Question

I have an XML with records, some records are associated with each other and, as such, I want to group them together in the output.

XML:

<Records>
   <Record id="1" group="10" />
   <Record id="2" group="20" />
   <Record id="3" group="20" />
   <Record id="4" group="20" />
</Records>

Currently, I display

<span>1</span><span>2</span><span>3</span><span>4</span>

What I would like to display is (based on the records having the same group)

<span>1</span><span>2-4</span>

I have looked into using preceding-sibling::Record/@group to see if the grouping has changed between iterations of Records but am struggling to figure out how to achieve the 2-4 grouping I require.

Here is what I have so far, interspersed with some comments to illustrate what I am trying to do:

<xsl:for-each select="Records/Record">
   <xsl:if test="@group != preceding-sibling::Record/@group">
      <!-- obviously here we need 2-4...somehow? -->
      <span><xsl:value-of="@id" /></span>
   </xsl:if>
</xsl:for-each>

Answer 1

This transformation:

<xsl:stylesheet version="1.0"
 xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
 <xsl:output omit-xml-declaration="yes"/>
 <xsl:strip-space elements="*"/>

 <xsl:key name="kFollowing" match="Record"
  use="generate-id(preceding-sibling::*
                         [not(@group = current()/@group)
                          ][1])"/>

 <xsl:template match="/*">
       <xsl:apply-templates mode="makeGroup" select=
       "Record[not(@group = preceding-sibling::*[1]/@group)]"/>
 </xsl:template>

 <xsl:template match="Record" mode="makeGroup">
  <xsl:variable name="vGroup"
    select="key('kFollowing', generate-id(preceding-sibling::*[1]))"/>
  <span>
   <xsl:value-of select="$vGroup[1]/@id"/>
   <xsl:if test="$vGroup[2]">
    <xsl:value-of select="concat('-', $vGroup[last()]/@id)"/>
   </xsl:if>
  </span>
 </xsl:template>
</xsl:stylesheet>

when applied on the provided XML document:

<Records>
    <Record id="1" group="10" />
    <Record id="2" group="20" />
    <Record id="3" group="20" />
    <Record id="4" group="20" />
</Records>

produces the wanted, correct result:

<span>1</span><span>2-4</span>

Explanation:

1. This is positional grouping using a key to define all adjacent Record elements that comprise a group.

2. This is an efficient (sublinear) algorithm because keys are used. Algorithms using a siblings axis are typically O(N^2) — quadratical in time complexity and can be too slow if the total number of siblings N is big.