Dumped the buggy auto parser. replaced it with a faster command-based system that doesn't fail as much

This commit is contained in:
Luke Rogers 2012-03-21 23:52:15 +13:00
parent ebfb6bd809
commit d935cccd18

View file

@ -4,46 +4,37 @@ from urllib2 import urlopen, Request, HTTPError
import re
import BeautifulSoup
ignored_urls = ["http://google.com", "http://youtube.com",
"http://pastebin.com", "http://mibpaste.com",
"http://fpaste.com", "http://git.io"]
titler = re.compile(r'(?si)<title>(.+?)</title>');
def parse(match):
url = urlnorm.normalize(match.encode('utf-8'))
if url not in ignored_urls:
url = url.decode('utf-8')
try:
soup = BeautifulSoup.BeautifulSoup(http.get(url))
return soup.title.string
except:
return "fail"
# there should be " after the ' in the regex string but I was unable to escape it properly
@hook.regex(r"(?i)\b((?:https?://|www\d{0,3}[.]|[a-z0-9.\-]+[.][a-z]{2,4}/)(?:[^\s()<>]+|\(([^\s()<>]+|(\([^\s()<>]+\)))*\))+(?:\(([^\s()<>]+|(\([^\s()<>]+\)))*\)|[^\s`!()\[\]{};:'.,<>?«»“”‘’]))")
def urlparser(match, say=None, bot=None):
def parse(url):
""" an improved version of our parsing code - now regex powered """
url = urlnorm.normalize(url.encode('utf-8'))
url = url.decode('utf-8')
# add http if its missing
if url[:7] != "http://" and url[:8] != "https://":
url = "http://" + url
try:
enabled = bot.config["plugins"]["urlparse"]["enabled"]
except KeyError:
enabled = False
if not enabled:
return
url = urlnorm.normalize(match.group().encode('utf-8'))
if url[:7] != "http://":
if url[:8] != "https://":
url = "http://" + url
for x in ignored_urls:
if x in url:
return
title = parse(url)
if title == "fail":
return
# get the title
print url
request = http.open(url)
real_url = request.geturl()
text = request.read()
text = text.decode('utf8')
match = titler.search(text)
title = match.group(1)
except:
return "Could not parse URL! Are you sure its valid?"
title = http.unescape(title)
realurl = http.get_url(url)
if realurl == url:
say(u"(Link) %s" % title)
return
# if the url has been redirected, show us
if real_url == url:
return title
else:
say(u"(Link) %s [%s]" % (title, realurl))
return
return u"%s [%s]" % (title, real_url)
@hook.command
def title(inp):
".title <url> -- gets the title of a web page"
return parse(inp)