Find possible random errors in Media's JavaScript tests

Based on https://www.drupal.org/node/2846936 all instances of ::assertWaitOnAjaxRequest() were replaced by the new methods.

Is anything else needed here?

In MediaViewsWizardTest::testMediaRevisionWizard there is an AJAX call that doesn't add or remove any element from the page. I couldn't find a proper way of waiting for this call to finish without using ::assertWaitOnAjaxRequest(). Left a TODO in the code.

@vishal9619 could you please point out the differences between your patch and the patch in #8?
Looks like it's the same file, renamed.

Once no DOM modification is done, perhaps any condition would be equivalent to what ::assertWaitOnAjaxRequest does?

We have removed 9 instances of ::assertWaitOnAjaxRequest, only this one is left. We are probably already reducing a lot the probabilities of random failures :)

Re-roll needed

+++ b/core/modules/media/tests/src/FunctionalJavascript/MediaSourceFileTest.php
@@ -46,7 +46,8 @@ public function testMediaFileSource() {
-    $assert_session->assertWaitOnAjaxRequest();
+    $result = $assert_session->waitForButton('Remove');
+    $this->assertNotEmpty($result);

@@ -63,7 +64,8 @@ public function testMediaFileSource() {
-    $assert_session->assertWaitOnAjaxRequest();
+    $result = $assert_session->waitForButton('Remove');
+    $this->assertNotEmpty($result);

+++ b/core/modules/media/tests/src/FunctionalJavascript/MediaSourceImageTest.php
@@ -60,7 +60,8 @@ public function testMediaImageSource() {
-    $assert_session->assertWaitOnAjaxRequest();
+    $result = $assert_session->waitForButton('Remove');
+    $this->assertNotEmpty($result);

+++ b/core/modules/media/tests/src/FunctionalJavascript/MediaSourceTestBase.php
@@ -112,11 +112,13 @@ public function doTestCreateMediaType($media_type_id, $source_id, array $provide
-    $assert_session->assertWaitOnAjaxRequest();
+    $result = $assert_session->waitForElementVisible('css', 'fieldset[data-drupal-selector="edit-source-configuration"]');
+    $this->assertNotEmpty($result);
...
-      $assert_session->selectExists("field_map[{$provided_field}]");
+      $result = $assert_session->waitForElementVisible('css', 'select[name="field_map[' . $provided_field . ']"]');
+      $this->assertNotEmpty($result);

+++ b/core/modules/media/tests/src/FunctionalJavascript/MediaTypeCreationTest.php
@@ -13,24 +13,27 @@ class MediaTypeCreationTest extends MediaJavascriptTestBase {
-    $this->assertSession()->assertWaitOnAjaxRequest();
+    $result = $assert_session->waitForElementVisible('css', 'fieldset[data-drupal-selector="edit-source-configuration"]');
+    $this->assertNotEmpty($result);

@@ -38,51 +41,51 @@ public function testMediaTypeCreationFormWithDefaultField() {
-    $this->assertSession()->assertWaitOnAjaxRequest();
+    $result = $assert_session->waitForElementVisible('css', 'fieldset[data-drupal-selector="edit-source-configuration"]');
+    $this->assertNotEmpty($result);
...
-    $this->assertSession()->assertWaitOnAjaxRequest();
+    $result = $assert_session->waitForElementVisible('css', 'fieldset[data-drupal-selector="edit-source-configuration"]');
+    $this->assertNotEmpty($result);

+++ b/core/modules/media/tests/src/FunctionalJavascript/MediaUiJavascriptTest.php
@@ -120,7 +120,8 @@ public function testMediaTypes() {
-    $assert_session->assertWaitOnAjaxRequest();
+    $result = $assert_session->waitForElementVisible('css', 'fieldset[data-drupal-selector="edit-source-configuration"]');
+    $this->assertNotEmpty($result);

+++ b/core/modules/media/tests/src/FunctionalJavascript/MediaViewsWizardTest.php
@@ -27,7 +27,8 @@ public function testMediaWizard() {
-    $assert_session->assertWaitOnAjaxRequest();
+    $result = $assert_session->waitForElementVisible('css', 'select[data-drupal-selector="edit-show-type"]');
+    $this->assertNotEmpty($result);

This is waiting for the concrete result of the AJAX requests instead of just some AJAX request to complete.

Good hardening.

However, everything else is pure clean-up, theoretically out of scope. At committer's discretion to accept it or require it to be moved to a separate patch.

Looks good!

The hardenings look good. The cleanups I am not very concerned about here, they look fine too. Thanks for making the test more robust/explicit!