[xsl] Is "A != B" equivalent to "not(A = B)"?

["Michael S." <mjs@xxxxxxxxx>]

I've got a stylesheet that produces different results depending on 
whether I use "A != B" or "not(A = B)"?  Is this supposed to happen?  
Can someone explain why?  It's not very intuitive.

This is my XML document:

<root>
<p>"Hallo, <cast>Eeyore</cast>!" said <cast>Pooh</cast>.  "This is 
<cast>Tigger</cast>."</p>
<p>"What is?" said <cast>Eeyore</cast>.</p>
<p>"This," explained <cast>Pooh</cast> and <cast>Piglet</cast> together, 
and <cast>Tigger</cast> smiled his
happiest smile and said nothing.</p>
</root>

The stylesheet that follows is supposed to pick out the cast members:

<xsl:stylesheet
  version="1.0"
  xmlns:xsl="http://www.w3.org/1999/XSL/Transform";
>

<xsl:template match="/">
<root>
<xsl:apply-templates select="//cast[text() != following::cast/text()]"/>
</root>
</xsl:template>

<xsl:template match="cast">
<p><xsl:value-of select="."/></p>
<xsl:text>
</xsl:text>
</xsl:template>

</xsl:stylesheet>

This actually "works" in xsltproc--I get the output I was expecting:

<?xml version="1.0"?>
<root><p>Eeyore</p>
<p>Pooh</p>
<p>Piglet</p>
<p>Tigger</p>
</root>

However, saxon, xalan and sablotron give:

<?xml version="1.0" encoding="utf-8"?><root><p>Eeyore</p>
<p>Pooh</p>
<p>Tigger</p>
<p>Eeyore</p>
<p>Pooh</p>
<p>Piglet</p>
</root>

That is, there are duplicates, which I wasn't expecting--but since three 
out of four processors give this result, I figure it must be the correct 
result.

However, if the XPath expression line is changed to:

<xsl:apply-templates select="//cast[not(text() = 
following::cast/text())]"/>

All four processors give the same result, the result given by xsltproc, 
above.  What's going on?



Michael

--
http://beebo.org


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