is-the-download-list-broken.py

#!/usr/bin/env python3
import bs4
import json
import logging
import os
import requests


log = logging.getLogger(__name__)


def xdg_cache_home():
    try:
        return os.environ['XDG_CACHE_HOME']
    except KeyError:
        return os.path.expanduser('~/.cache')


URL = 'https://endlessos.com/download/'
SELECTOR = '#version optgroup option'
CACHE = os.path.join(xdg_cache_home(), os.path.basename(__file__) + '.json')

RELEASES_JSON = 'https://d1anzknqnc1kmb.cloudfront.net/releases-eos-3.json'


class CheckFailedError(Exception):
    pass


def die(message, *args):
    raise CheckFailedError(message % args)


def check_website():
    log.info("Fetching %s", URL)
    r = requests.get(URL)
    r.raise_for_status()
    html = r.text

    log.info("Finding download options")
    soup = bs4.BeautifulSoup(html, "lxml")
    opts = soup.select(SELECTOR)

    if len(opts) == 0:
        die("No download options found")

    urls = [o.attrs['value'] for o in opts]
    n = len(urls)
    log.info("%d download URLs:\n%s", n, "\n".join(urls))

    try:
        log.info("Loading %s", CACHE)
        with open(CACHE, 'r') as f:
            cache = json.load(f)
    except FileNotFoundError:
        pass
    else:
        log.info("Loaded cache, checking options haven't decreased")
        cache_urls = cache['urls']
        m = len(cache_urls)
        log.info("%d cached urls:\n%s", m, "\n".join(cache_urls))
        if m > n:
            die("Number of download options fell from %d to %d", m, n)

    log.info("Persisting URL list to %s", CACHE)
    with open(CACHE, 'w') as f:
        json.dump({'urls': urls}, f)


def check_releases():
    r = requests.get(RELEASES_JSON)
    r.raise_for_status()

    j = r.json()
    if not j['images']:
        raise CheckFailedError('no images in json: %s', j)

    log.info('%s has some images, at least', RELEASES_JSON)

    # TODO: more thorough checks? Don't want to cross-check with website
    # because we know the website lags by up to an hour…


def main():
    logging.basicConfig(
        level='INFO',
        format='%(asctime)s %(levelname)8s [%(name)s] %(message)s')

    checks = [
        check_releases,
        check_website,
    ]
    errors = False
    for c in checks:
        try:
            c()
        except Exception:
            log.error('%s failed', c.__name__, exc_info=True)
            errors = True

    if errors:
        raise SystemExit(1)

if __name__ == '__main__':
    main()