see patch
| Comment | File | Size | Author |
|---|---|---|---|
| #7 | inline_comment_early_black_friday_special_for_xjm.patch | 1016 bytes | chx |
| 88KorUsO0VIhB55kRALbjV4hxbf4OzP9tyYw5nzEXa1vWmg0iZX61NeEFQXEBz8.patch | 964 bytes | chx |
see patch
| Comment | File | Size | Author |
|---|---|---|---|
| #7 | inline_comment_early_black_friday_special_for_xjm.patch | 1016 bytes | chx |
| 88KorUsO0VIhB55kRALbjV4hxbf4OzP9tyYw5nzEXa1vWmg0iZX61NeEFQXEBz8.patch | 964 bytes | chx |
Comments
Comment #1
chx commentedThe - private $baseContainer; is just cruft I nuked.
Comment #2
klausiduplicate of #1842726: Transliteration component must not contain drupal_alter(). Please post patches to the issue that caused the broken HEAD.
Comment #3
chx commentedThen I would need to reopen the bootstrap kernel patch which I won't.
Edit: the patch name comes from not having an issue id so I just thrown in some random garbage. Whatever to get the issue done at 5am.
Comment #4
chx commentedTo clarify, tests use $this->container which is set by TestBase to drupal_container() which worked fine until some test like the transliteration test decided to pull something out of CoreBundle -- and as $this->container was not reassigned after Kernel init, that service is not there. The transliteration test is not guilty. I think I am.
Comment #5
chx commentedWebTestBase calls TestBase::rebuildContainer which does
so those tests contain an up to date container. Now DrupalUnitTestBase does too by storing the container as it is being built. You know, PHP objects, they are passed around by handler so as the build finishes it's the exact same object in drupal_container and this->container.
Comment #6
xjmCan we add an inline comment explaining it? :)
Comment #7
chx commentedComment #8
xjmNow that I understand it, the fix makes sense. Thanks chx!
Comment #9
webchickLOL @ patch name. :)
Committed and pushed to 8.x. Thanks!