Ruby: Writing specs for (partially) memoized code

Posted Over 7 years ago. Visible to the public.

When you're writing specs for ActiveRecord models that use memoization Show archive.org snapshot , a simple #reload will not do:

it 'updates on changes' do
  subject.seat_counts = [5]
  subject.seat_total.should == 5
  # seat_total is either memoized itself, or using some
  # private memoized method
  
  subject.seat_counts = [5, 1]
  subject.seat_total.reload.should == 6 # => Still 5
end

You might be tempted to manually unmemoize any memoized internal method to get #seat_total to update, but that has two disadvantages:

  1. You're doing something that never happens in reality. Your tests shouldn't do things your code never does.
  2. You're fiddling with internal state. When writing tests, stay on your chosen abstraction level.

Solution 1

record.unmemoize_all

Solution 2

Perform a full reload by re-finding your record:

it 'updates on changes' do
  subject.seat_counts = [5]
  subject.seat_total.should == 5
  
  subject.seat_counts = [5, 1]
  # Re-find subject to wipe memoization
  subject = described_class.find(subject.id)
  subject.seat_total.reload.should == 6 # Passes
end

Note that this does not work if you're testing unpersisted new records. In this case, there's probably something conceptually wrong and you either shouldn't expect to see updates or shouldn't memoize at all.

Dominik Schöler
Last edit
Over 7 years ago
Henning Koch
License
Source code in this card is licensed under the MIT License.
Posted by Dominik Schöler to makandra dev (2016-11-15 14:32)