I have a problem I cannot solve with xslt 1.0. A solution for XSLT 2.0 was already given and works fine (see further below).
For each layer grouped by STRAT_ZONE the minimal values from DEPTHFROM_ZONE and maximum value DEPTHTO_ZONE are needed.
Grouping by STRAT_ZONE seems not an option, because the attribute can repeat itself. A checkup line by line was an idea.
Does anyone have any tips, how to try reach the desired goal?
Datablock:
<ZONES>
<ZONE STRAT_ZONE="Flysand unchained" DEPTHFROM_ZONE="0.00" DEPTHTO_ZONE="0.30" INDEX_ZONE="-1"/>
<ZONE STRAT_ZONE="Flysand unchained" DEPTHFROM_ZONE="0.30" DEPTHTO_ZONE="1.10" INDEX_ZONE="-1"/>
<ZONE STRAT_ZONE="Flysand unchained" DEPTHFROM_ZONE="1.10" DEPTHTO_ZONE="1.40" INDEX_ZONE="-1"/>
<ZONE STRAT_ZONE="Flysand unchained" DEPTHFROM_ZONE="1.40" DEPTHTO_ZONE="1.70" INDEX_ZONE="-1"/>
<ZONE STRAT_ZONE="Drifting sand" DEPTHFROM_ZONE="1.70" DEPTHTO_ZONE="1.80" INDEX_ZONE="-1"/>
<ZONE STRAT_ZONE="Drifting sand" DEPTHFROM_ZONE="1.80" DEPTHTO_ZONE="2.20" INDEX_ZONE="-1"/>
<ZONE STRAT_ZONE="Drifting sand" DEPTHFROM_ZONE="2.20" DEPTHTO_ZONE="2.60" INDEX_ZONE="-1"/>
<ZONE STRAT_ZONE="Flysand unchained" DEPTHFROM_ZONE="2.60" DEPTHTO_ZONE="2.80" INDEX_ZONE="-1"/>
<ZONE STRAT_ZONE="Flysand unchained" DEPTHFROM_ZONE="2.80" DEPTHTO_ZONE="2.90" INDEX_ZONE="-1"/>
<ZONE STRAT_ZONE="Drifting sand" DEPTHFROM_ZONE="2.90" DEPTHTO_ZONE="3.10" INDEX_ZONE="-1"/>
<ZONE STRAT_ZONE="Drifting sand" DEPTHFROM_ZONE="3.10" DEPTHTO_ZONE="3.40" INDEX_ZONE="-1"/>
<ZONE STRAT_ZONE="Drifting sand" DEPTHFROM_ZONE="3.40" DEPTHTO_ZONE="3.70" INDEX_ZONE="-1"/>
<ZONE STRAT_ZONE="Drifting sand" DEPTHFROM_ZONE="3.70" DEPTHTO_ZONE="4.50" INDEX_ZONE="-1"/>
<ZONE STRAT_ZONE="Drifting sand" DEPTHFROM_ZONE="4.50" DEPTHTO_ZONE="4.90" INDEX_ZONE="-1"/>
<ZONE STRAT_ZONE="Vulcanic" DEPTHFROM_ZONE="4.90" DEPTHTO_ZONE="6.00" INDEX_ZONE="-1"/>
<ZONE STRAT_ZONE="Vulcanic" DEPTHFROM_ZONE="6.00" DEPTHTO_ZONE="6.20" INDEX_ZONE="-1"/>
<ZONE STRAT_ZONE="Vulcanic" DEPTHFROM_ZONE="6.20" DEPTHTO_ZONE="6.30" INDEX_ZONE="-1"/>
</ZONES>
Desired output:
ZONE "Flysand unchained" 0.00 1.70 -1
ZONE "Drifting sand" 1.70 2.60 -1
ZONE "Flysand unchained" 2.60 2.90 -1
ZONE "Drifting sand" 2.90 4.90 -1
ZONE "Vulcanic" 4.90 6.30 -1
A Version 2.0 solution looks like that (Thanks Martin Honnen):
<xsl:template match="ZONES">
<xsl:for-each-group select="ZONE" group-adjacent="@STRAT_ZONE">
<xsl:value-of
select="'ZONE', current-grouping-key(), min(current-group()/@DEPTHFROM_ZONE), max(current-group()/@DEPTHTO_ZONE), @INDEX_ZONE"/>
<xsl:text> </xsl:text>
</xsl:for-each-group>
</xsl:template>
But how to realize the same result in XSLT 1.0 using MSXML 6 processor?
Thank you very much for any help.
There are several methods to implement group-adjacent in XSLT 1.0. I believe the most convenient one to use in your case would be this:
XSLT 1.0
<xsl:stylesheet version="1.0"
xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
<xsl:output method="text" encoding="UTF-8"/>
<xsl:key name="adj" match="ZONE" use="generate-id(preceding-sibling::ZONE[not(@STRAT_ZONE = current()/@STRAT_ZONE)][1])" />
<xsl:template match="/ZONES">
<xsl:for-each select="ZONE[generate-id() = generate-id(key('adj', generate-id(preceding-sibling::ZONE[not(@STRAT_ZONE = current()/@STRAT_ZONE)][1]))[1])]">
<xsl:variable name="current-group" select="key('adj', generate-id(preceding-sibling::ZONE[not(@STRAT_ZONE = current()/@STRAT_ZONE)][1]))" />
<xsl:text>ZONE "</xsl:text>
<xsl:value-of select="@STRAT_ZONE"/>
<xsl:text>" </xsl:text>
<xsl:for-each select="$current-group">
<xsl:sort select="@DEPTHFROM_ZONE" data-type="number" order="ascending"/>
<xsl:if test="position() = 1 ">
<xsl:value-of select="@DEPTHFROM_ZONE"/>
</xsl:if>
</xsl:for-each>
<xsl:text> </xsl:text>
<xsl:for-each select="$current-group">
<xsl:sort select="@DEPTHTO_ZONE" data-type="number" order="descending"/>
<xsl:if test="position() = 1 ">
<xsl:value-of select="@DEPTHTO_ZONE"/>
</xsl:if>
</xsl:for-each>
<xsl:text> </xsl:text>
<xsl:value-of select="@INDEX_ZONE"/>
<xsl:text> </xsl:text>
</xsl:for-each>
</xsl:template>
</xsl:stylesheet>
This groups the nodes by the id of the first preceding sibling whose value is different from the current node - and this id is common to all adjacent nodes with the same value.