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

[citibank] Relogin when the website returns garbage instead of a PDF. Closes #1820

Browse source
This commit is contained in:
Oleg Plakhotniuk 2015-04-12 13:42:21 -05:00 • committed by Romain Bignon
commit dbf13930a5
1 changed files with 20 additions and 5 deletions
Download patch file Download diff file Expand all files Collapse all files

25
modules/citibank/browser.py
View file

@ -32,6 +32,7 @@ import os
from datetime import datetime from datetime import datetime
from tempfile import mkstemp from tempfile import mkstemp
from subprocess import check_output, STDOUT from subprocess import check_output, STDOUT
from time import sleep
__all__ = ['Citibank'] __all__ = ['Citibank']
@ -81,7 +82,6 @@ class AccDetailsPage(JsonPage):
account.type = Account.TYPE_CARD account.type = Account.TYPE_CARD
account.label = re.sub(r'<[^>]+>', '', detact['accountName']) account.label = re.sub(r'<[^>]+>', '', detact['accountName'])
account.id = account.label[-4:] account.id = account.label[-4:]
account._innerId = details['instanceID']
for bal in details['accountBalances']: for bal in details['accountBalances']:
label, value = bal['label'], (bal['value'] or ['0'])[0] label, value = bal['label'], (bal['value'] or ['0'])[0]
if label == u'Current Balance:': if label == u'Current Balance:':
@ -128,8 +128,9 @@ class AccDetailsPage(JsonPage):
class StatementsPage(SomePage): class StatementsPage(SomePage):
def dates(self): def dates(self):
return self.doc.xpath( return [x for x in self.doc.xpath(
'//select[@id="currentStatementsDate"]/option/text()')[1:] u'//select[@id="currentStatementsDate"]/option/text()')
if re.match(u'^\d\d\d\d-\d\d-\d\d$', x)]
class StatementPage(RawPage): class StatementPage(RawPage):
@ -139,6 +140,9 @@ class StatementPage(RawPage):
RawPage.__init__(self, *args, **kwArgs) RawPage.__init__(self, *args, **kwArgs)
self._parser = StatementParser(self.doc) self._parser = StatementParser(self.doc)
def is_sane(self):
return self._parser.read_first_date_range() is not None
def transactions(self): def transactions(self):
return sorted(self._parser.read_transactions(), return sorted(self._parser.read_transactions(),
cmp=lambda t1, t2: cmp(t2.date, t1.date)) cmp=lambda t1, t2: cmp(t2.date, t1.date))
@ -188,7 +192,7 @@ class Citibank(LoginBrowser):
yield self.to_account(innerId).account() yield self.to_account(innerId).account()
def iter_history(self, account): def iter_history(self, account):
innerId = account._innerId innerId = self.to_accounts().inner_ids_dict()[account.id]
for trans in self.to_account(innerId).transactions(): for trans in self.to_account(innerId).transactions():
yield trans yield trans
for date in self.to_statements(innerId).dates(): for date in self.to_statements(innerId).dates():
@ -205,7 +209,18 @@ class Citibank(LoginBrowser):
return self.to_page(self.statements, accountID=innerId) return self.to_page(self.statements, accountID=innerId)
def to_statement(self, date): def to_statement(self, date):
return self.to_page(self.statement, date=date) # Sometimes the website returns non-PDF file.
# It recovers if we repeat whole browsing sequence all the way
# from home page up to the statement.
MAX_DELAY=10
for i in xrange(self.MAX_RETRIES):
if self.to_page(self.statement, date=date).is_sane():
return self.page
sleep(min(MAX_DELAY, 1 << i))
self.do_login()
innerId = self.to_accounts().inner_ids_dict().values()[0]
self.to_account(innerId)
self.to_statements(innerId)
@need_login @need_login
def to_page(self, url, **data): def to_page(self, url, **data):