Merge pull request #861 from cgill95/feature/import_unlabeled_dataset

Feature/import unlabeled dataset

app/api/tests/data/example.csv

@@ -1,4 +1,5 @@
-text,label
-AAA,Positive
-BBB,Positive
-CCC,Negative
+text,label,meta
+AAA
+BBB,Positive,The following is meta data
+CCC,Negative
+DDD,,This is meta data

app/api/tests/test_api.py

@@ -1212,11 +1212,11 @@ class TestUploader(APITestCase):
                                 file_format='csv',
                                 expected_status=status.HTTP_201_CREATED)
 
-    def test_cannot_upload_csv_file_does_not_match_column_and_row(self):
+    def test_can_upload_csv_file_does_not_match_column_and_row(self):
         self.upload_test_helper(project_id=self.classification_project.id,
-                                filename='example.invalid.1.csv',
+                                filename='example_column_and_row_not_matching.csv',
                                 file_format='csv',
-                                expected_status=status.HTTP_400_BAD_REQUEST)
+                                expected_status=status.HTTP_201_CREATED)
 
     def test_cannot_upload_csv_file_has_too_many_columns(self):
         self.upload_test_helper(project_id=self.classification_project.id,
@@ -1242,11 +1242,11 @@ class TestUploader(APITestCase):
                                 file_format='excel',
                                 expected_status=status.HTTP_201_CREATED)
 
-    def test_cannot_upload_excel_file_does_not_match_column_and_row(self):
+    def test_can_upload_excel_file_does_not_match_column_and_row(self):
         self.upload_test_helper(project_id=self.classification_project.id,
-                                filename='example.invalid.1.xlsx',
+                                filename='example_column_and_row_not_matching.xlsx',
                                 file_format='excel',
-                                expected_status=status.HTTP_400_BAD_REQUEST)
+                                expected_status=status.HTTP_201_CREATED)
 
     def test_cannot_upload_excel_file_has_too_many_columns(self):
         self.upload_test_helper(project_id=self.classification_project.id,
@@ -1419,10 +1419,10 @@ class TestParser(APITestCase):
                                parser=CoNLLParser())
 
     def test_give_classification_data_to_csv_parser(self):
-        self.parser_helper(filename='example.csv', parser=CSVParser())
+        self.parser_helper(filename='example.csv', parser=CSVParser(), include_label=False)
 
     def test_give_seq2seq_data_to_csv_parser(self):
-        self.parser_helper(filename='example.csv', parser=CSVParser())
+        self.parser_helper(filename='example.csv', parser=CSVParser(), include_label=False)
 
     def test_give_classification_data_to_json_parser(self):
         self.parser_helper(filename='classification.jsonl', parser=JSONParser())

app/api/utils.py

@@ -385,14 +385,17 @@ class ExcelParser(FileParser):
                 yield data
                 data = []
             # Only text column
-            if len(row) == len(columns) and len(row) == 1:
+            if len(row) <= len(columns) and len(row) == 1:
                 data.append({'text': row[0]})
             # Text, labels and metadata columns
-            elif len(row) == len(columns) and len(row) >= 2:
+            elif 2 <= len(row) <= len(columns):
                 datum = dict(zip(columns, row))
                 text, label = datum.pop('text'), datum.pop('label')
                 meta = FileParser.encode_metadata(datum)
-                j = {'text': text, 'labels': [label], 'meta': meta}
+                if label != '':
+                    j = {'text': text, 'labels': [label], 'meta': meta}
+                else:
+                    j = {'text': text, 'meta': meta}
                 data.append(j)
             else:
                 raise FileParseException(line_num=i, line=row)