The intention of the test case is to check the existence of a link node and it should not crash. This can be done by text-only results, which is cross-platform.
Created attachment 80656 [details] fix patch
Comment on attachment 80656 [details] fix patch View in context: https://bugs.webkit.org/attachment.cgi?id=80656&action=review > LayoutTests/ChangeLog:5 > + Convert the test case to text-only so the expected result is cross-platform. Please explicitly say we're converting this to a dumpAsText test. > LayoutTests/editing/deleting/5168598.html:1 > <body> Can we add <!DOCTYPE html>? > LayoutTests/editing/deleting/5168598.html:4 > +<div id="console"></div> We won't need this if we make the following changes. > LayoutTests/editing/deleting/5168598.html:14 > +function log(message) { > + var console = document.getElementById("console"); > + var text = document.createTextNode(message); > + console.appendChild(text); > +} I'm not sure if it makes sense to add this log function just to print one line. > LayoutTests/editing/deleting/5168598.html:21 > +log(document.getElementsByTagName('a').item(0) + "\n"); I would have done: var numberOfElements = document.getElementsByTagName('a').length; document.write(numberOfElements ? 'FAIL: there were ' + numberOfElements + ' anchor elements but expected none' : 'PASS');
Created attachment 80663 [details] fix patch 2
Comment on attachment 80663 [details] fix patch 2 View in context: https://bugs.webkit.org/attachment.cgi?id=80663&action=review r+ but you might want to consider hiding input element after printing PASS/FAIL. > LayoutTests/editing/deleting/5168598-expected.txt:3 > + PASS I think the whitespace before PASS will go away if you hid the input element after the test.
Comment on attachment 80663 [details] fix patch 2 Committed r77154: <http://trac.webkit.org/changeset/77154>
All reviewed patches have been landed. Closing bug.
Removing from Qtwebkit 2.1.1 Nice to have master bug.