Files
searxng/searxng_extra/update/update_ahmia_blacklist.py
T

33 lines
949 B
Python
Raw Normal View History

2020-10-18 23:55:57 -07:00
#!/usr/bin/env python
2021-10-03 15:12:09 +02:00
# SPDX-License-Identifier: AGPL-3.0-or-later
"""This script saves `Ahmia's blacklist`_ for onion sites.
2020-10-18 23:55:57 -07:00
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
2020-10-18 23:55:57 -07:00
import requests
from searx.data import data_dir
2020-10-18 23:55:57 -07:00
DATA_FILE = data_dir / 'ahmia_blacklist.txt'
2020-10-18 23:55:57 -07:00
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) # type: ignore
return resp.text.split()
2020-10-18 23:55:57 -07:00
if __name__ == '__main__':
blacklist = fetch_ahmia_blacklist()
blacklist.sort()
with DATA_FILE.open("w", encoding='utf-8') as f:
f.write('\n'.join(blacklist))