Extract html <title> from url and universally return as unicode

utf8-webtitles.py

import urllib2,httplib,re

def url_title(url,**kwargs):
    title = None
    request = urllib2.Request(url)
    try:
        response = urllib2.urlopen(request)
        data = response.read()
    except urllib2.HTTPError:
        data = None
    except urllib2.URLError:
        data = None
    except httplib.BadStatusLine:
        data = None
    except httplib.InvalidURL:
        data = None
    if data:
        if '<title>' in data:
            headers = response.info()
            raw_encoding = headers['content-type'].split('charset=')[-1]
            if 'text/html' in raw_encoding:
                encoding = 'unicode-escape'
            else:
                encoding = raw_encoding
            title_search = re.search('(?<=<title>).*(?=<\/title>)',data)
            if title_search:
                try:
                    if encoding:
                        title = unicode(title_search.group(0),encoding)
                    else:
                        title = title_search.group(0)
                except Exception, e:
                    print(e)
            return "%s | %s" % (encoding,title)

sample_urls = [
    'http://news.cnet.com/8301-31021_3-20035058-260.html?part=rss&subj=news&tag=2547-1_3-0-20', # ISO-8859-1
    'http://www.androidphonethemes.com/nec-medias-n-04c-is-7-7mm-think-%e2%80%93-runs-android-2-2/', # UTF-8
    'http://www.voiceblog.jp/semaasa/', # EUC-JP
    'http://taisyo.seesaa.net/article/187352622.html', # Shift_JIS
    'http://www.tuaw.com/2011/02/22/element-vapor-pro-iphone-4-case-exclusive-first-look-and-giveaw/', # text/html (does not contain unicode chars)
    'http://www.atpress.ne.jp/view/19160', # text/html (contains unicode)
    'http://we-newsing.com/archives/128', # text/html (contains unicode)
    'http://live.nicovideo.jp/watch/lv41334754', # text/html (contains unicode)
    'http://www.superdownloads.com.br/materias/acionistas-da-apple-querem-substituto-steve-jobs.html', # utf8 (claims utf8 but it /LIES/)
    'http://http://g3dfsdf.com', # None. Broken url
    'http://g3dfsdf.com', # None. Dead url
]

for url in sample_urls:
    print url_title(url)