45017 2010-08-31 21:28:05 -0700 CSS MediaQuery tests fail 2022-06-09 11:30:26 -0700 1 1 1 Unclassified WebKit CSS 528+ (Nightly build) All OS X 10.5 RESOLVED CONFIGURATION CHANGED http://www.w3.org/Style/CSS/Test/MediaQueries/20100726/media-queries-test.html P2 Normal --- 96752 97006 1 simon.fraser webkit-unassigned ahmad.saleem792 alexander.shalamov annevk ap daniel donggwan.kim hyatt kenneth koivisto rune yael oldest_to_newest 272621 0 simon.fraser 2010-08-31 21:28:05 -0700 180 of the 190 tests in the url fail. Firefox passes them all. 273961 1 66503 simon.fraser 2010-09-03 08:57:02 -0700 Created attachment 66503 Reduced testcase It seems that sheet.cssRules does not include the @media at-rules in WebKit. 274046 2 simon.fraser 2010-09-03 11:15:57 -0700 There are several issues here. First, we don't keep around empty CSSRules, so an empty @media rule doesn't show in sheet.cssRules. This patch fixes that: diff --git a/WebCore/css/CSSMediaRule.cpp b/WebCore/css/CSSMediaRule.cpp index d1c220b..9f8bb4a 100644 --- a/WebCore/css/CSSMediaRule.cpp +++ b/WebCore/css/CSSMediaRule.cpp @@ -40,9 +40,11 @@ CSSMediaRule::~CSSMediaRule() if (m_lstMedia) m_lstMedia->setParent(0); - int length = m_lstCSSRules->length(); - for (int i = 0; i < length; i++) - m_lstCSSRules->item(i)->setParent(0); + if (m_lstCSSRules) { + int length = m_lstCSSRules->length(); + for (int i = 0; i < length; i++) + m_lstCSSRules->item(i)->setParent(0); + } } unsigned CSSMediaRule::append(CSSRule* rule) diff --git a/WebCore/css/CSSParser.cpp b/WebCore/css/CSSParser.cpp index e20537a..4b5f620 100644 --- a/WebCore/css/CSSParser.cpp +++ b/WebCore/css/CSSParser.cpp @@ -5344,7 +5344,7 @@ CSSRule* CSSParser::createImportRule(const CSSParserString& url, MediaList* medi CSSRule* CSSParser::createMediaRule(MediaList* media, CSSRuleList* rules) { - if (!media || !rules || !m_styleSheet) + if (!media || !m_styleSheet) return 0; m_allowImportRules = m_allowNamespaceDeclarations = m_allowVariablesRules = false; RefPtr<CSSMediaRule> rule = CSSMediaRule::create(m_styleSheet, media, rules); Secondly, the error handling described here: http://dev.w3.org/csswg/css3-mediaqueries/#error-handling is not implemented. Even with that, I don't understand the test. It seems to test that: @media screen, not(orientation) (which does not parse) should result in @media screen, but I don't see how that is correct behavior, from the spec. 274050 3 ap 2010-09-03 11:21:36 -0700 > First, we don't keep around empty CSSRules, so an empty @media rule doesn't show in sheet.cssRules. I think that's bug 15649. 274055 4 simon.fraser 2010-09-03 11:22:58 -0700 Bug 15649 is about empty sheets, not empty @ rules. 274060 5 daniel 2010-09-03 11:29:06 -0700 > @media screen, not(orientation) > > (which does not parse) > should result in @media screen, but I don't see how that is correct behavior, > from the spec. The prose says the error handling cares about media queries. Here you have a group of two media queries... So not(orientation) is invalid and thrown away and screen is valid and preserved. This was also discussed recently inside the CSS WG. 274350 6 annevk 2010-09-04 01:58:19 -0700 It is turned into "not all" actually. See the bit about "Malformed media query" under error handling. 550786 7 simon.fraser 2012-02-07 03:32:58 -0800 *** Bug 77958 has been marked as a duplicate of this bug. *** 860184 8 rune 2013-03-21 03:42:14 -0700 I'm working on a patch that's for Media Queries error handling that's nearly finished. The Media Queries tests suite is using the media attribute which is currently split at commas, supporting HTML4 error handling, before fed into the CSS parser, so that needs to be fixed too. 1875021 9 ahmad.saleem792 2022-06-09 09:25:36 -0700 I tested attached test case with Safari 15.5 on macOS 12.4 and it shows dialog box with "sheet has 1 rules" and all other browsers (Chrome Canary 104 and Firefox Nightly 103) also show same behavior. I think it is now fixed since 2010 and can be marked as "RESOLVED CONFIGURATION CHANGED". If I am incorrect, please ignore. Thanks! 1875056 10 ap 2022-06-09 11:30:26 -0700 AFAICT passing this test does mean that the issue is fixed. Thank you for testing! 66503 2010-09-03 08:57:02 -0700 2010-09-03 08:57:02 -0700 Reduced testcase set-media-query.html text/html 688 simon.fraser PCFET0NUWVBFIGh0bWw+Cgo8aHRtbD4KPGhlYWQ+CiAgPHNjcmlwdCB0eXBlPSJ0ZXh0L2phdmFz Y3JpcHQiIGNoYXJzZXQ9InV0Zi04Ij4KICAgIAogICAgZnVuY3Rpb24gZG9UZXN0KCkgewogICAg ICB2YXIgc3R5bGVFbGVtZW50ID0gZG9jdW1lbnQuY3JlYXRlRWxlbWVudCgic3R5bGUiKTsKICAg ICAgc3R5bGVFbGVtZW50LnR5cGUgPSAidGV4dC9jc3MiOwogICAgICAvLyBwYXJzZV90ZXN0X3N0 eWxlX2VsZW1lbnQuZGlzYWJsZWQgPSB0cnVlOyAvLyBmb3IgcGVyZm9ybWFuY2UsIGhvcGVmdWxs eQogICAgICB2YXIgc3R5bGVDb250ZW50cyA9IGRvY3VtZW50LmNyZWF0ZVRleHROb2RlKCJAbWVk aWEgc2NyZWVuIHsgfSIpOwogICAgICBzdHlsZUVsZW1lbnQuYXBwZW5kQ2hpbGQoc3R5bGVDb250 ZW50cyk7CiAgICAgIGRvY3VtZW50LmdldEVsZW1lbnRzQnlUYWdOYW1lKCJoZWFkIilbMF0uYXBw ZW5kQ2hpbGQoc3R5bGVFbGVtZW50KTsKCiAgICAgIHZhciBzaGVldCA9IHN0eWxlRWxlbWVudC5z aGVldDsKICAgICAgYWxlcnQoJ3NoZWV0IGhhcyAnICsgc2hlZXQuY3NzUnVsZXMubGVuZ3RoICsg JyBydWxlcycpOwogICAgfQogICAgCiAgICB3aW5kb3cuYWRkRXZlbnRMaXN0ZW5lcignbG9hZCcs IGRvVGVzdCwgZmFsc2UpOwogIDwvc2NyaXB0Pgo8L2hlYWQ+Cjxib2R5PgoKPC9ib2R5Pgo8L2h0 bWw+Cg==