- From: Patrick Garies <pgaries@fastmail.us>
- Date: Sat, 01 Sep 2007 15:57:14 -0500
- To: Christof Höke <csad7@t-online.de>
- CC: www-style@w3.org
Christof Höke wrote:
> hope this is the right play to ask, well for the question:
I would guess that this is not the appropriate place, but I’ll answer
the question anyway.
Christof Höke wrote:
> As far as I understand the spec <!-- and --> in CSS Stylesheets should
> be ignored so it would lead to the following sheet:
>
> body { color: red }
> comment
> body { color: blue }
> body { color: pink }
> comment
> body { color: green }
>
> which definitely has errors in it. But how should the following
> statement be handled now? Ignored as Firefox does it or acknowledged as
> others do it?
I don’t know which part of which CSS specification says that, but
assuming that you’re right, then the style sheet would be interpreted as:
body { color: red }
comment body { color: blue }
body { color: pink }
comment body { color: green }
The first rule would be overridden while the second and fourth rules
would be ignored due to (probably) invalid selectors. Thus, the third
rule would (probably) be applied and the result would be that the text
is pink.
I just tried it in Firefox 2 though and that doesn’t seem to be what’s
happening. If it were, then “<!-- html --> body” would read as “html
body” and still select the body element. Instead, it seems that Firefox
ignores “<!--” and “-->” when they are the first and last ((WRT last)
effectively true even if not really) things in the style sheet.
Presumably, it’s reading something like:
body { color: red }
<!-- comment --> body { color: blue }
body { color: pink }
<!-- comment --> body { color: green }
The above code will still result in the text being pink, since “<!--
comment --> body” is not a valid selector. Thus, I would presume
Firefox’s interpretation to be correct.
You might look at
http://www.webdevout.net/articles/escaping-style-and-script-data for
some insight. It looks like ECMAScript has some specific provisions for
ignoring SGML comments, but I don’t see anything similar for style
elements in the CSS1/2.1/HTML4 specifications. It looks like Firefox is
doing whatever it does for ECMAScript.
Received on Saturday, 1 September 2007 20:57:45 UTC