[pmwiki-users] Re: Binding several classes in wiki styles

[Dominique Faure]

2005/11/22, Dominique Faure <dominique.faure at gmail.com>:
> 2005/11/22, Patrick R. Michaud <pmichaud at pobox.com>:
> > On Tue, Nov 22, 2005 at 03:58:14PM +0100, Dominique Faure wrote:
> > > I'm looking for a mean to "merge" style definitions into wiki text
> > > without the help of local customization file. Currently, the following
> > > text:
> > >
> > >   %style1 style2% Stylized text
> > >
> > > is rendered as:
> > >
> > >   <span class='style2' > Stylized text</span>
> > >
> > > using the last declared style class. How about a small style syntax
> > > enhancement to allow the rendering of:
> > >
> > >   %style1 +style2% Stylized text
> > >
> > > as:
> > >
> > >   <span class='style1 style2' > Stylized text</span>
> >
> > Would this be only for class= attributes?  In other words, if
> > we have
> >
> >   %define=style1 color=red%
> >   %define=style2 color=blue%
> >
> > would you expect %style1 +style2% to become ... ?
> >
> >    <span style='color:red; color:blue;'>
> >
>
> I was primarily thinking of the 'class' attributes.
> Anyway, I found a solution myself with regular style definitions as:
>
>  %class='style1 style2'%  Stylized text

If the above tips is working, it suffers for the tight CSS binding: My
first goal was to provide a way to allow users to remove the external
link decoration I introduced with the monobook skin.

Applying a %wikilink% style to these links works like a charm until
they are not already class-stylized (with the %frame% style for
example). In this case, the style definition should be %class='frame
wikilink'% to get the expected result.

This not-really-user-friendly syntax is getting worse when trying to
deal with already %lframe% or %rframe% stylized links. There, the
required style definition should become %class='lfloat frame
wikilink'% or %class='rfloat frame wikilink'% since %[rl]frame% aren't
"atomic CSS classes.

I just wanted to be able to write something like %rframe +wikilink%,
which in my case could resolve to a merge of class attributes only.
OTOH, CSS allow multiple declaration of the same attribute (with the
last one as winner), so this could be acceptable.

Regards,
Dom