From 228b12b4d2d89dd3e5b53a248aaf2a4ae2c7d57a Mon Sep 17 00:00:00 2001
From: Fabian Neugart <f.neugart@mytaxi.com>
Date: Fri, 28 Jan 2022 01:52:12 +0100
Subject: [PATCH] Fixing an issue with non recognized characters

---
 bbva2pandas/extractor.py   | 6 +++---
 requirements.txt           | 6 +++---
 tests/data/pdf-content.txt | 2 ++
 tests/test_extractor.py    | 8 ++++++--
 4 files changed, 14 insertions(+), 8 deletions(-)

diff --git a/bbva2pandas/extractor.py b/bbva2pandas/extractor.py
index d29e819..cdd36ba 100644
--- a/bbva2pandas/extractor.py
+++ b/bbva2pandas/extractor.py
@@ -8,15 +8,15 @@
     \s
     (\d\d/\d\d) #value date
     \s*
-    ([A-ZÑÁÉÍÓÚÜ\'\,\.\:\s]+) #concept
-    \s*
+    ([\wÀ-ÿ .,:*%\'\/()\-\\]+?) #concept
+    \s+
     (-?\d*.?\d*,\d*) #amount of the movement
     \s*
     (\d*.?\d*,\d*) #balance after movement
     \s*
     (\d*) # credit card number
     \s*
-    ([\d\wÑÁÉÍÓÚÜ \.\,\:\*\'\-\/\(\)]*) # subconcept
+    ([\wÀ-ÿ .,:*%\'\/()\-\\]*) # subconcept
     $''',
     re.MULTILINE | re.IGNORECASE | re.VERBOSE
 )
diff --git a/requirements.txt b/requirements.txt
index bf7cbe8..cdb1dba 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,3 +1,3 @@
-pandas==1.2.1
-numpy==1.20.2
-pdftotext==2.1.5
\ No newline at end of file
+pandas==1.4.0
+numpy==1.22.1
+pdftotext==2.2.2
\ No newline at end of file
diff --git a/tests/data/pdf-content.txt b/tests/data/pdf-content.txt
index 78a4c6e..7c7d7ac 100644
--- a/tests/data/pdf-content.txt
+++ b/tests/data/pdf-content.txt
@@ -6,6 +6,8 @@ F.Oper.       F.Valor                                         Concepto
                           SALDO ANTERIOR - - - - - - - - - - - - - - - - - - - - -                                                              0,00
 05/08 05/08               TRANSFERENCIAS                                                                         42,00                        42,00
                            X
+12/08 10/08            COMPRA EN COMERCIO EXTRANJERO-COMISIÓN 3 % INCLUÍDA                                                         -8,79                      13,55
+                            LIDL BCN-CAN BATLL\ BARCELONA
    Todos los importes de este extracto se expresan en:                             SALDO A NUESTRO FAVOR               SALDO A SU FAVOR
    EURO                                                                                                                                42,00
                                                                                                                                             F12345
diff --git a/tests/test_extractor.py b/tests/test_extractor.py
index 6e865ca..900e999 100644
--- a/tests/test_extractor.py
+++ b/tests/test_extractor.py
@@ -14,6 +14,10 @@ def test_movements_extraction(self):
         with open('tests/data/pdf-content.txt') as f:
             input = f.read()
         movements = extractor.find_movements(input)
-        expected = [('05/08', '05/08', 'TRANSFERENCIAS                                                                         ', '42,00', '42,00', '', 'X')]
-        self.assertEqual(1, len(movements))
+        expected = [('05/08', '05/08',
+                     'TRANSFERENCIAS', '42,00',
+                     '42,00', '', 'X'),
+                    ('12/08', '10/08', 'COMPRA EN COMERCIO EXTRANJERO-COMISIÓN 3 % INCLUÍDA', '-8,79', '13,55', '',
+                     'LIDL BCN-CAN BATLL\ BARCELONA')]
+        self.assertEqual(2, len(movements))
         self.assertEqual(expected, movements)