get_cook_data.py 文件源码

def get_senate_ratings():
    url = get_latest_url('http://www.cookpolitical.com/node/4060')
    doc = parse(url).getroot()

    good_tds = []

    for td in doc.cssselect('td'):
        d = dict(td.items())
        if not d.has_key('width') or not d['width'] == '92':
            continue
        data = [x for x in list(td.itertext()) if x.strip()]
        if len(data) == 1:
            continue

        rating = re.sub(r' \(.*$', '', data[0]) \
                .lower() \
                .replace(' ', '_') \
                .replace('toss_up', 'tossup') \

        data = data[1:]

        for race in data:
            state = re.search(r'[A-Z]{2}', race).group()
            district = ''
            body = 'S'

            cr, created = CookRating.objects.get_or_create(body=body,
                                           state=state,
                                           district=district,
                                           rating=rating)
            cr.save()