do not follow spam links
This commit is contained in:
parent
847c91b1b3
commit
e1eb1b80af
2 changed files with 10 additions and 9 deletions
|
|
@ -140,8 +140,7 @@ class Cragr(BaseBrowser):
|
||||||
accounts_list.extend(self.page.get_list())
|
accounts_list.extend(self.page.get_list())
|
||||||
|
|
||||||
# credit cards
|
# credit cards
|
||||||
cards_page = self.page.cards_page()
|
for cards_page in self.page.cards_pages():
|
||||||
if cards_page:
|
|
||||||
self.location(cards_page)
|
self.location(cards_page)
|
||||||
assert self.is_on_page(CardsPage)
|
assert self.is_on_page(CardsPage)
|
||||||
accounts_list.extend(self.page.get_list())
|
accounts_list.extend(self.page.get_list())
|
||||||
|
|
|
||||||
|
|
@ -102,13 +102,15 @@ class _AccountsPage(BasePage):
|
||||||
|
|
||||||
yield account
|
yield account
|
||||||
|
|
||||||
def cards_page(self):
|
def cards_pages(self):
|
||||||
try:
|
for line in self.document.xpath('//table[@class="ca-table"]/tr[@class="ligne-connexe"]'):
|
||||||
return self.document.xpath('//table[@class="ca-table"]' +
|
try:
|
||||||
'/tr[@class="ligne-connexe"]' +
|
link = line.xpath('.//a/@href')[0]
|
||||||
'//a/@href')[0]
|
except IndexError:
|
||||||
except IndexError:
|
pass
|
||||||
pass
|
else:
|
||||||
|
if not link.startswith('javascript:'):
|
||||||
|
yield link
|
||||||
|
|
||||||
|
|
||||||
class CardsPage(BasePage):
|
class CardsPage(BasePage):
|
||||||
|
|
|
||||||
Loading…
Add table
Add a link
Reference in a new issue