Hi,

thanks for the detailed explanation.

You are right, these are two 'issues', I confused them because
the Python SAX parser I use replaces both the predefined and the not
predefined entity references, which is ok. I simply assumed an XSLT
processor would also replace both, but that assumption is probably wrong.

I don't know why I prefer ä over 'ä', maybe because 7-bit
ASCI seems to be more portable, but I can't really find a use case
where 'ä' would be less portable.

Thanks, Stephan

Martin Honnen wrote:
> 
> 
> Stephan Hoffmann wrote:
> 
> 
>> I use XML mainly as a source for HTML. HTML browsers 'know'
>> certain entity references like é or ä.
>> 
>> When I use XSL to transform XML to HTML or XML, these entities are
>> replaced by what they refer to.
>> 
>> Is there a way to avoid that?
> 
> XSLT/XPath 1.0 at least which is the current version and the one
> implemented by lots of processors and in wide-spread use does not
> provide anything in its data model or in its instructions to create
> entity references and to ensure that these are preserved and not
> replaced by the entity content when the result of a transformation is
> serialized.
> You would need to look at a specific XSLT processor and check whether it
> provides any mechanisms outside the standards to deal with entity and
> entity references.
> Saxon 6 has an extension function documented here:
> <http://saxon.sourceforge.net/saxon6.5.4/extensions.html#saxon:entity-ref>
> 
>> Two reasons to avoid that:
>> - On my linux machine xsltproc replaced the entities in a way that
>> my browser did not correctly display the resulting HTML
>> (I updated my linux distribution and it now works).
>> 
>> - &lt; is replaced by < and the output is no longer valid XML/HTML
> 
> But &lt; and &gt; are references to entities predefined in XML and
> certainly if any application supposed to output XML or HTML outputs &lt;
> as a plain '<' character then the application is seriously broken.
> This is a different issue, those characters '<' and '>' are obviously
> special as they delimit tags in both XML and HTML and therefore need to
> be escaped as &lt; respectively &gt;.
> &auml; in HTML 4 stands for the character 'ä' and that has no special
> meaning in XML or HTML so if an XSLT processor or other application
> supposed to output XML or HTML simply inserts 'ä' instead of &auml; in a
> document properly encoded and with the proper encoding used and declared
> then there are no problems with well-formedness (or even validity).
> 
> 
>