Stop find_in_batches using the records after yielding.

Currently if the code that calls .find_in_batches modifies the yielded array in place then .find_in_batches can enter an infinite loop searching with ruby object ids in the database instead of the primary key of records in the database. This happens because it naively assumes the yielded array hasn't been modified before calling #id on the last object in the array. And ruby (1.8 at least) alias' #id to #object_id so an integer is still returned no matter what the last object is.

By moving finding the #id of the last object before yielding the array it means the calling code can do whatever it wants to the array in terms of modifying it in place, and .find_in_batches doesn't care.

activerecord/lib/active_record/relation/batches.rb

@@ -68,11 +68,14 @@ module ActiveRecord
       records = relation.where(table[primary_key].gteq(start)).all
 
       while records.any?
+        records_size = records.size
+        primary_key_offset = records.last.id
+
         yield records
 
-        break if records.size < batch_size
+        break if records_size < batch_size
 
-        if primary_key_offset = records.last.id
+        if primary_key_offset
           records = relation.where(table[primary_key].gt(primary_key_offset)).to_a
         else
           raise "Primary key not included in the custom select clause"

activerecord/test/cases/batches_test.rb

@@ -93,4 +93,20 @@ class EachTest < ActiveRecord::TestCase
       end
     end
   end
+
+  def test_find_in_batches_should_not_use_records_after_yielding_them_in_case_original_array_is_modified
+    not_a_post = "not a post"
+    not_a_post.stubs(:id).raises(StandardError, "not_a_post had #id called on it")
+
+    assert_nothing_raised do
+      Post.find_in_batches(:batch_size => 1) do |batch|
+        assert_kind_of Array, batch
+        assert_kind_of Post, batch.first
+
+        batch.map! { not_a_post }
+      end
+    end
+
+  end
+
 end