[xsl] d-o-e in extension element contents (was Re: Welcome comments...)

Subject: [xsl] d-o-e in extension element contents (was Re: Welcome comments...)
From: Mike Brown <mike@xxxxxxxx>
Date: Fri, 7 Dec 2001 10:42:47 -0700 (MST)
Joerg Pietschmann wrote:
> No processor will honor a d-o-e directive unless it serializes the 
> result.

True, but there is some ambiguity in when serialization of the result may
occur. It can happen after the result tree is created, but it may also
happen when processing an extension element. The XSLT spec doesn't seem to
address the possibility that d-o-e can be honored by an extension element
as it processes the xsl:value-of and xsl:text elements in its contents.

For example, here's a situation we have in 4XSLT with 4Suite Server:

We have an extension element that acts upon the serialized-as-XML form of
any result fragments that are generated when instantiating its content. 


In this example, the foo:our-element implementation takes the 'greeting' 
element and its 'hello' text node child, serializes it as XML into the 
string '<greeting>hello</greeting>', and passes this string to some other 
function in the server for additional voodoo to be performed on it.

There are times when you may already have the desired XML serialization as
a string. In such a case you can't just plop it in an xsl:value-of inside
the extension element, like in the following example, because you'll
subject it to escaping as it is reserialized.

<xsl:variable name="XMLstring" select="'&lt;greeting&gt;hello&lt;/greeting&gt;'"/>
  <xsl:value-of select="$XMLstring"/>

And you can't just change the extension element's implementation by 
plopping the string into an attribute on the element, because attributes 
are subjected to normalization by the XML parser.

Another idea might be to implement an extension function that takes the
string as an argument -- but this has the problem that it might be
subjected to lazy evaluation. That is, all functions in XPath/XSLT must
return the same result when given the same arguments, so a processor might
cache the results the first time around and not bother, when processing
subsequent function calls, to do the entire instantiation with whatever
side-effects it may produce.

So in this situation, unless I'm overlooking something, we are left with
no option but to honor disable-output-escaping on the xsl:value-of inside
the extension element.

  <xsl:value-of select="$XMLstring" disable-output-escaping="yes"/>

   - Mike
  mike j. brown, fourthought.com  |  xml/xslt: http://skew.org/xml/
  denver/boulder, colorado, usa   |  personal: http://hyperreal.org/~mike/

 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list

Current Thread