Bug 142221 – Test legacy event handler attributes (ones with names like "onclick")

• Status: RESOLVED
• Resolution: FIXED
• Component: DOM

[Darin Adler]

Test legacy event handler attributes (ones with names like "onclick")

[Darin Adler]

Created attachment 247765 [details]
Patch

[Darin Adler]

Going to use this test to make sure I don’t break anything while refactoring the handling for these attributes.

[WebKit Commit Bot]

Comment on attachment 247765 [details]
Patch

Clearing flags on attachment: 247765

Committed r180952: <http://trac.webkit.org/changeset/180952>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.

[Brent Fulgham]

This causes a test failure on Windows. Should I rebaseline the Windows tests to match this test's output?

Based on the test output difference, this might just be some kind of mistake in the Windows reporting logic:

--- /home/buildbot/slave/win-release-tests/build/layout-test-results/fast/dom/legacy-event-handler-attributes-expected.txt
+++ /home/buildbot/slave/win-release-tests/build/layout-test-results/fast/dom/legacy-event-handler-attributes-actual.txt
@@ -479,9 +479,9 @@
 FAIL testElementAttribute(audioElement, "webkitanimationend") should be target. Was script: none; content: target.
 FAIL testElementAttribute(audioElement, "webkitanimationiteration") should be target. Was script: none; content: target.
 FAIL testElementAttribute(audioElement, "webkitanimationstart") should be target. Was script: none; content: target.
-FAIL testElementAttribute(audioElement, "webkitkeyadded") should be target. Was script: target; content: none.
-FAIL testElementAttribute(audioElement, "webkitkeyerror") should be target. Was script: target; content: none.
-FAIL testElementAttribute(audioElement, "webkitkeymessage") should be target. Was script: target; content: none.
+FAIL testElementAttribute(audioElement, "webkitkeyadded") should be target. Was none.
+FAIL testElementAttribute(audioElement, "webkitkeyerror") should be target. Was none.
+FAIL testElementAttribute(audioElement, "webkitkeymessage") should be target. Was none.
 FAIL testElementAttribute(audioElement, "webkitpresentationmodechanged") should be target. Was none.
 FAIL testElementAttribute(audioElement, "webkittransitionend") should be target. Was script: none; content: target.
 PASS testElementAttribute(audioElement, "wheel") is "target"
@@ -579,9 +579,9 @@
 FAIL testElementAttribute(videoElement, "webkitanimationend") should be target. Was script: none; content: target.
 FAIL testElementAttribute(videoElement, "webkitanimationiteration") should be target. Was script: none; content: target.
 FAIL testElementAttribute(videoElement, "webkitanimationstart") should be target. Was script: none; content: target.
-FAIL testElementAttribute(videoElement, "webkitkeyadded") should be target. Was script: target; content: none.
-FAIL testElementAttribute(videoElement, "webkitkeyerror") should be target. Was script: target; content: none.
-FAIL testElementAttribute(videoElement, "webkitkeymessage") should be target. Was script: target; content: none.
+FAIL testElementAttribute(videoElement, "webkitkeyadded") should be target. Was none.
+FAIL testElementAttribute(videoElement, "webkitkeyerror") should be target. Was none.
+FAIL testElementAttribute(videoElement, "webkitkeymessage") should be target. Was none.
 FAIL testElementAttribute(videoElement, "webkitpresentationmodechanged") should be target. Was none.
 FAIL testElementAttribute(videoElement, "webkittransitionend") should be target. Was script: none; content: target.
 PASS testElementAttribute(videoElement, "wheel") is "target"

[Brent Fulgham]

I landed a Windows baseline in r180975. Let me know if we should track down the cause of the handful of differences in the output.

[Darin Adler]

My strategy was to leave out events that were different on different platforms. It looks like keyadded, keyerror, and keymessage are feature-specific and the feature is not turned on for Windows.

The preferred fix for now would be to remove those three events from the test and put them into the section below listing events that are platform specific that we might want to test one day.

Then we could remove the Windows-specific baseline.

[Lucas Forschler]

Mass moving XML DOM bugs to the "DOM" Component.