removing duplicate nodes

nguna · February 16th, 2009, 08:50 AM

Hi,
I'm trying to the the following

<xsl:variable name="locationCodes">
<xsl:for-each select="Output/../../../insureditems"> <xsl:variable name="code">
<xsl:call-template name="scrollLocationLink">
<xsl:with-param name="polbr" select="../POLBRR"/> </xsl:call-template>
</xsl:variable>
<xsl:variable name="occupancy">
<xsl:value-of select="OPCT"/>
</xsl:variable>
<xsl:value-of select="concat($code,'/' ,$occupancy)"/>
<xsl:if test="not(position( ) = last( ) )">
<xsl:text>#</xsl:text>
</xsl:if> </xsl:for-each>
</xsl:variable>


<xsl:variable name="listLocationCodes" select="tokenize($locationCodes,'#') "/>

I'm trying to elimate duplicate nodes by doing the following
<xsl:variable name="listLocationCodesNoDups" select="$listLocationCodes[not( .= preceding-sibling:: *)] "/>

this is not working i'm using the xmlspy built in parser.

Anyone could assis me in solving this

Thanks

Martin Honnen · February 16th, 2009, 09:17 AM

tokenize suggests you use XSLT 2.0.
If you want to remove duplicates from a sequence of strings like that returned by the tokenize function then simply use the function distinct-values http://www.w3.org/TR/xpath-functions...istinct-values

Code:

<xsl:variable name="listLocationCodesNoDups" select="distinct-values($listLocationCodes)"/>

mhkay · February 16th, 2009, 09:28 AM

Further to Martin's response, the reason your code doesn't work is that tokenize() returns a sequence of strings (not nodes), and strings don't have preceding siblings. Only nodes have siblings.

nguna · February 16th, 2009, 10:19 AM

martin & micheal ,
thanks for your reply.

I would like to know how i can elimate duplicate nodes if i create a node set as follows

<xsl:variable name="locationCodes">
<xsl:for-each select="Output/../../../insureditems"> <r>
<c1>
<xsl:call-template name="scrollLocationLink">
<xsl:with-param name="polbr" select="../POLBRR"/>
</xsl:call-template>
</c1>
<c2><xsl:value-of select="OPCT"/></c2>
</r>
</xsl:for-each>
</xsl:variable>

<xsl:for-each select="exslt:node-set($locationCodes)/*">
<xsl:if test="$locationCodes[not(.=preceding-sibling:: *)]">
<locCode><xsl:value-of select="c1"/></locCode>
<occupancy><xsl:value-of select="c2"/></occupancy>
</xsl:if>
</xsl:for-each>

i would like to know what is it that i'm not doing correctly

thanks

mhkay · February 16th, 2009, 10:32 AM

I'm having to make guesses because you haven't shown your called template. However,

(a) exslt:node-set($locationCodes)/* selects a sequence that contains <r> elements, whose children are <c1> and <c2> elements, and in that situation, comparing the string value of each <r> element with its preceding siblings, (the string value being the concatenation of the <c1> and <c2> values), would seem a bit surprising.

(b) There's something wrong here:

<xsl:for-each select="exslt:node-set($locationCodes)/*">
<xsl:if test="$locationCodes[not(.=preceding-sibling:: *)]">

because the value of the test doesn't depend on which node you are currenly processing. I think you meant either

<xsl:for-each select="exslt:node-set($locationCodes)/*[not(.=preceding-sibling:: *)]">

or

<xsl:for-each select="exslt:node-set($locationCodes)/*">
<xsl:if test="not(.=preceding-sibling:: *)">

Martin Honnen · February 16th, 2009, 10:32 AM

Try whether

Code:

<xsl:for-each select="exslt:node-set($locationCodes)/*[not(c1 = preceding-sibling::*/cs)]">
  <locCode><xsl:value-of select="c1"/></locCode>
  <occupancy><xsl:value-of select="c2"/></occupancy>
</xsl:for-each>

works.

But I am confused as to why you make use of exslt:node-set, your original post seemed to be XSLT 2.0 which does not need that function at all and which would allow you to use xsl:for-each-group e.g.

Code:

<xsl:for-each-group select="$locationCodes/*" group-by="c1">

nguna · February 16th, 2009, 10:40 AM

Thank you for both of your help.

I wanted to get a solution that works for xsl 1.0 that's why i tried the node-set method.

Thanks