Re: [Toaster] [PATCH 1/3] Toaster: Toaster: Write UI TestCase -> Visualize all projects

Richard Purdie <[email protected]> Thu, 23 Nov 2023 12:10:37 +0000
Newsgroups org.yoctoproject.lists.toaster,org.openembedded.lists.bitbake-devel
Message-ID <33255f18bffaaf54ff6337c0d3917e34929cc5d0.camel@linuxfoundation.org>
Hi Alassane,

On Tue, 2023-11-21 at 14:47 +0100, Alassane Yattara wrote:
> 
> @@ -205,3 +218,24 @@ class TestAllProjectsPage(SeleniumTestCase):
>          expected_url = reverse('project', args=(self.project.id,))
>          msg = 'link on project name should point to configuration but was %s' % link_url
>          self.assertTrue(link_url.endswith(expected_url), msg)
> +
> +    def test_allProject_table_search_box(self):
> +        """ Test the search box in the all project table on the all projects page """
> +        self._create_projects()
> +
> +        url = reverse('all-projects')
> +        self.get(url)
> +
> +        # Chseck search box is present and works
> +        self.wait_until_present('#projectstable tbody tr')
> +        search_box = self.find('#search-input-projectstable')
> +        self.assertTrue(search_box.is_displayed())
> +
> +        # Check that we can search for a project by project name
> +        search_box.send_keys('test project 10')
> +        search_btn = self.find('#search-submit-projectstable')
> +        search_btn.click()
> +        self.wait_until_present('#projectstable tbody tr')
> +        time.sleep(1)
> +        rows = self.find_all('#projectstable tbody tr')
> +        self.assertTrue(len(rows) == 1)


I've merged these with some tweaks to the commit messages, basically
changing "Toaster:" to "toaster/tests:" and "->" to "-" to make things
more consistent with our other commits. 

The time.sleep() calls in these tests worry me a bit as the values are
a little arbitrary. I noticed one does have to increase the timeout
depending on how much work is being done. Does the test framework not
have a better way to handle this such as a settle function or something
similar? I'm thinking this could be fixed in a follow up commit.

Cheers,

Richard