diff --git a/account_statement_base_import/parser/file_parser.py b/account_statement_base_import/parser/file_parser.py
index 1e8f0e17..5b27c738 100644
--- a/account_statement_base_import/parser/file_parser.py
+++ b/account_statement_base_import/parser/file_parser.py
@@ -19,8 +19,6 @@
 ##############################################################################
 
 from openerp.tools.translate import _
-import base64
-import csv
 import tempfile
 import datetime
 from parser import BankStatementImportParser
@@ -111,20 +109,16 @@ class FileParser(BankStatementImportParser):
         return True
 
 
-    def _parse_csv(self, delimiter=';'):
+    def _parse_csv(self):
         """
-        :return: dict of dict from csv file (line/rows)
+        :return: list of dict from csv file (line/rows)
         """
         csv_file = tempfile.NamedTemporaryFile()
         csv_file.write(self.filebuffer)
-        # We ensure that cursor is at beginig of file
-        csv_file.seek(0)
-        reader = UnicodeDictReader(
-                    open(csv_file.name).readlines(),
-                    delimiter=delimiter,
-                    fieldnames=self.fieldnames
-        )
-        return [x for x in reader]
+        csv_file.flush()
+        with open(csv_file.name, 'rU') as fobj: 
+            reader = UnicodeDictReader(fobj, fieldnames=self.fieldnames)
+            return list(reader)
 
     def _parse_xls(self):
         """
diff --git a/account_statement_base_import/parser/parser.py b/account_statement_base_import/parser/parser.py
index c39098fa..a2dff32d 100644
--- a/account_statement_base_import/parser/parser.py
+++ b/account_statement_base_import/parser/parser.py
@@ -22,7 +22,12 @@ import base64
 import csv
 
 def UnicodeDictReader(utf8_data, **kwargs):
-    csv_reader = csv.DictReader(utf8_data, **kwargs)
+    sniffer = csv.Sniffer()
+    pos = utf8_data.tell()
+    sample_data = utf8_data.read(1024)
+    utf8_data.seek(pos)
+    dialect = sniffer.sniff(sample_data, delimiters=',;\t')
+    csv_reader = csv.DictReader(utf8_data, dialect=dialect, **kwargs)
     for row in csv_reader:
         yield dict([(key, unicode(value, 'utf-8')) for key, value in row.iteritems()])
 
@@ -210,4 +215,4 @@ def new_bank_statement_parser(parser_name, *args, **kwargs):
         if cls.parser_for(parser_name):
             return cls(parser_name, *args, **kwargs)
     raise ValueError
-            
\ No newline at end of file
+