range(1, pgno)])
os.system('img2pdf %s -o %s.pdf' % (pages, title))
print('Done')

get_urls.py

/*
 * Click on a reading in the Perusall web interface,
 * and run this script in the developer console.
 * Copy-and-paste the console.info output to data.py.
 */
var len = 0; 
var times = 0;
var i = setInterval(() => { 
  var img = document.querySelectorAll("img.chunk"); img[img.length-1].scrollIntoView(); 
  if (len < img.length) {
    len = img.length;
  } else if (times > 3) {
    var urls = [];
    img.forEach((e) => urls.push(e.src));
    var spl = location.pathname.split('/');
    console.info('urls = """\n'+urls.join('\n')+'\n"""\n\ntitle="'+spl[spl.length-1]+'"\n');
    clearInterval(i);
  } else {
      times++;
  }
}, 2000);

I would suggest adding magick to line 20 in the download_perusall.py as windows often has more convert.exe commands. Apart of that great little script.
os.system('magick convert -append %s %s/page_%s.png' % (f, folder, pgno))

Thank you

I've tried to run the script and it works but not well.
I will explaing what is my issue.
I'm having problems, i think, in the part where I have to scrape the file online on perusall to get the urls.
The problem is that the get_urls script work and return me the urls but not all of them, and when the download _perusall.py script run it create a pdf that show only the initial and last pages.
Probably the problem is that perusall dosen't load the page in time for being captured
Any suggestion of what to do? Thanks.

Yo this was cool and works well, just a warning.
If running in Linux make sure you have the following packages installed:
sudo apt-get install imagemagick
sudo apt install img2pdf
If not you would spend too much time wondering why the os.system() calls are not working properly.
Also, make sure you load the entire document in perusall before running the js script to pull the png URLs.

Other than that, it works well.