xslt gmuenchian grouping with sub-groups - xslt

Question

I am trying to re-structure xml data into groups and sub-group. I was able to get it to work, but my code has to include something, that looks (at least to me) like a workaround. Here are my sample files:
Data.xml:
<data>
<record Group="g1" SubGroup="sg1">Record 1</record>
<record Group="g2" SubGroup="sg1">Record 2</record>
<record Group="g1" SubGroup="sg1">Record 3</record>
<record Group="g2" SubGroup="sg1">Record 4</record>
<record Group="g2" SubGroup="sg2">Record 5</record>
<record Group="g1" SubGroup="sg2">Record 6</record>
</data>
Stylesheet.xsl:
<?xml version="1.0" encoding="UTF-8"?>
<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" version="1.0">
<xsl:output method="xml" version="1.0" indent="yes" encoding="UTF-8"/>
<xsl:key name="Group" match="record" use="#Group" />
<xsl:key name="SubGroup" match="record" use="#SubGroup" />
<xsl:template match="/data">
<xsl:variable name="Records" select="record"/>
<data>
<xsl:for-each select="$Records[generate-id(.)=generate-id(key('Group',#Group)[1])]">
<xsl:sort select="#Group"/>
<xsl:variable name="Group" select="#Group"/>
<xsl:call-template name="Group">
<xsl:with-param name="Records" select="$Records[#Group = $Group]"/>
<xsl:with-param name="Group" select="$Group"/>
</xsl:call-template>
</xsl:for-each>
</data>
</xsl:template>
<xsl:template name="Group">
<xsl:param name="Records"/>
<xsl:param name="Group"/>
<group name="{$Group}">
<xsl:for-each select="$Records[generate-id(.)=generate-id(key('SubGroup',#SubGroup)[1])]">
<!-- this works: <xsl:for-each select="$Records[generate-id(.)=generate-id(key('SubGroup',#SubGroup)[#Group = $Group][1])]"> -->
<xsl:sort select="#SubGroup"/>
<xsl:variable name="SubGroup" select="#SubGroup"/>
<xsl:call-template name="SubGroup">
<xsl:with-param name="Records" select="$Records[#SubGroup = $SubGroup]"/>
<xsl:with-param name="Group" select="$Group"/>
<xsl:with-param name="SubGroup" select="$SubGroup"/>
</xsl:call-template>
</xsl:for-each>
</group>
</xsl:template>
<xsl:template name="SubGroup">
<xsl:param name="Records"/>
<xsl:param name="Group"/>
<xsl:param name="SubGroup"/>
<subgroup name="{$SubGroup}">
<xsl:for-each select="$Records">
<xsl:copy-of select="."/>
</xsl:for-each>
</subgroup>
</xsl:template>
</xsl:stylesheet>
This is the output generated:
<?xml version="1.0" encoding="UTF-8"?>
<data>
<group name="g1">
<subgroup name="sg1">
<record Group="g1" SubGroup="sg1">Record 1</record>
<record Group="g1" SubGroup="sg1">Record 3</record>
</subgroup>
</group>
<group name="g2">
<subgroup name="sg2">
<record Group="g2" SubGroup="sg2">Record 5</record>
</subgroup>
</group>
</data>
but this is the output, I want to have:
<?xml version="1.0" encoding="UTF-8"?>
<data>
<group name="g1">
<subgroup name="sg1">
<record Group="g1" SubGroup="sg1">Record 1</record>
<record Group="g1" SubGroup="sg1">Record 3</record>
</subgroup>
<subgroup name="sg2">
<record Group="g1" SubGroup="sg2">Record 6</record>
</subgroup>
</group>
<group name="g2">
<subgroup name="sg1">
<record Group="g2" SubGroup="sg1">Record 2</record>
<record Group="g2" SubGroup="sg1">Record 4</record>
</subgroup>
<subgroup name="sg2">
<record Group="g2" SubGroup="sg2">Record 5</record>
</subgroup>
</group>
</data>
The problem is the for-each loop in the tempalte named "Group". It seems, that the key()-function is not working on the nodes contained in $Records but on the entire input XML file.
I get identical results with xsltproc and with saxon, so I do not think, it is a bug in my xslt processor. It seems, that I did not completely understand, how key() works.
If I add an additional selector [#Group = $Group] to key()'s output, I get the expected result.
Can somebody explain what is going on and why the additional selector [#Group = $Group] is needed.
Mario

Answer 1

When you want to do sub-grouping, you need to use a concatenated key of both the main group and the sub-group
<xsl:key name="SubGroup" match="record" use="concat(#Group,'|', #SubGroup)" />
Then, just use it in the same way as before, with the concatenation
<xsl:for-each select="$Records[generate-id(.)=generate-id(key('SubGroup',concat(#Group,'|', #SubGroup))[1])]">
Try this XSLT (which I have also simplified to utilise the key when calling your named templates with the records)
<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" version="1.0">
<xsl:output method="xml" version="1.0" indent="yes" encoding="UTF-8"/>
<xsl:key name="Group" match="record" use="#Group" />
<xsl:key name="SubGroup" match="record" use="concat(#Group,'|', #SubGroup)" />
<xsl:template match="/data">
<data>
<xsl:for-each select="record[generate-id(.)=generate-id(key('Group',#Group)[1])]">
<xsl:sort select="#Group"/>
<xsl:call-template name="Group">
<xsl:with-param name="Records" select="key('Group',#Group)"/>
</xsl:call-template>
</xsl:for-each>
</data>
</xsl:template>
<xsl:template name="Group">
<xsl:param name="Records"/>
<group name="{#Group}">
<xsl:for-each select="$Records[generate-id(.)=generate-id(key('SubGroup',concat(#Group,'|', #SubGroup))[1])]">
<xsl:sort select="#SubGroup"/>
<xsl:call-template name="SubGroup">
<xsl:with-param name="Records" select="key('SubGroup',concat(#Group,'|', #SubGroup))"/>
</xsl:call-template>
</xsl:for-each>
</group>
</xsl:template>
<xsl:template name="SubGroup">
<xsl:param name="Records"/>
<subgroup name="{#SubGroup}">
<xsl:for-each select="$Records">
<xsl:copy-of select="."/>
</xsl:for-each>
</subgroup>
</xsl:template>
</xsl:stylesheet>