From 0be9eac469c7eb7699042288060d9b65eea48a53 Mon Sep 17 00:00:00 2001 From: Chris Oliver Date: Thu, 9 Feb 2012 22:32:52 -0600 Subject: Added regex to parse urls. Closes #14 --- logbot.py | 12 ++++++++++++ 1 file changed, 12 insertions(+) (limited to 'logbot.py') diff --git a/logbot.py b/logbot.py index ecdcaaa..8691d9d 100644 --- a/logbot.py +++ b/logbot.py @@ -1,4 +1,6 @@ #!/usr/bin/env python +# coding: utf-8 + """ LogBot @@ -45,6 +47,15 @@ except: from ircbot import SingleServerIRCBot from irclib import nm_to_n +import re + +pat1 = re.compile(r"(^|[\n ])(([\w]+?://[\w\#$%&~.\-;:=,?@\[\]+]*)(/[\w\#$%&~/.\-;:=,?@\[\]+]*)?)", re.IGNORECASE | re.DOTALL) + +#urlfinder = re.compile("(?i)\b((?:[a-z][\w-]+:(?:/{1,3}|[a-z0-9%])|www\d{0,3}[.]|[a-z0-9.\-]+[.][a-z]{2,4}/)(?:[^\s()<>]+|\(([^\s()<>]+|(\([^\s()<>]+\)))*\))+(?:\(([^\s()<>]+|(\([^\s()<>]+\)))*\)|[^\s`!()\[\]{};:'\".,<>?«»“”‘’]))") + +def urlify2(value): + return pat1.sub(r'\1\3', value) + #return urlfinder.sub(r'\1', value) ### Configuration options DEBUG = False @@ -190,6 +201,7 @@ class Logbot(SingleServerIRCBot): # Format the event properly chans = event.target() msg = self.format_event(name, event, params) + msg = urlify2(msg) # Quit goes across all channels if not chans or not chans.startswith("#"): -- cgit v1.2.3