Re: Expensive XSLT2 - suggestions for improving?

Right after putting together the test case I tried some xsl:key
myself... I should have done it before.

Your method of integrating the check into a common template with the
'first' node as a variable serves another purpose: I need multiple
attribute values from that node to construct my unique identifier.

For my 20 MByte XML file, the processing is now down from 42 minutes
(!) to 6 seconds. This is more than 99% percent saved just because of
a xsl:key!

something like this:

<xsl:key name="v" match="value" use="."/>

<xsl:template match="value">
  <xsl:copy>
    <xsl:apply-templates select="@*"/>
    <xsl:variable name="first" select="key('v',')[1]"/>
    <xsl;choose>
    <xsl:when test=". is $first">
    <xsl:attribute name="refoid" select="$first/@oid"/>
    </xsl:when>
    <xsl:otherwise>
     <xsl:apply-templates/>
    </xsl:otherwise>
    </xsl:choose>

  </xsl:copy>
</xsl:template>

if you need to do same in xslt1 then

. is $first

becomes

generate-id(.) = generate-id($first)

and

<xsl:attribute name="refoid" select="$first/@oid"/>

becomes

    <xsl:attribute name="refoid"><xsl:value-of select="$first/@oid"/
></xsl:attribute>

David

--
_______________________________________________________________
Michael M|ller-Hillebrand: Dokumentations-Technologie
Adobe Certified Expert, FrameMaker
Lvsungen und Training, FrameScript, XML/XSL, Unicode
<http://cap-studio.de/> -- Tel. +49 (9131) 28747