This shifts the behat test run to be triggered form composer activity
within the framework module directly,
* silverstripe/serve is used to provide a webserver, based on the
php -S command
* se/selenium-server-standalone is used to install selenium rather than
a download command
Because we’re using serve, the behat configuration can be locked down.
Further refinements could be made on this:
* the behat-extension could be responsible for installing and
starting/stopping selenium, making these tests more portable
* xvfb initialisation could be provided with another bin tool in the
begat-extension: vendor/bin/xvfb 1024x768
* The bootstrap-file argument to serve could be provided as part of a
composer.json setting. This would make it easier for developers to
start a dev server simply by running vendor/bin/serve
* the behat-extension could be responsible for installing and
starting/stopping silverstripe/serve, removing the need for
specifying base_url at all, and possibly utilising the same bootstrap
file between serve and behat.
The recent upgrade to Chosen.js 1.5 changes the generated markup:
It still renders an <a> tag, but without the "href" attribute.
This attribute isn't actually required to uniquely identify
the link in this structure.
This makes tests more robust at the expense of an additional
check with every step. The "ajax steps" logic works similarly,
but does not respect the loading indicator. While this logic
waits for a short time after ajax requests are finished,
in some cases that isn't enough time for the browser to
process the response and remove the loading indicator.
Since the "ajax steps" logic applies to operations outside of the CMS
as well, we can't add the loading indicator check there.
The main benefit of this is so that authors who make use of
.editorconfig don't end up with whitespace changes in their PRs.
Spaces vs. tabs has been left alone, although that could do with a
tidy-up in SS4 after the switch to PSR-1/2.
The command used was this:
for match in '*.ss' '*.css' '*.scss' '*.html' '*.yml' '*.php' '*.js' '*.csv' '*.inc' '*.php5'; do
find . -path ./thirdparty -not -prune -o -path ./admin/thirdparty -not -prune -o -type f -name "$match" -exec sed -E -i '' 's/[[:space:]]+$//' {} \+
find . -path ./thirdparty -not -prune -o -path ./admin/thirdparty -not -prune -o -type f -name "$match" | xargs perl -pi -e 's/ +$//'
done
- Content filters included in SiteTree view
- View (tree/list) buttons included in SiteTree view
- Update view button styles for new layout
- Updated breadcrumbs for new layout
- The filter panel has been removed in favour of a search menu in the header.
- The multi-select component has been updated:
- Now called 'Bulk actions'
- Styling updated
- Added placeholder text / removed redundant option
- Now also appears in SiteTree view
The CmsUiContext->theIFillInTheDropdownWith() method was written
primarily for TreeDropdownField, which don't have a select tag (only an input tag).
The method currently fails for CMS dropdowns (Dropdown form field class),
since they have a hidden select tag.
I've checked through core feature files and confirmed that every use
of the method relates to TreeDropdownField, which is why this bug hasn't ocurred earlier.
The behat rule for text selection wasn't working due to the JavaScript not executing properly. I have
also updated the code to traverse all childNodes, which is important if you have text like this:
<p>text1 <b>text2</b> text3</p>
And you are trying to select 'text3'
getAttribute('value') behaves inconsistently with Selenium drivers
(fails on Travis and TeamCity, but works locally). Selenium2Driver
in Mink provides a JS wrapper for getting the value, which is more reliable.
This fixes "insert a link" failures, see https://travis-ci.org/silverstripe/silverstripe-cms/jobs/14281251
Technically a textarea DOM node doesn't have a 'value' attribute,
but rather a HTML content. This used to work, but likely broke either
by updated browser handling or updated selenium logic.
Fixes "Scenario: I can edit title and content and see the changes on draft"
