Question

I am trying to produce a nested XML document from a flat XML document as below. After searching here and other resources available on the net it seems that that the muenchain method could be the solution however I am struggling applying it to my situation.

The XML I need to translate is of form :

<i>
  <item id="1" name="one" sub_id="10" sub_name="s1" detail_id="t1" detail_name="aaaa"/>
  <item id="1" name="one" sub_id="10" sub_name="s1" detail_id="t2" detail_name="bbb"/>
  <item id="1" name="one" sub_id="20" sub_name="s2" detail_id="t1" detail_name="ccc"/>
  <item id="1" name="one" sub_id="20" sub_name="s2" detail_id="t2" detail_name="ddd"/>
  <item id="2" name="two" sub_id="10" sub_name="s1" detail_id="t1" detail_name="eee"/>
  <item id="2" name="two" sub_id="10" sub_name="s1" detail_id="t2" detail_name="fff"/>
  <item id="2" name="two" sub_id="20" sub_name="s2" detail_id="t1" detail_name="ggg"/>
  <item id="2" name="two" sub_id="20" sub_name="s2" detail_id="t2" detail_name="hhh"/>
  <item id="3" name="three" />
  <item id="4" name="four" sub_id="10" sub_name="s1" detail_id="t1" detail_name="mmm"/>
  <item id="4" name="four" sub_id="10" sub_name="s1" detail_id="t2" detail_name="nnn"/>
  <item id="4" name="four" sub_id="20" sub_name="s2" detail_id="t1" detail_name="ooo"/>
  <item id="4" name="four" sub_id="20" sub_name="s2" detail_id="t2" detail_name="ppp"/>
</i>

I would like to tranform this to XML in the following form:

  <i>
    <item id="1" name="one" sub_items="true">
      <sub_item sub_id="10" sub_name="s1">
        <detail detail_id="t1" detail_name="aaaa"/>
        <detail detail_id="t2" detail_name="bbb"/>
      </sub_item>
      <sub_item sub_id="20" sub_name="s2">
        <detail detail_id="t1" detail_name="ccc"/>
        <detail detail_id="t2" detail_name="ddd"/>
      </sub_item>
    </item>
    <item id="2" name="two" sub_items="true">
      <sub_item sub_id="10" sub_name="s1">
        <detail detail_id="t1" detail_name="eee"/>
        <detail detail_id="t2" detail_name="fff"/>
      </sub_item>
      <sub_item sub_id="20" sub_name="s2">
        <detail detail_id="t1" detail_name="ggg"/>
        <detail detail_id="t2" detail_name="hhh"/>
      </sub_item>
    </item>
    <item id="3" name="three" sub_items="false"/>
    <item id="4" name="four" sub_items="true">
      <sub_item sub_id="10" sub_name="s1">
        <detail detail_id="t1" detail_name="mmm"/>
        <detail detail_id="t2" detail_name="nnn"/>
      </sub_item>
      <sub_item sub_id="20" sub_name="s2">
        <detail detail_id="t1" detail_name="ooo"/>
        <detail detail_id="t2" detail_name="ppp"/>
      </sub_item>
    </item>
  </i>

From my research I have the following XLST to perform the transform. I use the key method on the item id attribute. This doesn’t group the data correctly repeating everything for the given item id at each level – which makes sense given the key. So my problem is how do I go about selecting the nodes required to output each nest level do I need to use another key statement?

<?xml version="1.0" encoding="utf-8"?>
<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
  <xsl:key name="indexes" match="i/item" use="@id"/>
  <xsl:template match="i">
    <i>
    <xsl:for-each select="item[count(. | key('indexes',@id)[1]) = 1]"  >
      <xsl:sort select="@id"/>
      <item>
        <xsl:attribute name="id">
          <xsl:value-of select="@id"/>
        </xsl:attribute>
        <xsl:attribute name="name">
          <xsl:value-of select="@name"/>
        </xsl:attribute>
        <xsl:attribute name="hasRows">
          <xsl:value-of select="@id"/>
        </xsl:attribute>
        <xsl:for-each select="key('indexes',@id)">
          <subitem>
            <xsl:attribute name ="sub_id">
              <xsl:value-of select="@sub_id"/>
            </xsl:attribute>
            <xsl:attribute name ="sub_name">
              <xsl:value-of select="@sub_name"/>
            </xsl:attribute>
            <xsl:for-each select="key('indexes',@id)">
              <segment>
                <xsl:attribute name ="detail_id">
                  <xsl:value-of select="@detail_id"/>
                </xsl:attribute>
                <xsl:attribute name ="detail_name">
                  <xsl:value-of select="@detail_name"/>
                </xsl:attribute>
              </segment>
            </xsl:for-each>
          </subitem>
        </xsl:for-each>
      </item>
    </xsl:for-each>
    </i>
  </xsl:template>
</xsl:stylesheet>

Also is it possible to populate the *sub_items* attribute with true when subitems/details for an item exist and false when they don’t?

Lastly to improve XSLT understanding/skills can anyone recommend good learning resources?

Answer 1

Notice this XSLT 1.0 transform with use of:

• AVT (attribute value templates) to simplify the attributes definition)
• one main key, used to group first level item
• One second compound key to perform sublevel groupings

• extensive use of xsl:apply-templates with modes, versus xsl:for-each

  <xsl:stylesheet version="1.0"
   xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
  
  <xsl:output method="xml" indent="yes"/>
  <xsl:strip-space elements="*"/>
  
  <xsl:key name="kSubs" match="item" use="concat(@id,'|',@sub_id)"/> 
  <xsl:key name="kItems" match="item" use="@id"/> 
  
  <xsl:template match="i">
      <i>
          <xsl:apply-templates select="item[
              generate-id()
                  = generate-id(key('kItems',@id)[1])]"/>
      </i>
  </xsl:template>
  
  <xsl:template match="item">
      <item id="{@id}" name="{@name}" sub_items="{boolean(@sub_id)}">
          <xsl:apply-templates select=".|following-sibling::item[
              (generate-id()
                      = generate-id(key('kSubs',concat(@id,'|',@sub_id))[1]))
              and
              (./@id 
                      = current()/@id)
              ]" mode="sub_id"/>
      </item>
  </xsl:template>
  
  <xsl:template match="item[@sub_id]" mode="sub_id">
      <sub_item sub_id="{@sub_id}" sub_name="{@sub_name}">
          <xsl:apply-templates select="key('kSubs',concat(@id,'|',@sub_id))" 
           mode="detail"/>
      </sub_item>
  </xsl:template>
  
  <xsl:template match="item" mode="detail">
      <detail detail_id="{@detail_id}" detail_name="{@detail_name}"/>
  </xsl:template>
  
  </xsl:stylesheet>
  

Applied on the input shown in your question, produces:

<i>
   <item id="1" name="one" sub_items="true">
      <sub_item sub_id="10" sub_name="s1">
         <detail detail_id="t1" detail_name="aaaa"/>
         <detail detail_id="t2" detail_name="bbb"/>
      </sub_item>
      <sub_item sub_id="20" sub_name="s2">
         <detail detail_id="t1" detail_name="ccc"/>
         <detail detail_id="t2" detail_name="ddd"/>
      </sub_item>
   </item>
   <item id="2" name="two" sub_items="true">
      <sub_item sub_id="10" sub_name="s1">
         <detail detail_id="t1" detail_name="eee"/>
         <detail detail_id="t2" detail_name="fff"/>
      </sub_item>
      <sub_item sub_id="20" sub_name="s2">
         <detail detail_id="t1" detail_name="ggg"/>
         <detail detail_id="t2" detail_name="hhh"/>
      </sub_item>
   </item>
   <item id="3" name="three" sub_items="false"/>
   <item id="4" name="four" sub_items="true">
      <sub_item sub_id="10" sub_name="s1">
         <detail detail_id="t1" detail_name="mmm"/>
         <detail detail_id="t2" detail_name="nnn"/>
      </sub_item>
      <sub_item sub_id="20" sub_name="s2">
         <detail detail_id="t1" detail_name="ooo"/>
         <detail detail_id="t2" detail_name="ppp"/>
      </sub_item>
   </item>
</i>