sort published drafts by last updated

[TildaDares]

Fixes #9985

[gitpod-io]

[codecov]

Codecov Report

Merging #9995 (1055cfe) into main (f41f8e7) will increase coverage by 0.05%.
The diff coverage is 100.00%.

❗ Current head 1055cfe differs from pull request most recent head ce374d1. Consider uploading reports for the commit ce374d1 to get more accurate results

@@            Coverage Diff             @@
##             main    #9995      +/-   ##
==========================================
+ Coverage   82.13%   82.19%   +0.05%     
==========================================
  Files          98       98              
  Lines        5968     5971       +3     
==========================================
+ Hits         4902     4908       +6     
+ Misses       1066     1063       -3     

Impacted Files	Coverage Δ
app/controllers/home_controller.rb	98.66% <100.00%> (ø)
app/controllers/images_controller.rb	70.27% <100.00%> (ø)
app/controllers/notes_controller.rb	85.09% <100.00%> (+0.22%)	⬆️
app/helpers/application_helper.rb	87.62% <100.00%> (+2.21%)	⬆️
app/services/search_service.rb	95.00% <100.00%> (-0.10%)	⬇️
app/controllers/user_tags_controller.rb	83.78% <0.00%> (+1.35%)	⬆️

[TildaDares]

@jywarren None of the tests are failing

[noi5e]

Nice @TildaDares!

Is there a way we can get this tested in some way before merging? Screenshots, or even with a scrubbed database? That seems important for the original issue at #9985.

[ebarry]

Thank you for writing this @TildaDares !!!!

[jywarren]

We discussed using Timecop on the weekly call, i think that could be a nice test -- maybe a functional test? Or integration test?

[jywarren]

This is awesome though, thank you!!!

[TildaDares]

We discussed using Timecop on the weekly call, i think that could be a nice test -- maybe a functional test? Or integration test?

@jywarren Is it how notes on the activity feed are sorted I'm supposed to test? Currently, I'm testing this by changing the timestamp of an old note to a time far in the future so that it will appear as the first item in the activity feed but this line sorts the items in the activity feed by newly created.

plots2/app/controllers/home_controller.rb

Line 155 in cff3581

activity = (notes + wikis + comments + answer_comments).sort_by(&:created_at).reverse

[jywarren]

So as to the test, it looks like we actually have one here!

plots2/test/functional/notes_controller_test.rb

Lines 840 to 866 in cff3581

	test 'draft author can publish the draft' do
	perform_enqueued_jobs do
	UserSession.create(users(:jeff))
	node = nodes(:draft)
	old_created = node['created']
	old_changed = node['changed']
	assert_equal 3, node.status
	ActionMailer::Base.deliveries.clear
	Timecop.freeze(Date.today + 1) do
	get :publish_draft, params: { id: node.id }
	assert_response :redirect
	assert_equal "Thanks for your contribution. Research note published! Now, it's visible publicly.", flash[:notice]
	node = assigns(:node)
	assert_equal 1, node.status
	assert_not_equal old_changed, node['changed'] # these should have been forward dated!
	assert_not_equal old_created, node['created']
	assert_equal 1, node.author.status
	assert_redirected_to '/notes/' + users(:jeff).username + '/' + (Time.now).strftime('%m-%d-%Y') + '/' + node.title.parameterize
	email = ActionMailer::Base.deliveries.last
	assert_equal '[PublicLab] ' + node.title + " (##{node.id}) ", email.subject
	end
	end
	end

But are we not testing enough? We want to:

1. create a draft
2. confirm the draft has current timestamp
3. move time forward more than 1 day
4. publish the draft
5. confirm draft has new (current day + one day) timestamp
6. confirm draft has new URL too

It does look like that line 155 is sorting a second time by created_at, but for nodes, created_at is just a function which formats the timestamp, i believe? see def created_at in node.rb to confirm.

Hope this helps!

[jywarren]

Here could we assert what it should be equal to? Also, we could assert the wrong (original) timestamp before we publish the draft, to show that it was different before vs after.

This looks great! Thanks Tilda!!!

[TildaDares]

@jywarren I've made the changes

[codeclimate]

Code Climate has analyzed commit 6ac3a98 and detected 0 issues on this pull request.

View more on Code Climate.

[jywarren]

This is fantastic. Great job!!!!