Overriding generated content inserted via the :before pseudo element

Thread view:

Enable EMail Alerts

Start New Thread

Thread rating:

Spartanicus - 27 Sep 2005 22:57 GMT

http://homepage.ntlworld.com/spartanicus/temp3.htm

p:before{content:"foo"}
p:before{content:none}

<p>bar</p>

renders as "foobar". I thought that according to the "latter specified
wins if two rules have the same weight, origin and specificity" cascade
rule it would render as "bar", but apparently I'm missing something. Can
someone point me to the spec section that explains this behaviour?

Signature

Spartanicus

Reply to this Message

Sander Tekelenburg - 28 Sep 2005 03:20 GMT

In article
<4afjj1hgrq4p9a3fq89fqacp3sukk38hcd@news.spartanicus.utvinternet.ie>,

> http://homepage.ntlworld.com/spartanicus/temp3.htm
>
[quoted text clipped - 7 lines]
> rule it would render as "bar", but apparently I'm missing something. Can
> someone point me to the spec section that explains this behaviour?

content:none isn't valid CSS, so it's up to the user-agent to decide
what to do. Refusing to render invalid code isn't considered acceptable
by the public, so it is to be expected that a user-agent will simply
ignore the rule.

If you change to this

p:before{content:"foo"}
p:before{content:""}

<p>bar</p>

the user-agenst should render "bar". iCab, Safari, Mozilla and Opera all
do.

Signature

Sander Tekelenburg, <http://www.euronet.nl/%7Etekelenb/>

Reply to this Message

Spartanicus - 28 Sep 2005 07:34 GMT

>content:none isn't valid CSS

http://www.w3.org/TR/CSS21/generate.html#content

Signature

Spartanicus

Reply to this Message

Sander Tekelenburg - 28 Sep 2005 10:05 GMT

In article
<6fekj1dn2d9jihstdlfj8rngnsupo39l03@news.spartanicus.utvinternet.ie>,

> >content:none isn't valid CSS
>
> http://www.w3.org/TR/CSS21/generate.html#content

Oh my. Turns out I was looking at a local copy of an earlier draft
(2004-02-25) which didn't yet contain this...

Btw, the section on changes between CSS 2.0 and 2.1 doesn't list this
change. Seems like this change is so recent it didn't even make it into
the change log. Not too surprising then that browsers don't support it
yet (besides, CSS 2.1 is still only a draft anyway). Or perhaps the
change log simply contains a typo? Something about that sentence at
<http://www.w3.org/TR/CSS21/changes.html#q4> confuses me.

Signature

Sander Tekelenburg, <http://www.euronet.nl/%7Etekelenb/>

Reply to this Message

Spartanicus - 28 Sep 2005 11:18 GMT

>Btw, the section on changes between CSS 2.0 and 2.1 doesn't list this
>change. Seems like this change is so recent it didn't even make it into
>the change log.

Good catch.

>Not too surprising then that browsers don't support it
>yet (besides, CSS 2.1 is still only a draft anyway).

I tend to think of CSS 2.1 primarily as clarifying issues with 2.0 and
removing stuff that wasn't implemented, I forgot that there were minor
changes, and a few additions other than inline-block.

>Or perhaps the change log simply contains a typo?

I'll mention it on the w3c style mailing list.

>Something about that sentence at
><http://www.w3.org/TR/CSS21/changes.html#q4> confuses me.

Same here :)

Signature

Spartanicus

Reply to this Message

Spartanicus - 28 Sep 2005 08:00 GMT

>content:none isn't valid CSS

Just discovered that it's in CSS 2.1, but not in CSS 2.0. That's the
first time that I've been caught out in using CSS 2.1 and finding that
modern browsers still implement the CSS 2.0 values.

Signature

Spartanicus

Reply to this Message

Webmaster Forum / HTML, CSS, Scripts / CSS / September 2005

Overriding generated content inserted via the :before pseudo element