Bring time series stats back into line with what was before. Note there was an off-by-one-day bug in the month-binning in prior releases.

- Legacy-Id: 16075

ietf/review/utils.py

@@ -180,10 +180,10 @@ def extract_review_assignment_data(teams=None, reviewers=None, time_from=None, t
         filters &= Q(reviewer__person__in=reviewers)
 
     if time_from:
-        filters &= Q(assigned_on__gte=time_from)
+        filters &= Q(review_request__time__gte=time_from)
 
     if time_to:
-        filters &= Q(assigned_on__lte=time_to)
+        filters &= Q(review_request__time__lte=time_to)
 
     # This doesn't do the left-outer join on docevent that the previous code did. These variables could be renamed
     event_qs = ReviewAssignment.objects.filter(filters)
@@ -193,7 +193,7 @@ def extract_review_assignment_data(teams=None, reviewers=None, time_from=None, t
         "reviewer__person", "assigned_on", "completed_on"
     )
 
-    event_qs = event_qs.order_by(*[o.replace("reviewer", "reviewer__person").replace("team","review_request__team") for o in ordering] + ["assigned_on", "pk", "completed_on"])
+    event_qs = event_qs.order_by(*[o.replace("reviewer", "reviewer__person").replace("team","review_request__team") for o in ordering] + ["review_request__time", "assigned_on", "pk", "completed_on"])
 
     def positive_days(time_from, time_to):
         if time_from is None or time_to is None:

ietf/stats/views.py

@@ -1143,7 +1143,8 @@ def review_stats(request, stats_type=None, acronym=None):
         def time_key_fn(t):
             d = t[req_time_index].date()
             #d -= datetime.timedelta(days=d.weekday()) # weekly
-            d -= datetime.timedelta(days=d.day) # monthly
+            # NOTE: Earlier releases had an off-by-one error here - some stat counts may move a month.
+            d -= datetime.timedelta(days=d.day-1) # monthly 
             return d
 
         found_results = set()