From 99053629d16bfdd3e4f08de68e4ab57d5bd5d7d3 Mon Sep 17 00:00:00 2001
From: Florent Viard <florent.viard@budget-insight.com>
Date: Tue, 2 Feb 2021 11:57:01 +0100
Subject: [PATCH] [bp] Fixes coming history that was returning incorrect labels
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Here are exemples of labels that were returned for coming transactions:
data-titrepopinv2="Détail d'une opération" >PREL D
data-titrepopinv2="Détail d'une opération" >FR6720
data-titrepopinv2="Détail d'une opération" >200410

It should have been:
PREL DE MATMUT ROUEN
FR672045353464765875867865556435
200410414141414141414343

The previous code was not working well because there are two "<a>"
elements, one at the root and one inside a "<noscript>" element.
Sadly the one inside the noscript element is kind of broken:
<a class="popinV3"
href="../../comptesCommun/detailoperation/operation-detail_operation.ea?idxOperation=0&amp;typeOperation=OAV"
target="_blank"> data-titrepopinv2="D&#233;tail d\'une op&#233;ration"
&gt;PREL DE MATMUT ROUEN

The a element is incorrectly closed after "blank" and the real end
symbol is escaped.

To be noted: this way to extract the transaction is only related to
coming transactions. For normal history, the line content is extracted
by the previous part of the code, and is based on a span element and no
"noscript".
---
 modules/bp/pages/accounthistory.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/modules/bp/pages/accounthistory.py b/modules/bp/pages/accounthistory.py
index 08486daa8f..e61b6d20c1 100644
--- a/modules/bp/pages/accounthistory.py
+++ b/modules/bp/pages/accounthistory.py
@@ -34,7 +34,7 @@
 from weboob.browser.filters.html import Link, TableCell
 from weboob.browser.filters.standard import (
     CleanDecimal, CleanText, Eval, Async, AsyncLoad, Date, Env, Format,
-    Regexp,
+    Regexp, Base,
 )
 from weboob.tools.compat import urljoin
 
@@ -217,7 +217,7 @@ def parse(self, el):
                 if (label and label.split()[0] != raw_label.split()[0]) or not label:
                     label = raw_label
 
-                self.env['raw_label'] = CleanText(TableCell('label')(self)[0].xpath('./noscript'))(self) or label
+                self.env['raw_label'] = Base(TableCell('label'), CleanText('a'))(self) or label
 
             obj_raw = Transaction.Raw(Env('raw_label'))
 
-- 
GitLab