fix: csv import should throw error on repeat user

if the user show up more than once in the csv, the import will fail and send back the error message with line number.

Ticket[AU-20]

update error message for repeat user

version bump

CHANGELOG.rst

@@ -13,8 +13,10 @@ Change Log
 
 Unreleased
 ~~~~~~~~~~
+[0.8.13] - 2021-07-12
+* Fix bug where we ignore repeat user in the csv import
 
-[0.8.12] - 2012-06-21
+[0.8.12] - 2021-06-21
 * Fixed import csv to not working with multiple sections per-user override
 
 [0.8.11] - 2021-07-09

bulk_grades/__init__.py

@@ -2,6 +2,6 @@
 Support for bulk scoring and grading.
 """
 
-__version__ = '0.8.12'
+__version__ = '0.8.13'
 
 default_app_config = 'bulk_grades.apps.BulkGradesConfig'  # pylint: disable=invalid-name

bulk_grades/api.py

@@ -357,12 +357,13 @@ def preprocess_row(self, row):
         Preprocess the CSV row.
         """
         operation = {}
-        if row['user_id'] in self._users_seen:
-            return operation
+        user_id = row['user_id']
+        if user_id in self._users_seen:
+            raise ValidationError(_('Repeated user_id: ') + str(user_id))
 
         operation['new_override_grades'] = []
         operation['course_id'] = self.course_id
-        operation['user_id'] = row['user_id']
+        operation['user_id'] = user_id
 
         for key in row:
             if key.startswith('new_override-'):
@@ -380,7 +381,7 @@ def preprocess_row(self, row):
                             raise ValidationError(_('Grade must not be negative'))
                         operation['new_override_grades'].append((block_id, new_grade))
 
-        self._users_seen.add(row['user_id'])
+        self._users_seen.add(user_id)
         return operation
 
     def process_row(self, row):

bulk_grades/views.py

@@ -91,6 +91,11 @@ def post(self, request, course_id, *args, **kwargs):
             the_file = request.FILES['csv']
             self.processor.process_file(the_file, autocommit=True)
             data = self.processor.status()
+            data['error_messages'] = []
+            for error_message in self.processor.error_messages:
+                line_numbers = ', #'.join(str(line_number+1) for line_number in self.processor.error_messages[error_message])
+                data['error_messages'].append(f'{error_message} on lines (#{line_numbers})')
+
             log.info('Processed file %s for %s -> %s saved, %s processed, %s error. (async=%s)',
                      the_file.name,
                      course_id,

tests/test_api.py

@@ -475,10 +475,12 @@ def test_multiple_subsection_override(self):
 
     def test_repeat_user(self):
         processor = api.GradeCSVProcessor(course_id=self.course_id)
+        username = 'ditto'
         row = {
             'block_id': self.usage_key,
             'new_override-85bb02db': '1',
             'user_id': self.learner.id,
+            'username': username,
             'Previous Points': '',
         }
         operation = processor.preprocess_row(row)
@@ -488,12 +490,13 @@ def test_repeat_user(self):
             'block_id': self.usage_key,
             'new_override-123402db': '2',
             'user_id': self.learner.id,
+            'username': username,
             'Previous Points': ''
         }
 
-        # different row with the same user id should not get processed
-        operation = processor.preprocess_row(row2)
-        assert not operation
+        # different row with the same user id throw error
+        with self.assertRaisesMessage(ValidationError, 'Repeated user'):
+            processor.preprocess_row(row2)
 
     def test_empty_grade(self):
         processor = api.GradeCSVProcessor(course_id=self.course_id)