Merge pull request ckan#208 from ckan/github-206-empty-lines

Skip empty lines instead of erroring
# Conflicts:
#	ckanext/xloader/loader.py
### RESOLVED.

ckanext/xloader/loader.py

@@ -194,6 +194,7 @@ def load_csv(csv_filepath, resource_id, mimetype='text/csv', dialect=None, encod
     # Get the list of rows to skip. The rows in the tabulator stream are
     # numbered starting with 1.
     skip_rows = list(range(1, header_offset + 1))
+    skip_rows.append({'type': 'preset', 'value': 'blank'})
 
     # Get the delimiter used in the file
     delimiter = stream.dialect.get('delimiter')
@@ -426,17 +427,21 @@ def load_table(table_filepath, resource_id, mimetype='text/csv', dialect=None, e
     try:
         file_format = os.path.splitext(table_filepath)[1].strip('.')
         with UnknownEncodingStream(table_filepath, file_format, decoding_result,
-                                   post_parse=[TypeConverter().convert_types], dialect=dialect,
+                                   dialect=dialect,
                                    force_encoding=bool(encoding),
+                                   skip_rows=[{'type': 'preset', 'value': 'blank'}],
+                                   post_parse=[TypeConverter().convert_types],
                                    logger=(logger if not has_logged_dialect else None)) as stream:
             header_offset, headers = headers_guess(stream.sample)
             has_logged_dialect = True
     except TabulatorException:
         try:
             file_format = mimetype.lower().split('/')[-1]
             with UnknownEncodingStream(table_filepath, file_format, decoding_result,
-                                       post_parse=[TypeConverter().convert_types], dialect=dialect,
+                                       dialect=dialect,
                                        force_encoding=bool(encoding),
+                                       skip_rows=[{'type': 'preset', 'value': 'blank'}],
+                                       post_parse=[TypeConverter().convert_types],
                                        logger=(logger if not has_logged_dialect else None)) as stream:
                 header_offset, headers = headers_guess(stream.sample)
                 has_logged_dialect = True
@@ -459,6 +464,7 @@ def load_table(table_filepath, resource_id, mimetype='text/csv', dialect=None, e
     # Get the list of rows to skip. The rows in the tabulator stream are
     # numbered starting with 1. We also want to skip the header row.
     skip_rows = list(range(1, header_offset + 2))
+    skip_rows.append({'type': 'preset', 'value': 'blank'})
 
     TYPES, TYPE_MAPPING = get_types()
     # (canada fork only): add config option for strict guessing

ckanext/xloader/tests/samples/sample_with_empty_lines.csv

@@ -0,0 +1,10 @@
+date,temperature,place
+2011-01-01,1,Galway
+2011-01-02,-1,Galway
+2011-01-03,0,Galway
+2011-01-01,6,Berkeley
+
+,,Berkeley
+2011-01-03,5,
+
+

ckanext/xloader/tests/test_loader.py

@@ -632,6 +632,18 @@ def test_with_blanks(self, Session):
         )
         assert len(self._get_records(Session, resource_id)) == 3
 
+    def test_with_empty_lines(self, Session):
+        csv_filepath = get_sample_filepath("sample_with_empty_lines.csv")
+        resource = factories.Resource()
+        resource_id = resource['id']
+        loader.load_csv(
+            csv_filepath,
+            resource_id=resource_id,
+            mimetype="text/csv",
+            logger=logger,
+        )
+        assert len(self._get_records(Session, resource_id)) == 6
+
     def test_with_quoted_commas(self, Session):
         csv_filepath = get_sample_filepath("sample_with_quoted_commas.csv")
         resource = factories.Resource()