115954 2013-05-11 01:08:33 -0700 [att~=val] attribute selectors fail when tab is used as the whitespace character 2014-09-17 14:58:45 -0700 1 1 1 Unclassified WebKit CSS 528+ (Nightly build) Unspecified Unspecified RESOLVED DUPLICATE 136807 BlinkMergeCandidate P2 Normal --- 1 dan webkit-unassigned ap benjamin koivisto msaboff richard.gibson rniwa oldest_to_newest 887764 0 201453 dan 2013-05-11 01:08:33 -0700 Created attachment 201453 Test case showing the problem Note: I have also reported this bug in the Chromium bug tracker... presumably now they've forked WebKit to Blink, we need to report issues both here and there? (https://code.google.com/p/chromium/issues/detail?id=239786) Steps to reproduce the problem: 1. Create an element that has an attribute with a whitespace-delimited list of values. Make the whitespace a space character 2. Create another element that has an attribute with a whitespace-delimited list of values. Make the whitespace a tab character 3. Assign styles to those values using the [att~=val] attribute selector 4. Load the page What is the expected behaviour? Both elements should have identical styling What went wrong? Only the element whose attribute is delimited with space characters will see the styling applied Does this work in other browsers? Yes This fails in the following browsers: WebKit r149907 nightly build from 10 May 2013 (OSX 10.8.3) Safari 6.0.4 (8536.29.13) (OSX 10.8.3) Chrome 28.0.1500.6 dev (OSX 10.8.3) This passes in the following browsers: Opera 12.15 (build 1748) (OSX 10.8.3) Firefox 20.0 (OSX 10.8.3) IE10 (Win 7) IE9 (Win 7) IE8 (Win 7) IE7 (Win XP) I've not tested on Windows in any non-IE browser. Interestingly enough, the Chromium team have marked this as being part of the Blink-JavaScript component, whereas I'd say it was a CSS bug. For this reason (they certainly know more about browser internals than I do), I've left the "component" field here as "New Bugs" rather than setting it to "CSS". 887868 1 ap 2013-05-11 17:03:58 -0700 > presumably now they've forked WebKit to Blink, we need to report issues both here and there? This is correct. Cross-referencing the bugs would be helpful to avoid duplicating work (think you for providing a link to chromium bug tracker). 891860 2 ap 2013-05-21 10:03:52 -0700 *** Bug 116504 has been marked as a duplicate of this bug. *** 891861 3 ap 2013-05-21 10:04:21 -0700 This was fixed in Blink: https://chromium.googlesource.com/chromium/blink/+/3b062d8c310875be0fe6f26a89e2592104944e83 1035773 4 richard.gibson 2014-09-17 14:45:26 -0700 I didn't see this bug when I filed #136807, but they are identical. I suppose this one should be closed as a duplicate, temporal anomalies notwithstanding. 1035777 5 ap 2014-09-17 14:58:45 -0700 As that one is fixed already, forward duping. *** This bug has been marked as a duplicate of bug 136807 *** 201453 2013-05-11 01:08:33 -0700 2013-05-11 01:08:33 -0700 Test case showing the problem whitespaceAttributeSelectorBug.html text/html 583 dan PCFET0NUWVBFIGh0bWw+CjxodG1sIGNsYXNzPSJuby1qcyI+CjxoZWFkPgoJPG1ldGEgY2hhcnNl dD0idXRmLTgiPgoJPHRpdGxlPkF0dHJpYnV0ZSBzZWxlY3RvciBidWc8L3RpdGxlPgoJCgk8c3R5 bGUgdHlwZT0idGV4dC9jc3MiPgoJCXBbZGF0YS1mb3JtYXR0aW5nfj0iYm9sZCJdIHsKCQkJZm9u dC13ZWlnaHQ6IGJvbGQ7CgkJfQoKCQlwW2RhdGEtZm9ybWF0dGluZ349Iml0YWxpYyJdIHsKCQkJ Zm9udC1zdHlsZTogaXRhbGljOwoJCX0KCgkJcFtkYXRhLWZvcm1hdHRpbmd+PSJ1bmRlcmxpbmUi XSB7CgkJCXRleHQtZGVjb3JhdGlvbjogdW5kZXJsaW5lOwoJCX0KCTwvc3R5bGU+CjwvaGVhZD4K Cjxib2R5PgoJPHAgZGF0YS1mb3JtYXR0aW5nPSJib2xkIGl0YWxpYyB1bmRlcmxpbmUiPgoJCVRo aXMgdGV4dCBzaG91bGQgYmUgYm9sZCwgaXRhbGljaXNlZCwgYW5kIHVuZGVybGluZWQKCTwvcD4K CTxwIGRhdGEtZm9ybWF0dGluZz0iYm9sZAlpdGFsaWMJdW5kZXJsaW5lIj4KCQlUaGlzIHRleHQg c2hvdWxkIGFsc28gYmUgYm9sZCwgaXRhbGljaXNlZCwgYW5kIHVuZGVybGluZWQKCTwvcD4KPC9i b2R5Pgo8L2h0bWw+Cg==