Mikaela/Limnoria
1
0
Fork 0
mirror of https://github.com/Mikaela/Limnoria.git synced 2024-11-19 08:59:27 +01:00
Code Issues Projects Releases Wiki Activity

Google: update scraping regexp

Browse Source
This commit is contained in:
Valentin Lorentz 2019-07-22 21:39:41 +02:00
parent 540b827213
commit dba0ae3771
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
plugins/Google/plugin.py
View File

@ -75,7 +75,7 @@ class Google(callbacks.PluginRegexp):
msg = ircmsgs.privmsg(msg.args[0], s, msg=msg) msg = ircmsgs.privmsg(msg.args[0], s, msg=msg)
return msg return msg
_decode_re = re.compile(r'<div><div class="\w+"><a href="/url\?q=(?P<url>[^"]+)&[^"]+"><div class="(\w| )+">(?P<title>.*?)</div><div class="(\w| )+">(?P<breadcrumbs>.*?)</div></a></div>(?P<content><div class="(\w| )+">.*?</div></div>)', re.DOTALL | re.MULTILINE) _decode_re = re.compile(r'<div class="\w+"><a href="/url\?q=(?P<url>[^"]+)&[^"]+"><div class="(\w| )+">(?P<title>.*?)</div><div class="(\w| )+">(?P<breadcrumbs>.*?)</div></a></div>(?P<content><div class="(\w| )+">.*?</div></div>)', re.DOTALL | re.MULTILINE)
@classmethod @classmethod
def decode(cls, text): def decode(cls, text):
matches = cls._decode_re.finditer(text) matches = cls._decode_re.finditer(text)