Re: [docbook] Too Strict Content Model for abstract?

[Fekete, Róbert <robert.fekete@balabit.com>]

On Wed, Feb 4, 2015 at 9:10 AM, Thomas Schraitle <tom_schr@web.de> wrote:

Hi Robert,

On Wed, 4 Feb 2015 07:55:38 +0100
"Fekete, Róbert" <robert.fekete@balabit.com> wrote:

> I agree with Thomas, I also dislike embedding not-inline elements in
> <para>.

Thanks, but my main concern is the limited content model of abstract.
The issue with para is more or less my own taste. ;) I wouldn't touch
para, because for some it could still be useful.


> @Simon: A wrapper element that could group any other elements would be
> immensely useful for single-sourcing. Currently I use <topic> for
> that, and have a preprocessing script that deletes the <topic> tags
> after xincludes are resolved, but that makes validating documents a
> pain.

Yes, topic could also be used, however, it can't be nested.

Do you use your XIncludes with XPointer to skip topics or do you apply
a XSLT transformation? Just curious. ;)

 I xinclude the entire topic, resolve the xincludes while generating the profiled xml, then I simply delete the <topic> and </topic> tags with a shell script before processing the profiled xml further. :)

So the topic is essentially a wrapper for me.

 

--
Gruß/Regards,
    Thomas Schraitle