searxng/searxng_extra/update/update_ahmia_blacklist.py
Markus Heiser 4c06837a50 [mod] make python code pylint 2.16.1 compliant
Signed-off-by: Markus Heiser <markus.heiser@darmarit.de>
2023-02-10 13:59:21 +01:00

37 lines
1 KiB
Python
Executable file

#!/usr/bin/env python
# lint: pylint
# SPDX-License-Identifier: AGPL-3.0-or-later
"""This script saves `Ahmia's blacklist`_ for onion sites.
Output file: :origin:`searx/data/ahmia_blacklist.txt` (:origin:`CI Update data
... <.github/workflows/data-update.yml>`).
.. _Ahmia's blacklist: https://ahmia.fi/blacklist/
"""
# pylint: disable=use-dict-literal
from os.path import join
import requests
from searx import searx_dir
URL = 'https://ahmia.fi/blacklist/banned/'
def fetch_ahmia_blacklist():
resp = requests.get(URL, timeout=3.0)
if resp.status_code != 200:
# pylint: disable=broad-exception-raised
raise Exception("Error fetching Ahmia blacklist, HTTP code " + resp.status_code)
return resp.text.split()
def get_ahmia_blacklist_filename():
return join(join(searx_dir, "data"), "ahmia_blacklist.txt")
if __name__ == '__main__':
blacklist = fetch_ahmia_blacklist()
with open(get_ahmia_blacklist_filename(), "w", encoding='utf-8') as f:
f.write('\n'.join(blacklist))