Re: [Xsltforms-support] New release: Please give it a try!

[Steven P. <ste...@cw...>]

Am I right in thinking that

    <output class="{@foo} value="bar"/>

doesn't work yet?

(Working to get the testsuite running under the new release).

Best wishes,

Steven

On Wed, 05 Feb 2020 21:15:57 +0100, Alain Couthures  
<ala...@ag...> wrote:

> Hello,
>
> Please find a new release for XSLTForms at  
> www.agencexml.com/1.5beta/xsltforms.zip
>
> It has not yet been fully tested because a lot of changes have been made  
> and you are welcome to locate remaining issues with your own forms.
>
> The XSLT part has been reduced to minimal for better performance.  
> Instead of parsing the XPath expressions and transforming all the XForms  
> >elements into HTML elements, it basically just transposes the non-HTML  
> elements into sort-of custom elements: xforms:* elements become xforms-*  
> >elements with xf-* and ev-* attributes.
>
> Have a look with your favorite browser debugger! Actually, authors could  
> even prefer to directly write/generate forms with this new notation and  
> forget >about the XSLT step. You can compare two sources for the same  
> form: hello.xml and hello.htm
>
> XSLTForms Javascript classes constructors are obtaining their properties  
> directly from xf-* attributes and XPath parsing is then performed.
>
> No ids are automatically added as previously.
>
> Extra xf-* attributes and extra xforms-* elements are used to embed  
> effective HTML rendering elements, for example, xforms-body or  
> xforms-repeat->item while, before, span or div elements where used.
>
> XSLTForms classes for xforms:select and xforms:itemset had to be  
> partially rewritten.
>
> SVG support has been basically tested too.
>
> CSS styling is not anymore based on xforms-* classes but on custom  
> element names and attribute selectors. For example, the extra xf-bound  
> >attribute, when present, says that the XForms control is bound to a  
> node, eventually a not relevant one, and the extra xf-notrelevant  
> attribute can, then, >be checked...
>
> Thank you for your contribution!
>
> --Alain