parse categories

This commit is contained in:
Romain Bignon 2012-02-26 15:05:37 +01:00
commit defcacb9db
6 changed files with 147 additions and 151 deletions

View file

@ -1,6 +1,6 @@
# -*- coding: utf-8 -*-
# Copyright(C) 2010-2011 Romain Bignon
# Copyright(C) 2010-2012 Romain Bignon
#
# This file is part of weboob.
#
@ -40,7 +40,7 @@ class BNPorcBackend(BaseBackend, ICapBank):
DESCRIPTION = 'BNP Paribas French bank website'
CONFIG = BackendConfig(ValueBackendPassword('login', label='Account ID', masked=False),
ValueBackendPassword('password', label='Password', regexp='^(\d{6}|)$'),
ValueBackendPassword('rotating_password',
ValueBackendPassword('rotating_password', default='',
label='Password to set when the allowed uses are exhausted (6 digits)',
regexp='^(\d{6}|)$'))
BROWSER = BNPorc
@ -76,13 +76,11 @@ class BNPorcBackend(BaseBackend, ICapBank):
def iter_history(self, account):
with self.browser:
for history in self.browser.get_history(account.id):
yield history
return self.browser.iter_history(account.id)
def iter_operations(self, account):
with self.browser:
for coming in self.browser.get_coming_operations(account.id):
yield coming
return self.browser.iter_coming_operations(account.id)
def iter_transfer_recipients(self, ignored):
for account in self.browser.get_transfer_accounts().itervalues():

View file

@ -1,6 +1,6 @@
# -*- coding: utf-8 -*-
# Copyright(C) 2009-2011 Romain Bignon
# Copyright(C) 2009-2012 Romain Bignon
#
# This file is part of weboob.
#
@ -129,16 +129,41 @@ class BNPorc(BaseBrowser):
return None
def get_history(self, id):
self.location('/banque/portail/particulier/FicheA?contractId=%d&pageId=releveoperations&_eventId=changeOperationsPerPage&operationsPerPage=200' % int(id))
return self.page.get_operations()
#def get_history(self, id):
# data = {'contactId': int(id),
# 'pageId': 'releveoperations',
# '_eventId': 'pastOperations',
# 'operationsPerPage': 200,
# 'groupId': -2
# }
# self.location('/banque/portail/particulier/FicheA#pageId=releveoperations', urllib.urlencode(data))
# return self.page.get_operations()
def get_coming_operations(self, id):
#def get_coming_operations(self, id):
# if not self.is_on_page(AccountsList):
# self.location('/NSFR?Action=DSP_VGLOBALE')
# execution = self.page.get_execution_id()
# data = {'externalIAId': 'IAStatements',
# 'contactId': int(id),
# 'pastOrPendingOperations': 2,
# 'pageId': 'mouvementsavenir',
# 'execution': execution,
# }
# self.location('/banque/portail/particulier/FicheA#pageId=mouvementsavenir', urllib.urlencode(data))
# return self.page.get_operations()
def iter_history(self, id):
self.location('/banque/portail/particulier/FicheA?contractId=%d&pageId=releveoperations&_eventId=changeOperationsPerPage&operationsPerPage=200' % int(id))
return self.page.iter_operations()
def iter_coming_operations(self, id):
if not self.is_on_page(AccountsList):
self.location('/NSFR?Action=DSP_VGLOBALE')
execution = self.page.get_execution_id()
self.location('/banque/portail/particulier/FicheA?externalIAId=IAStatements&contractId=%d&pastOrPendingOperations=2&pageId=mouvementsavenir&execution=%s' % (int(id), execution))
return self.page.get_operations()
return self.page.iter_operations()
@check_expired_password
def get_transfer_accounts(self):

View file

@ -1,6 +1,6 @@
# -*- coding: utf-8 -*-
# Copyright(C) 2009-2011 Romain Bignon
# Copyright(C) 2009-2012 Romain Bignon
#
# This file is part of weboob.
#
@ -19,8 +19,7 @@
from .accounts_list import AccountsList
from .account_coming import AccountComing
from .account_history import AccountHistory
from .transactions import AccountHistory, AccountComing
from .transfer import TransferPage, TransferConfirmPage, TransferCompletePage
from .login import LoginPage, ConfirmPage, ChangePasswordPage, MessagePage

View file

@ -1,69 +0,0 @@
# -*- coding: utf-8 -*-
# Copyright(C) 2009-2011 Romain Bignon
#
# This file is part of weboob.
#
# weboob is free software: you can redistribute it and/or modify
# it under the terms of the GNU Affero General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# weboob is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU Affero General Public License for more details.
#
# You should have received a copy of the GNU Affero General Public License
# along with weboob. If not, see <http://www.gnu.org/licenses/>.
import re
from datetime import date
from weboob.tools.browser import BasePage
from weboob.capabilities.bank import Operation
__all__ = ['AccountComing']
class AccountComing(BasePage):
LABEL_PATTERNS = [('^FACTURE CARTE DU (?P<dd>\d{2})(?P<mm>\d{2})(?P<yy>\d{2}) (?P<text>.*)',
u'CB %(yy)s-%(mm)s-%(dd)s: %(text)s'),
('^PRELEVEMENT(?P<text>.*)', 'Order: %(text)s'),
('^ECHEANCEPRET(?P<text>.*)', u'Loan payment n°%(text)s'),
]
def on_loaded(self):
self.operations = []
for tr in self.document.xpath('//table[@id="tableauOperations"]//tr'):
if 'typeop' in tr.attrib:
tds = tr.findall('td')
if len(tds) != 3:
continue
d = tr.attrib['dateop']
d = date(int(d[4:8]), int(d[2:4]), int(d[0:2]))
label = tds[1].text or u''
label = label.replace(u'\xa0', u'')
for child in tds[1].getchildren():
if child.text: label += child.text
if child.tail: label += child.tail
label = label.strip()
for pattern, text in self.LABEL_PATTERNS:
m = re.match(pattern, label)
if m:
label = text % m.groupdict()
amount = tds[2].text.replace('.','').replace(',','.').strip(u' \t\u20ac\xa0\n\r')
operation = Operation(len(self.operations))
operation.date = d
operation.label = label
operation.amount = float(amount)
self.operations.append(operation)
def get_operations(self):
return self.operations

View file

@ -1,68 +0,0 @@
# -*- coding: utf-8 -*-
# Copyright(C) 2009-2011 Romain Bignon
#
# This file is part of weboob.
#
# weboob is free software: you can redistribute it and/or modify
# it under the terms of the GNU Affero General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# weboob is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU Affero General Public License for more details.
#
# You should have received a copy of the GNU Affero General Public License
# along with weboob. If not, see <http://www.gnu.org/licenses/>.
import re
from datetime import date
from weboob.tools.browser import BasePage
from weboob.capabilities.bank import Operation
from weboob.capabilities.base import NotAvailable
__all__ = ['AccountHistory']
class AccountHistory(BasePage):
LABEL_PATTERNS = [(u'^CHEQUEN°(?P<no>.*)', u'CHEQUE', u'%(no)s')]
def on_loaded(self):
self.operations = []
for tr in self.document.xpath('//table[@id="tableCompte"]//tr'):
if len(tr.xpath('td[@class="debit"]')) == 0:
continue
id = tr.find('td').find('input').attrib['value']
op = Operation(id)
op.label = tr.findall('td')[2].text.replace(u'\xa0', u'').strip()
op.date = date(*reversed([int(x) for x in tr.findall('td')[1].text.split('/')]))
op.category = NotAvailable
for pattern, _cat, _lab in self.LABEL_PATTERNS:
m = re.match(pattern, op.label)
if m:
op.category = _cat % m.groupdict()
op.label = _lab % m.groupdict()
break
else:
if ' ' in op.label:
op.category, useless, op.label = [part.strip() for part in op.label.partition(' ')]
debit = tr.xpath('.//td[@class="debit"]')[0].text.replace('.','').replace(',','.').strip(u' \t\u20ac\xa0\n\r')
credit = tr.xpath('.//td[@class="credit"]')[0].text.replace('.','').replace(',','.').strip(u' \t\u20ac\xa0\n\r')
if len(debit) > 0:
op.amount = - float(debit)
else:
op.amount = float(credit)
self.operations.append(op)
def get_operations(self):
return self.operations

View file

@ -0,0 +1,111 @@
# -*- coding: utf-8 -*-
# Copyright(C) 2009-2012 Romain Bignon
#
# This file is part of weboob.
#
# weboob is free software: you can redistribute it and/or modify
# it under the terms of the GNU Affero General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# weboob is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU Affero General Public License for more details.
#
# You should have received a copy of the GNU Affero General Public License
# along with weboob. If not, see <http://www.gnu.org/licenses/>.
import re
from datetime import date
from weboob.tools.browser import BasePage
from weboob.capabilities.bank import Transaction
from weboob.capabilities.base import NotAvailable
__all__ = ['AccountHistory', 'AccountComing']
class TransactionsBasePage(BasePage):
LABEL_PATTERNS = [(re.compile(u'^CHEQUEN°(?P<no>.*)'),
Transaction.TYPE_CHECK, u'%(no)s'),
(re.compile('^FACTURE CARTE DU (?P<dd>\d{2})(?P<mm>\d{2})(?P<yy>\d{2}) (?P<text>.*)'),
Transaction.TYPE_CARD, u'20%(yy)s-%(mm)s-%(dd)s: %(text)s'),
(re.compile('^(PRELEVEMENT|TELEREGLEMENT) (?P<text>.*)'),
Transaction.TYPE_ORDER, '%(text)s'),
(re.compile('^ECHEANCEPRET(?P<text>.*)'),
Transaction.TYPE_LOAN_PAYMENT, u'%(text)s'),
(re.compile('^RETRAIT DAB (?P<dd>\d{2})/(?P<mm>\d{2})/(?P<yy>\d{2}) (?P<HH>\d+)H(?P<MM>\d+) (?P<text>.*)'),
Transaction.TYPE_WITHDRAWAL, u'20%(yy)s-%(mm)s-%(dd)s %(HH)s:%(MM)s: %(text)s'),
(re.compile('^VIREMENT (?P<text>.*)'),
Transaction.TYPE_TRANSFER, u'%(text)s'),
(re.compile('^REMBOURST (?P<text>.*)'),
Transaction.TYPE_PAYBACK, '%(text)s'),
(re.compile('^COMMISSIONS (?P<text>.*)'),
Transaction.TYPE_BANK, '%(text)s'),
]
def parse_text(self, op):
op.category = NotAvailable
if ' ' in op.text:
op.category, useless, op.label = [part.strip() for part in op.label.partition(' ')]
else:
op.label = op.text
for pattern, _type, _label in self.LABEL_PATTERNS:
m = pattern.match(op.text)
if m:
op.type = _type
op.label = (_label % m.groupdict()).strip()
return
class AccountHistory(TransactionsBasePage):
def iter_operations(self):
for tr in self.document.xpath('//table[@id="tableCompte"]//tr'):
if len(tr.xpath('td[@class="debit"]')) == 0:
continue
id = tr.find('td').find('input').attrib['value']
op = Transaction(id)
op.text = tr.findall('td')[2].text.replace(u'\xa0', u'').strip()
op.date = date(*reversed([int(x) for x in tr.findall('td')[1].text.split('/')]))
self.parse_text(op)
debit = tr.xpath('.//td[@class="debit"]')[0].text.replace('.','').replace(',','.').strip(u' \t\u20ac\xa0\n\r')
credit = tr.xpath('.//td[@class="credit"]')[0].text.replace('.','').replace(',','.').strip(u' \t\u20ac\xa0\n\r')
if len(debit) > 0:
op.amount = - float(debit)
else:
op.amount = float(credit)
yield op
class AccountComing(TransactionsBasePage):
def iter_operations(self):
i = 0
for tr in self.document.xpath('//table[@id="tableauOperations"]//tr'):
if 'typeop' in tr.attrib:
tds = tr.findall('td')
if len(tds) != 3:
continue
d = tr.attrib['dateop']
d = date(int(d[4:8]), int(d[2:4]), int(d[0:2]))
text = tds[1].text or u''
text = text.replace(u'\xa0', u'')
for child in tds[1].getchildren():
if child.text: text += child.text
if child.tail: text += child.tail
amount = tds[2].text.replace('.','').replace(',','.').strip(u' \t\u20ac\xa0\n\r')
i += 1
operation = Transaction(i)
operation.date = d
operation.text = text.strip()
self.parse_text(operation)
operation.amount = float(amount)
yield operation