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

cragr: Support of a new history layout

Browse source 
Much easier to parse due to having properly separated in table elements,
and having parent div with meaningful IDs.
Found on the Toulouse region.
This commit is contained in:
Laurent Bachelier 2011-04-10 19:37:43 +02:00 • committed by Romain Bignon
commit 73137b681d
2 changed files with 15 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

1
weboob/backends/cragr/browser.py
View file

@ -42,6 +42,7 @@ class Cragr(BaseBrowser):
('https://%s/login/process' % website) + self.SESSION_REGEXP: pages.AccountsList,
'https://%s/accounting/listAccounts' % website: pages.AccountsList,
'https://%s/accounting/listOperations' % website: pages.AccountsList,
'https://%s/accounting/showAccountDetail.+' % website: pages.AccountsList,
}
BaseBrowser.__init__(self, *args, **kwargs)

17
weboob/backends/cragr/pages/accounts_list.py
View file

@ -75,7 +75,7 @@ class AccountsList(CragrBasePage):
the history of a specific account.
"""
# tested on CA Lorraine, Paris, Toulouse
title_spans = self.document.xpath('/html/body/div[@class="dv"]/span')
title_spans = self.document.xpath('/html/body//div[@class="dv"]/span')
for title_span in title_spans:
title_text = title_span.text_content().strip().replace("\n", '')
if (re.match('.*Compte.*n.*[0-9]+.*au.*', title_text)):
@ -206,12 +206,23 @@ class AccountsList(CragrBasePage):
else:
left_div_count += 1
# new layout that is somewhat easier to parse (found at Toulouse)
table_layout = len(self.document.xpath("id('operationsHeader')")) > 0
# So, how are data laid out?
toulouse_way_of_life = (left_div_count == 2 * right_div_count)
alternate_layout = (left_div_count == 2 * right_div_count)
# we'll have: one left-aligned div for the date, one right-aligned
# div for the amount, and one left-aligned div for the label. Each time.
if (not toulouse_way_of_life):
if table_layout:
lines = self.document.xpath('id("operationsContent")//table[@class="tb"]/tr')
for line in lines:
operation = Operation(index)
index += 1
operation.date = self.extract_text(line[0])
operation.label = self.extract_text(line[1])
operation.amount = clean_amount(self.extract_text(line[2]))
yield operation
elif (not alternate_layout):
for body_elmt in interesting_divs:
if (self.is_right_aligned_div(body_elmt)):
# this is the second line of an operation entry, displaying the amount