| 1 |
# Website |
| 2 |
# |
| 3 |
# Douglas Thrift |
| 4 |
# |
| 5 |
# $Id$ |
| 6 |
|
| 7 |
from BeautifulSoup import MinimalSoup |
| 8 |
import copy |
| 9 |
import mechanize |
| 10 |
import re |
| 11 |
|
| 12 |
# XXX: hack to make Wells Fargo http-equiv redirects actually work |
| 13 |
mechanize._http.AbstractHeadParser.head_elems = tuple(list(mechanize._http.AbstractHeadParser.head_elems) + ['body']) |
| 14 |
mechanize._http.HTTPRedirectHandler.max_redirections = 20 |
| 15 |
|
| 16 |
class DuckSoup(MinimalSoup): |
| 17 |
MARKUP_MASSAGE = copy.copy(MinimalSoup.MARKUP_MASSAGE) |
| 18 |
|
| 19 |
MARKUP_MASSAGE.append((re.compile(r'''</?[a-z]+['"]\+['"][a-z]+'''), lambda match: '')) |
| 20 |
|
| 21 |
class Website(object): |
| 22 |
Soup = DuckSoup |
| 23 |
|
| 24 |
def __init__(self, debug): |
| 25 |
self.browser = mechanize.Browser() |
| 26 |
self.browser.addheaders = [('User-agent', 'Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US) AppleWebKit/525.19 (KHTML, like Gecko) Chrome/0.4.154.25 Safari/525.19')] |
| 27 |
|
| 28 |
if debug: |
| 29 |
self.browser.set_debug_http(True) |
| 30 |
|
| 31 |
self.browser.set_handle_equiv(True) |
| 32 |
self.browser.set_handle_redirect(True) |
| 33 |
self.browser.set_handle_refresh(True) |
| 34 |
self.browser.set_handle_robots(False) |
