Website change detection

I have used https://www.changedetection.com/ for a while. But it has always annoyed me that they don't send me the actual diff when they send me a change notification, just a link to a status page. So here is a simple self hosted version, written in Python, using a headless Chrome (to ensure we render JavaScript content). You specify URLs and the divs' id/class in the file - internal state is tracked in some plain text files - and voilà.

This way you can track changes to local restaurant's weekly menu, slow moving government cases, job listings or privacy policies that your providers chooses to change whenever they feel like it.

You can download both selenium and BeautifulSoup from pip. You will also need to download ChromeDriver.

change_detection.py

#!/usr/bin/python3

from bs4 import BeautifulSoup
from selenium import webdriver
from selenium.webdriver.chrome.options import Options
import difflib
import hashlib
import os

state_directory = "/tmp/"

pages = [
    ("https://stortinget.no/no/Stottemeny/Stilling-ledig/", ".jobbnorge-joblist-table"),
    ("https://stortinget.no/no/Saker-og-publikasjoner/Sporsmal/Skriftlige-sporsmal-og-svar/Skriftlig-sporsmal/?qid=74380", "#main-content"),
    ("https://www.smalhans.no/matogvin", '.menu-block'),
    ("https://www.digitalocean.com/legal/privacy-policy/", ".www-Section")
]

chrome_options = Options()
chrome_options.add_argument("--headless")
chrome_driver = "/path/to/chromedriver"

def get_element_from_url(url, selector):
    driver = webdriver.Chrome(
        chrome_options=chrome_options, executable_path=chrome_driver
    )
    driver.get(url)
    html = driver.execute_script("return document.body.innerHTML")
    soup = BeautifulSoup(html)
    selected = soup.select(selector)
    if not selected:
        print("Selector {} not found at {}".format(selector, url))
        return None
    if len(selected) > 1:
        print("Found multiple items with selector {} at {}. Selecting first.".format(selector, url))

    return selected[0].prettify()

for url, selector in pages:
    identifier = hashlib.md5((url+selector).encode('utf-8')).hexdigest()
    file_name = "{}{}.txt".format(state_directory, identifier)

    html_new = get_element_from_url(url, selector)

    # cound not get content
    if not html_new:
        continue

    # save new input and continue, we will check diff next time
    if not os.path.exists(file_name):
        f = open(file_name, "w+")
        f.write(html_new)
        continue

    f = open(file_name, "r")
    html_old = f.read()

    diff = difflib.unified_diff(
        html_old.splitlines(),
        html_new.splitlines()
    )

    diff_output = '\n'.join(list(diff)[3:]) # The four first lines are just +++---

    if diff_output:
        print(' ***** {} ***** '.format(url))
        print("\n" + diff_output + "\n")
        # save new version to file
        f = open(file_name, "w+")
        f.write(html_new)

Put this in your crontab on a server near you and let crontab send you an email when the script outputs a change.

0 * * * * /home/badguy90/bin/change_detection.py

A bonus; colorize the content of the emails in mutt to make the diff easier to read, just like git. Put this in your .muttrc.

# colorfull diffs in email
color body green default "^diff \-.*"
color body green default "^index [a-f0-9].*"
color body green default "^\-\-\- .*"
color body green default "^[\+]{3} .*"
color body cyan default "^[\+][^\+]+.*"
color body red  default "^\-[^\-]+.*"
color body brightblue default "^@@ .*"

Nice.