refactor template stats endpoint to read from new redis keys

New redis keys are partitioned per service per day. New process is as
follows:

* require a count of days to filter by. Currently admin always gives 7.
* for each day, check and see if there's anything in redis. There won't
  be if either a) redis is/was down or b) the service didn't send any
  notifications that day
  - if there isn't, go to the database and get a count out.
* combine all these stats together
* get the names/template types etc out of the DB at the end.

app/utils.py

@@ -80,6 +80,9 @@ def cache_key_for_service_template_counter(service_id, limit_days=7):
 
 
 def cache_key_for_service_template_usage_per_day(service_id, datetime):
+    """
+    You should probably pass a BST datetime into this function
+    """
     return "service-{}-template-usage-{}".format(service_id, datetime.date().isoformat())
 
 
@@ -96,4 +99,19 @@ def days_ago(number_of_days):
     """
     Returns midnight a number of days ago. Takes care of daylight savings etc.
     """
-    return get_london_midnight_in_utc(datetime.utcnow() - timedelta(number_of_days))
+    return get_london_midnight_in_utc(datetime.utcnow() - timedelta(days=number_of_days))
+
+
+def last_n_days(limit_days):
+    """
+    Returns the last n dates, oldest first. Takes care of daylight savings (but returns a date, be careful how you manipulate it later! Don't
+    directly use the date for comparing to UTC datetimes!). Includes today.
+    """
+    return [
+        datetime.combine(
+            (convert_utc_to_bst(datetime.utcnow()) - timedelta(days=x)),
+            datetime.min.time()
+        )
+        # reverse the countdown, -1 from first two args to ensure it stays 0-indexed
+        for x in range(limit_days - 1, -1, -1)
+    ]