RE: [xsl] <xsl:number> starting at a preceding-sibling

Subject: RE: [xsl] <xsl:number> starting at a preceding-sibling
From: Hermann Stamm-Wilbrandt <STAMMW@xxxxxxxxxx>
Date: Tue, 27 Jan 2009 14:39:06 +0100
Michael,

> The specification for xsl:number in XSLT 1.0 is less than perfect,
> particularly as regards the meaning of the "from" pattern, and as a
result
> there are quite a few differences between implementations.
thank you for your assessment!

> Why not make the move to XSLT 2.0 and Saxon?
I am developer of another XSLT processor ... :-)

But later I will try Saxon-B on below problem.


Hermann Stamm-Wilbrandt
Developer, XML Compiler
WebSphere DataPower SOA Appliances
----------------------------------------------------------------------
IBM Deutschland Research & Development GmbH
Vorsitzender des Aufsichtsrats: Martin Jetter
Geschdftsf|hrung: Erich Baier
Sitz der Gesellschaft: Bvblingen
Registergericht: Amtsgericht Stuttgart, HRB 243294

----- Forwarded by Hermann Stamm-Wilbrandt/Germany/IBM on 01/27/2009 02:29
PM -----

             "Michael Kay"
             <mike@xxxxxxxxxxx
             m>                                                         To
                                       <xsl-list@xxxxxxxxxxxxxxxxxxxxxx>
             01/27/2009 11:57                                           cc
             AM
                                                                   Subject
                                       RE: [xsl] <xsl:number> starting at
             Please respond to         a preceding-sibling
             xsl-list@xxxxxxxx
              lberrytech.com








> While this works for xsltproc (compiled against libxml 20626,
> libxslt 10117 and libexslt 813) it does not work for
> xalan-j_2_7_0 -- is that an xalan bug?

The specification for xsl:number in XSLT 1.0 is less than perfect,
particularly as regards the meaning of the "from" pattern, and as a result
there are quite a few differences between implementations. I can't
immediately see how this result can be justified, so by all means raise it
as a bug, but you may not get very far.

Why not make the move to XSLT 2.0 and Saxon?

Michael Kay
http://www.saxonica.com/

Current Thread