nojhan/weboob-devel
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

Fuck anti-scrapping

Browse source 
A lot of similar forms in the source (ten...), with a hidden
value to validate a real user (form selected by javascript).

But we are lucky, this id can be read is the newsletter form or
in the free call form.
This commit is contained in:
Florent 2013-09-26 12:56:04 +02:00
commit 6d0dd80522
1 changed files with 7 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

7
modules/poivy/pages.py
View file

@ -47,11 +47,18 @@ class LoginPage(BasePage):
captcha = self.document.xpath('//label[@class="label_captcha_input"]') captcha = self.document.xpath('//label[@class="label_captcha_input"]')
if len(captcha) > 0: if len(captcha) > 0:
return False return False
form_newsletter = self.document.xpath('//form[@id="newsletter_form"]')[0]
hidden_input = form_newsletter.xpath('./input[@type="hidden"]')[0]
hidden_id = hidden_input.attrib["value"]
hidden_name = hidden_input.attrib["name"]
# Form without name # Form without name
self.browser.select_form(predicate=self._predicate_form) self.browser.select_form(predicate=self._predicate_form)
self.browser.set_all_readonly(False) self.browser.set_all_readonly(False)
self.browser['login[username]'] = login.encode('iso-8859-1') self.browser['login[username]'] = login.encode('iso-8859-1')
self.browser['login[password]'] = password.encode('iso-8859-1') self.browser['login[password]'] = password.encode('iso-8859-1')
self.browser[hidden_name] = hidden_id
self.browser.submit(nologin=True) self.browser.submit(nologin=True)
return True return True