Update functional tests #839
Conversation
This commit adds a browser test that checks the functionality of searching a regulation. Specifically, the tests check to make sure that clicking the search icon in the nav panel displays the search bar, searching a term displays a results page, and clicking a search result takes the user to the appropriate subsection.
Update the functional test for the table of contents to more thoroughly check that the correct content and effective date is displayed when a TOC link is clicked.
Update the functional test for navigation to check that internal references to regulation subsections link to the proper subsection.
Update the functional test for interpretations to more fully check for the proper "show/hide" controls and to check for a link to the correct interpretation section in the interpretation expandable.
The functional test for search runs in Sauce Labs in a window that is too small to accommodate the 1366x768 resolution that was being used in this test. Also, the IE window cut off the result link that the test was trying to click, making it unclickable and making the test error. Both problems are now fixed: the window is set to 1024x600 (as in the table of contents test) and the last result link on the first page of results—which is reliably always in the viewport and clickable—is now being tested.
There was a problem hiding this comment.
Tests work great for me!
The use of nosetests -tc here seems to require an installation of nose-testconfig, which, as far as I can tell, isn't in the requirements files anywhere.
When the tests do run, I get an ImportError (No module named selenium), I think again because there's a missing dependency on selenium.
regulations/uitests/diff_test.py
Outdated
| @@ -4,6 +4,7 @@ | |||
| from selenium import webdriver | |||
| from selenium.webdriver.support.ui import WebDriverWait | |||
| from selenium.webdriver.support.select import Select | |||
| from selenium.webdriver.support.color import Color | |||
There was a problem hiding this comment.
Minor Python thing, we typically try to alphabetize these imports from their source module (e.g. selenium.webdriver.support.color should come before the other selenium.webdriver.support imports).
regulations/uitests/diff_test.py
Outdated
| '2011-11111', | ||
| '2012-12121' | ||
| ] | ||
| for index, item in enumerate(timeline_list_items): |
There was a problem hiding this comment.
It might be better to write this test backwards, so iterate over the expected output and confirm that it matches what you get. Here you iterate over the output you see, but let's say you didn't get back any .history-toc-list elements, this test would still pass incorrectly.
There was a problem hiding this comment.
GOOD CATCH! It turns out that, thanks to a dumb mistake I made with the selector, this exact thing was happening, and the test was passing despite an empty timeline_list_items list. This should be fixed (and now correctly passing) in 6381fa7.
regulations/uitests/diff_test.py
Outdated
| self.test_url + '/1005-2/2011-11111', | ||
| self.test_url + '/1005-2/2012-12121' | ||
| ] | ||
| for index, link in enumerate(timeline_links): |
There was a problem hiding this comment.
Fixed in 6381fa7. I also updated a similar test in search_test.py to iterate over the expected output instead of the actual output.
regulations/uitests/interp_test.py
Outdated
| @@ -26,6 +26,9 @@ def test_interps(self): | |||
| # should have the appropriate header | |||
| self.assertTrue('OFFICIAL INTERPRETATION TO 2(h)' in interp_dropdown.text) | |||
|
|
|||
| # should have the "SHOW" link | |||
| self.assertTrue('SHOW' in interp_dropdown.text) | |||
There was a problem hiding this comment.
Slightly nicer way to do this in Python: self.assertIn('SHOW', interp_dropdown.text)
regulations/uitests/search_test.py
Outdated
| for index, result in enumerate(search_results): | ||
| self.assertEqual(result.text, expected_results[index]) | ||
|
|
||
| # clicking a serrch result loads the appropriate subsection |
Fix a few functional test issues uncovered during code review: - Alphabetize imports - Iterate over lists of expected outputs rather than lists of actual outputs - Use `assertIn` instead of `assertTrue` where applicable - Fix typos in comments
|
Approved on my end. Let's see this 👶 merged! |
Update functional tests to more fully check that the eRegs user interface is working as expected
Additions
Changes
Testing
regulations-site(e.g.python manage.py runserver 8001)regulations-siteand start up the dummy API with./dummy_api/start.sh 0.0.0.0:8282(or whatever port you usually run the real API fromregulations-coreon)http://localhost:8001/(or whatever port you're runningregulations-siteon) in a browser; you should see only a single regulation, 1005, on the home screen, and all the content for that regulation should be placeholder gibberishregulations-siteand rungrunt testScreenshots
You should see this in your terminal when
grunt testfinishes:Notes
Todos
search_test.py, search isn't actually working locally. When you search for any term, the app returns every section of the dummy reg as a result. So right now the search test can't fully test searching a reg, but it does check to make sure that all the UI elements used to do a search work. When we get search working locally, it should be pretty easy to updatesearch_test.pyto account for correct search results. Should we hold off on addingsearch_test.pyuntil then, though?