Import: improve bad encoding detection

1 files changed, 13 insertions, 4 deletions

diff --git a/ishtar_common/forms_common.py b/ishtar_common/forms_common.py
index 7ad89ed3e..f14799f53 100644
--- a/ishtar_common/forms_common.py
+++ b/ishtar_common/forms_common.py
@@ -213,6 +213,8 @@ class BaseImportForm(IshtarForm, forms.ModelForm):
         self.fields["imported_file"].validators = [file_size_validator]
         self._post_init()
 
+    BAD_CHARS = ["é", "³", "ô", "Ã\xa0", "é"]
+
     def _clean_csv(self, is_csv=False):
         imported_file = self.cleaned_data.get("imported_file", None)
         encoding = self.cleaned_data.get("encoding", None)
@@ -220,15 +222,22 @@ class BaseImportForm(IshtarForm, forms.ModelForm):
             try:
                 if not imported_file.name.lower().endswith(".csv"):
                     if is_csv:
-                        raise AssertionError()
+                        raise ValueError()
                     else:
                         return
                 imported_file.seek(0)
                 reader = csv.reader(StringIO(imported_file.read().decode(encoding)))
-                for __ in reader:
-                    break
+                idx = 0
+                for row in reader:
+                    for col in row:
+                        for char in self.BAD_CHARS:
+                            if char in col:
+                                raise ValueError()
+                    idx += 1
+                    if idx >= 200:
+                        break
                 imported_file.seek(0)
-            except (AssertionError, UnicodeDecodeError) as e:
+            except (UnicodeDecodeError, ValueError) as e:
                 raise forms.ValidationError(
                     _("This is not a valid CSV file. Check file format and encoding.")
                 )