Fix counter_cache decremented twice#32372
Conversation
|
Thanks for the pull request, and welcome! The Rails team is excited to review your changes, and you should hear from @kaspth (or someone else) soon. If any changes to this PR are deemed necessary, please add them as extra commits. This ensures that the reviewer can see what has changed since they last reviewed the code. Due to the way GitHub handles out-of-date commits, this should also make it reasonably obvious what issues have or haven't been addressed. Large or tricky changes may require several passes of review and changes. This repository is being automatically checked for code quality issues using Code Climate. You can see results for this analysis in the PR status below. Newly introduced issues should be fixed before a Pull Request is considered ready to review. Please see the contribution instructions for more information. |
|
Hey, thanks for following up on this! For good measure, can you use http://contributors.rubyonrails.org/faq#credit-several-people so the original author also gets credit? :) @kamipo can you take a look here? |
|
@kaspth Thanks. I updated my commit message to giving credits according to the contribution guidelines you linked me too. Sorry I missed this step. In the mean time, I updated the PR to simplify the unit tests. The original PR was changing the definitions of two models, which was in fact not necessary to reproduce the bug. Hope it's better then way. |
|
@kamipo Any feedback? :) Will this fix still be released on the 5.2 branch? |
|
I have been busy lately. I will visit again in weekend. |
There was a problem hiding this comment.
repace => replace here and above.
There was a problem hiding this comment.
I thought has_key?was deprecated in favor of key?? I prefer the latter anyway.
There was a problem hiding this comment.
I find this test unwieldy, can we simplify to this?
def test_decrement_counter_cache_once
original_trainer = DogLover.create!
dog = Dog.create!(trainer: original_trainer)
assert_difference -> { original_trainer.reload.trained_dogs_count }, -1 do
dog.update!(trainer: DogLover.create!)
end
end|
@kaspth Thank you! I fixed each of your review point in a specific commit. There are conflicts with master though. Should I fix the conflicts before any further review? |
This commit fixes the bug reported in rails#31493 and rails#32079. It includes the following test that would fail without the patch as follows: Failure: BelongsToAssociationsTest#test_decrement_counter_cache_once [rails/activerecord/test/cases/associations/belongs_to_associations_test.rb:778]: Expected: -1 Actual: -2 The fix was proposed in rails#31494 by "gm_zhou <gm_zhou@ctrip.com>". The test was added by the author of this commit. [Guangming Zhou, Malik Olivier Boussejra]
Test case was using changing the a dog's nested attributes using `accepts_nested_attributes_for`. However, this is not required to reproduce the bug fixed by 6107b37. Indeed, the bug can be reproduced simply by having a model with two different `counter_cache`. As we do not change the definitions of dogs or dog_lovers, the overall impact of the PR over the project is reduced by this commit.
key? is preferable over has_key?
|
I rebased and fixed the conflicts. |
This is a 4th attempt to make counter cache transactional completely. Past attempts: rails#9236, rails#14849, rails#23357. All existing counter cache issues (increment/decrement twice, lost increment) are caused due to updating counter cache on the outside of the record saving transaction by assigning belongs_to record, even though assigning that doesn't cause the record saving. We have the `@_after_replace_counter_called` guard condition to mitigate double increment/decrement issues, but we can't completely prevent that inconsistency as long as updating counter cache on the outside of the transaction, since saving the record is not always happened after that. We already have handling counter cache after create/update/destroy, https://github.com/rails/rails/blob/1b90f614b1b3d06b7f02a8b9ea6cd84f15d58643/activerecord/lib/active_record/counter_cache.rb#L162-L189 https://github.com/rails/rails/blob/1b90f614b1b3d06b7f02a8b9ea6cd84f15d58643/activerecord/lib/active_record/associations/builder/belongs_to.rb#L33-L59 so just removing assigning logic on the belongs_to association makes counter cache transactional completely. Closes rails#14849. Closes rails#23357. Closes rails#31493. Closes rails#31494. Closes rails#32372. Closes rails#33113. Closes rails#33117 Closes rails#33129. Closes rails#33458.
This commit fixes the bug reported in #31493 and #32079.
It includes the following test that would fail without the patch as
follows:
The fix was proposed first in #31494 by "gm_zhou gm_zhou@ctrip.com".
I included a failing test case on top of his contribution.