Don't include public bodies tagged 'test' in public body statistics

In the initial release of public body statistics to WhatDoTheyKnow
a public body only intended for testing ("mySociety Test Quango")
was included in the statistics.  This commit causes public bodies
tagged with "test" to be excluded from the public body statistics
page.

Fixes #1115.

2 files changed, 56 insertions, 2 deletions

diff --git a/app/models/public_body.rb b/app/models/public_body.rb
index dcff761e2..fbe2956e3 100644
--- a/app/models/public_body.rb
+++ b/app/models/public_body.rb
@@ -655,13 +655,22 @@ class PublicBody < ActiveRecord::Base
         end
     end
 
+    def self.where_clause_for_stats(minimum_requests, total_column)
+        # When producing statistics for public bodies, we want to
+        # exclude any that are tagged with 'test' - we use a
+        # sub-select to find the IDs of those public bodies.
+        test_tagged_query = "SELECT model_id FROM has_tag_string_tags" \
+            " WHERE model = 'PublicBody' AND name = 'test'"
+        "#{total_column} >= #{minimum_requests} AND id NOT IN (#{test_tagged_query})"
+    end
+
     # Return data for the 'n' public bodies with the highest (or
     # lowest) number of requests, but only returning data for those
     # with at least 'minimum_requests' requests.
     def self.get_request_totals(n, highest, minimum_requests)
         ordering = "info_requests_count"
         ordering += " DESC" if highest
-        where_clause = "info_requests_count >= #{minimum_requests}"
+        where_clause = where_clause_for_stats minimum_requests, 'info_requests_count'
         public_bodies = PublicBody.order(ordering).where(where_clause).limit(n)
         public_bodies.reverse! if highest
         y_values = public_bodies.map { |pb| pb.info_requests_count }
@@ -682,7 +691,8 @@ class PublicBody < ActiveRecord::Base
         ordering = "y_value"
         ordering += " DESC" if highest
         y_value_column = "(cast(#{column} as float) / #{total_column})"
-        where_clause = "#{total_column} >= #{minimum_requests} AND #{column} IS NOT NULL"
+        where_clause = where_clause_for_stats minimum_requests, total_column
+        where_clause += " AND #{column} IS NOT NULL"
         public_bodies = PublicBody.select("*, #{y_value_column} AS y_value").order(ordering).where(where_clause).limit(n)
         public_bodies.reverse! if highest
         y_values = public_bodies.map { |pb| pb.y_value.to_f }
diff --git a/spec/models/public_body_spec.rb b/spec/models/public_body_spec.rb
index 3ef4c71e7..23842ccff 100644
--- a/spec/models/public_body_spec.rb
+++ b/spec/models/public_body_spec.rb
@@ -532,6 +532,7 @@ describe PublicBody, " when override all public body request emails set" do
 end
 
 describe PublicBody, "when calculating statistics" do
+
     it "should not include unclassified or hidden requests in percentages" do
         with_hidden_and_successful_requests do
             totals_data = PublicBody.get_request_totals(n=3,
@@ -559,4 +560,47 @@ describe PublicBody, "when calculating statistics" do
             percentages_data['y_values'][geraldine_index].should == 50
         end
     end
+
+    it "should only return totals for those with at least a minimum number of requests" do
+        minimum_requests = 1
+        with_enough_info_requests = PublicBody.where(["info_requests_count >= ?",
+                                                      minimum_requests]).length
+        all_data = PublicBody.get_request_totals 4, true, minimum_requests
+        all_data['public_bodies'].length.should == with_enough_info_requests
+    end
+
+    it "should only return percentages for those with at least a minimum number of requests" do
+        with_hidden_and_successful_requests do
+            # With minimum requests at 3, this should return nil
+            # (corresponding to zero public bodies) since the only
+            # public body with just more than 3 info requests (The
+            # Geraldine Quango) has a hidden and an unclassified
+            # request within this block:
+            minimum_requests = 3
+            with_enough_info_requests = PublicBody.where(["info_requests_visible_classified_count >= ?",
+                                                          minimum_requests]).length
+            all_data = PublicBody.get_request_percentages(column='info_requests_successful_count',
+                                                          n=10,
+                                                          true,
+                                                          minimum_requests)
+            all_data.should be_nil
+        end
+    end
+
+    it "should only return those with at least a minimum number of requests, but not tagged 'test'" do
+        hpb = PublicBody.find_by_name 'Department for Humpadinking'
+
+        original_tag_string = hpb.tag_string
+        hpb.add_tag_if_not_already_present 'test'
+
+        begin
+            minimum_requests = 1
+            with_enough_info_requests = PublicBody.where(["info_requests_count >= ?", minimum_requests])
+            all_data = PublicBody.get_request_totals 4, true, minimum_requests
+            all_data['public_bodies'].length.should == 3
+        ensure
+            hpb.tag_string = original_tag_string
+        end
+    end
+
 end