205733 2020-01-03 10:24:31 -0800 Unitless zero should not always be allowed 2023-05-28 09:37:38 -0700 1 1 1 Unclassified WebKit CSS Safari Technology Preview Unspecified Unspecified RESOLVED CONFIGURATION CHANGED https://bugs.webkit.org/show_bug.cgi?id=207439 P2 Normal --- 1 simon.fraser webkit-unassigned ahmad.saleem792 annevk ap bfulgham dino emilio graouts koivisto rniwa sam simon.fraser zalan oldest_to_newest 1602503 0 simon.fraser 2020-01-03 10:24:31 -0800 inline bool shouldAcceptUnitlessValue(double value, CSSParserMode cssParserMode, UnitlessQuirk unitless) { // FIXME: Presentational HTML attributes shouldn't use the CSS parser for lengths return value == 0 || isUnitLessValueParsingEnabledForMode(cssParserMode) || (cssParserMode == HTMLQuirksMode && unitless == UnitlessQuirk::Allow); } Note the "value == 0". This is wrong; unitless zero should only be accepted in some legacy situations. See https://www.w3.org/TR/css3-values/#angles and related. 1889511 1 ahmad.saleem792 2022-08-07 05:52:30 -0700 This FIXME is present but the code is bit changed. https://github.com/WebKit/WebKit/blob/10224b907dbca4a066401003bd0734bdefbadf1c/Source/WebCore/css/parser/CSSPropertyParserHelpers.cpp#L93 Is this still required? Thanks! 1958529 2 annevk 2023-05-28 09:37:38 -0700 Comment 0 is addressed due to: (value == 0 && unitlessZero == UnitlessZeroQuirk::Allow) I.e., comparison to 0 is no longer done unconditionally. The FIXME is a separate issue and sounds like it relates to callers of this method. Best tackled separately.