mirror of https://gitgud.io/fatchan/globalafk
You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
30 lines
1.2 KiB
30 lines
1.2 KiB
import logging
|
|
import re
|
|
from abc import ABC, abstractmethod
|
|
|
|
from urlextract import URLExtract
|
|
|
|
|
|
class Evaluator(ABC):
|
|
@abstractmethod
|
|
def eval(self, text, *args, **kwargs):
|
|
raise NotImplementedError
|
|
|
|
|
|
class PostEvaluator(Evaluator):
|
|
def __init__(self, blacklist=(), url_whitelist=()):
|
|
self.blacklist_re = re.compile('|'.join(blacklist), re.IGNORECASE) if blacklist else None
|
|
if url_whitelist:
|
|
self.url_blacklist_re = re.compile(f'(?!{"|".join(url_whitelist)})', re.IGNORECASE)
|
|
self._url_extractor = URLExtract(extract_localhost=False)
|
|
else:
|
|
self.url_blacklist_re = None
|
|
logging.debug(f'Compiled regex\nblacklist:{self.blacklist_re}\nurl blacklist:{self.url_blacklist_re}')
|
|
|
|
def eval(self, text, *args, **kwargs):
|
|
trigger_urls = [
|
|
*filter(self.url_blacklist_re.match, self._url_extractor.find_urls(text, only_unique=True))
|
|
] if self.url_blacklist_re and text else []
|
|
trigger_entries = re.findall(self.blacklist_re, text) if self.blacklist_re and text else []
|
|
logging.debug(f'Evaluated text:{text}\ntrigger urls:{trigger_urls}\ntrigger entries:{trigger_entries}')
|
|
return trigger_urls, trigger_entries
|
|
|