DF/93: unify data_format value format (list).

It is better to have field value of the same type whatever happens. Having both `str` and `list` for different cases is to invite errors to the further data processing.
PanDAWMS · Feb 2, 2018 · f0b7620 · f0b7620
1 parent 4ef0e45
commit f0b7620
Show file tree

Hide file tree

Showing 3 changed files with 30 additions and 30 deletions.
diff --git a/Utils/Dataflow/093_datasetsFormat/README b/Utils/Dataflow/093_datasetsFormat/README
@@ -5,7 +5,7 @@
 1. Description
 --------------
 Update dataset metadata:
-* add "data_format" field (field value type: str|tuple)
+* add "data_format" field (field value type: list)
 
 
 2. Input

diff --git a/Utils/Dataflow/093_datasetsFormat/datasets_format.py b/Utils/Dataflow/093_datasetsFormat/datasets_format.py
@@ -63,15 +63,15 @@ def dataset_format(datasetname):
     and first part ('DAOD'), defining the general name of the data format.
 
     :param datasetname:
-    :return: str|tuple
+    :return: list
     """
     if not datasetname:
         return None
     ds_format = datasetname.split('.')[4]
     if re.match("\w+_\w+", ds_format) is not None:
-        result = (ds_format, ds_format.split('_')[0])
+        result = [ds_format, ds_format.split('_')[0]]
     else:
-        result = ds_format
+        result = [ds_format]
     return result