Bug 59305

weee!

This should be smart to ignore text inputs and other elements whose text dump didn't include the shadow DOM before.

Created attachment 118336 [details] Simple layout test for this

It looks like dumpAsText iterates over frames and prints the innerText. Since we don’t want to change innerText, I guess we need something that uses TextIterator etc.?

I was extending TextIterator to handle this, but I found that a shadow content element does not have a renderer. So I cannot emit 'Light', though I can emit 'Shadow A' and 'Shadow B'. How do I add a renderer to shadow content element?

(In reply to comment #5) > How do I add a renderer to shadow content element? Shadow content element should not have a renderer because it is not rendered. Content element has a list of inclusions which you could walk to get the elements output at that point. You should be careful not to change the behavior of eg textContent when implementing this in DRT. hayato is working on visual tree iteration including shadow/content; you should be able to use the API he is designing.

Created attachment 120137 [details] Test Patch

(In reply to comment #7) > Created an attachment (id=120137) [details] > Patch I've uploaded a patch which does not consider ShadowContentElement, and only works on chromium. I think we can use it for discussing the problem.

Attachment 120137 [details] did not pass style-queue: Failed to run "['Tools/Scripts/check-webkit-style', '--diff-files', u'LayoutTests/ChangeLog', u'LayoutTests/fast..." exit_code: 1 Source/WebCore/editing/TextIterator.cpp:400: Boolean expressions that span multiple lines should have their operators on the left side of the line instead of the right side. [whitespace/operators] [4] Total errors found: 1 in 12 files If any of these errors are false positives, please file a bug against check-webkit-style.

Please wait for approval from fishd@chromium.org before submitting because this patch contains changes to the Chromium public API.

Comment on attachment 120137 [details] Test Patch Attachment 120137 [details] did not pass chromium-ews (chromium-xvfb): Output: http://queues.webkit.org/results/10993515 New failing tests: http/tests/inspector-enabled/dedicated-workers-list.html http/tests/media/reload-after-dialog.html fast/dom/shadow/activeelement-should-be-shadowhost.html http/tests/inspector/resource-tree/resource-tree-document-url.html http/tests/security/local-video-poster-from-remote.html http/tests/security/local-video-source-from-remote.html fast/dom/HTMLProgressElement/progress-element-form.html compositing/geometry/clipped-video-controller.html http/tests/media/text-served-as-text.html http/tests/media/video-served-as-text.html http/tests/media/pdf-served-as-pdf.html fast/css/first-letter-block-form-controls-crash.html fast/dom/shadow/layout-tests-can-access-shadow.html fast/dom/HTMLMeterElement/meter-percent-size.html fast/canvas/webgl/tex-image-and-sub-image-2d-with-video.html accessibility/anchor-linked-anonymous-block-crash.html http/tests/security/local-video-src-from-remote.html fast/dom/HTMLProgressElement/indeterminate-progress-002.html http/tests/media/video-error-abort.html

(In reply to comment #11) As you see in comment #11, the patch will break many existing tests. Adding Chromium-specific test results for these tests is not acceptable. We can't change the default behavior until all of platforms implement it.

Comment on attachment 120137 [details] Test Patch View in context: https://bugs.webkit.org/attachment.cgi?id=120137&action=review > Source/WebKit/chromium/public/WebElement.h:62 > + WEBKIT_EXPORT WebString innerTextWithShadow(); LGTM for WebKit API change.

Closing as WONTFIx per comment #12 We are writing shadow related tests even without this.