A quick-and-dirty mass domain blocker plugin for searxng

blocklist.py

from urllib import request
from urllib.parse import urlparse

from flask_babel import gettext as _
from searx.plugins import Plugin, PluginInfo

block_lists = [
    'https://raw.githubusercontent.com/NotaInutilis/Super-SEO-Spam-Suppressor/refs/heads/main/hosts.txt',
    'https://raw.githubusercontent.com/laylavish/uBlockOrigin-HUGE-AI-Blocklist/refs/heads/main/noai_hosts.txt',
]

def load_block_lists():
    rv = set()
    for link in block_lists:
        with request.urlopen(link) as data:
            for line in data:
                line = line.decode().strip()
                if line and not line.startswith('#'):
                    parts = line.split(' ')
                    if len(parts) == 2:
                        host = parts[1]
                        rv.add(host)
                        if not host.startswith('www.'):
                            rv.add('www.' + host)
    return rv

def mk_url_filter(blocklist):
    
    def url_filter(result, field_name, url_src) -> bool | str:
        hostname = urlparse(url_src).hostname
        if not hostname:
            return True
        return hostname not in blocklist
    
    return url_filter

class MyPlugin(Plugin):

    id = "xr_domain_block"

    def __init__(self, plg_cfg):
        super().__init__(plg_cfg)
        self.info = PluginInfo(id=self.id, name=_("Simple Domain Blocker"), description=_("PoC of a mass domain blocker"))
        self.blocklist = load_block_lists()
        self.url_filter = mk_url_filter(self.blocklist)

    def on_result(self, request, search, result) -> bool:
        result.filter_urls(self.url_filter)
        return True