Do no ignore none in latest tx data.

Still ignoring none from other sources (in latest_dataset_value). This
may or may not be right but this is just meant as a broad strokes demo
of how sparse just the latest data is. Also, data directly from
upload/done pages etc may be rightly ignoring nones as this is the sign
of an error - someone has deliberately uploaded a none when they should
either upload a value or not upload at all.

backdrop/transformers/tasks/latest_transaction_explorer_values.py

@@ -31,9 +31,8 @@
 
 def _get_latest_data_point(data, data_point_name):
     def _use_data_point(data_point, name, ignore):
-        has_data = (name in data_point and data_point[name] is not None)
         should_not_be_ignored = (ignore != data_point['type'])
-        return has_data and should_not_be_ignored
+        return should_not_be_ignored
 
     name = data_point_name['name']
     ignore = data_point_name['ignore']
@@ -64,7 +63,7 @@ def _get_stripped_down_data_for_data_point_name_only(
         if field in latest_data_points:
             new_data[field] = latest_data_points[field]
         else:
-            new_data[field] = None
+            return None
     for field in ADDITIONAL_FIELDS:
         if field in latest_data_points:
             new_data[field] = latest_data_points[field]

tests/transformers/tasks/test_latest_transaction_explorer_values.py

@@ -33,14 +33,14 @@
         "type": "quarterly"
     },
     {
+        '_timestamp': u'2014-12-12T00:00:00+00:00',
         "_id": encode_id('quarterly-nonsense', 'digital_takeup'),
-        "_timestamp": "2013-12-12T00:00:00+00:00",
-        "digital_takeup": 0,
-        "end_at": "2013-01-01T00:00:00+00:00",
-        "period": "year",
-        "service_id": "service-with-quarterly-data",
-        "dashboard_slug": "quarterly-nonsense",
-        "type": "quarterly"
+        'period': u'year',
+        'end_at': u'2013-01-01T00:00:00+00:00',
+        'dashboard_slug': 'quarterly-nonsense',
+        'service_id': u'service-with-quarterly-data',
+        'digital_takeup': None,
+        'type': u'quarterly'
     },
     {
         "_id": encode_id(
@@ -55,14 +55,14 @@
         "type": "quarterly"
     },
     {
+        '_timestamp': u'2014-12-12T00:00:00+00:00',
         "_id": encode_id('quarterly-nonsense2', 'digital_takeup'),
-        "_timestamp": "2013-12-12T00:00:00+00:00",
-        "digital_takeup": 0,
-        "end_at": "2013-01-01T00:00:00+00:00",
-        "period": "year",
-        "service_id": "service-with-quarterly-data",
-        "dashboard_slug": "quarterly-nonsense2",
-        "type": "quarterly"
+        'period': u'year',
+        'end_at': u'2013-01-01T00:00:00+00:00',
+        'dashboard_slug': 'quarterly-nonsense2',
+        'service_id': u'service-with-quarterly-data',
+        'digital_takeup': None,
+        'type': u'quarterly'
     },
     {
         "_id": encode_id('sorn', 'cost_per_transaction'),
@@ -118,6 +118,16 @@
         "total_cost": 11767069.6,
         "type": "seasonally-adjusted"
     },
+    {
+        '_timestamp': u'2013-04-01T00:00:00+00:00',
+        'period': u'year',
+        'end_at': u'2012-04-01T00:00:00+00:00',
+        'number_of_transactions': None,
+        'dashboard_slug': 'sorn',
+        'service_id': u'sorn-innit',
+        "_id": encode_id('sorn', 'number_of_transactions'),
+        'type': u'seasonally-adjusted'
+    },
     {
         "_id": encode_id('bis-returns', 'cost_per_transaction'),
         "_timestamp": "2013-04-01T00:00:00+00:00",
@@ -148,16 +158,6 @@
         "dashboard_slug": "bis-returns",
         "type": "seasonally-adjusted"
     },
-    {
-        "_id": encode_id('bis-returns', 'number_of_digital_transactions'),
-        "_timestamp": "2012-12-12T00:00:00+00:00",
-        "end_at": "2013-01-01T00:00:00+00:00",
-        "number_of_digital_transactions": 2301214,
-        "period": "year",
-        "service_id": "bis-annual-returns",
-        "dashboard_slug": "bis-returns",
-        "type": "seasonally-adjusted"
-    },
     {
         "_id": encode_id('bis-returns', 'number_of_transactions'),
         "_timestamp": "2013-04-01T00:00:00+00:00",