<xsl:for-each select="//eSubpackages/eClassifiers">
<xsl:variable name="classifier" select="@name"/>
<xsl:for-each select="eStructuralFeatures/eAnnotations/details[@key='tabGroup'][not(@value=preceding::eClassifiers/[@name=$classifier]/eStructuralFeatures/eAnnotations/details[@key='tabGroup']/@value)]">

Can some one help me why the preceding axis is not working when I try to scope (the element where I would like to search) as in the sample above?

I think
preceding::eClassifiers/[@name=$classifier]
needs to be
preceding::eClassifiers[@name=$classifier]

--
  Martin Honnen
  Microsoft MVP - XML

You say its not working but you don't say what you are trying to achieve - nor what error you are getting (if indeed you are getting one). Why is it not working?

You also don't show us an example of your XML, so we have no idea what the above is going to generate...


/- Sam Judson : Wrox Technical Editor -/

The only thing obviously wrong with your code is a syntax error

preceding::eClassifiers/[@name=$classifier]

I suspect the "/" should simply be removed.

If that's not your problem, then we would need to know something about your input and your desired output.

I don't know what you mean by "try to scope (the element where I would like to search)".

I should also point out that this construct is likely to be very inefficient if your source document is large. To eliminate duplicates, it's much more efficient to use xsl:for-each-group in XSLT 2.0, or Muenchian grouping in XSLT 1.0.

Michael Kay
http://www.saxonica.com/
Author, XSLT 2.0 and XPath 2.0 Programmer's Reference

No issue with the syntax. That was a typo in the post. Sorry for that.I have no issue with efficiency as the documents are not huge in size. What I could not get right was duplicate checking within a particular subtree rather than the whole document

Following is the info about the issue

<xsl:for-each select="//eSubpackages/eClassifiers">
<xsl:for-each select="eStructuralFeatures/eAnnotations/details[@key='tabGroup'][not(@value=preceding::eStructuralFeatures/eAnnotations/details[@key='tabGroup']/@value)]">
<xsl:value-of select="@value"/>
</xsl:for-each>
</xsl:for-each>

Preceding is filtering the duplicates as expected using the above piece of code i.e it is able to output a "tabGroup" only when there is no ssame preceding "tabGroup". However, duplicates are no problem across "eClassifiers" but they are a problem only within a "eClassifier". Using the above approach, duplicates are being avoided even across "eClassifiers" i.e if a tabgroup is already present in one "eClassifier", it is filtered in the other "eClassifier" as well.

So, I have used the following to restrict the duplicate checking within an "eClassifier"

<xsl:for-each select="//eSubpackages/eClassifiers">
<xsl:variable name="classifier" select="@name"/>
<xsl:for-each select="eStructuralFeatures/eAnnotations/details[@key='tabGroup'][not(@value=preceding::eClassifiers[@name=$classifier]/eStructuralFeatures/eAnnotations/details[@key='tabGroup']/@value)]">
</xsl:for-each>
</xsl:for-each>

problem is that I get duplicates now. Not sure what is missing.

We've still no idea what your input XML looks like, what you want your output to look like, or the output you are currently getting.

/- Sam Judson : Wrox Technical Editor -/

 <eClassifiers xsi:type="ecore:EClass" name="BusinessProcess" eSuperTypes="platform:/plugin/com.ibm.team.repository.common/models/repository.ecore#//repository/SimpleItem">
  <eStructuralFeatures xsi:type="ecore:EReference" name="subProcess" upperBound="-1"
          eType="#//blueharmonypilot/SubProcess">
        <eAnnotations source="sednaProperty">
          <details key="inverse" value="processes"/>
          <details key="tabGroup" value="Process Details"/>
          <details key="tab" value="Sub Processes"/>
        </eAnnotations>
      </eStructuralFeatures>
      <eStructuralFeatures xsi:type="ecore:EReference" name="gap" upperBound="-1"
          eType="#//blueharmonypilot/Gap">
        <eAnnotations source="sednaProperty">
          <details key="inverse" value="processes"/>
          <details key="tabGroup" value="Design Considerations"/>
          <details key="tab" value="Gaps"/>
        </eAnnotations>
<eStructuralFeatures>
<eStructuralFeatures xsi:type="ecore:EReference" name="processModel" eType="#//blueharmonypilot/ProcessModel">
        <eAnnotations source="sednaProperty">
          <details key="tabGroup" value="Process Details"/>
          <details key="tab" value="Process Model"/>
        </eAnnotations>
</eStructuralFeatures>
</eClassifiers>

<eClassifiers xsi:type="ecore:EClass" name="SubProcess" eSuperTypes="platform:/plugin/com.ibm.team.repository.common/models/repository.ecore#//repository/SimpleItem">
  <eStructuralFeatures xsi:type="ecore:EReference" name="requirement" upperBound="-1"
          eType="#//blueharmonypilot/Requirement">
        <eAnnotations source="sednaProperty">
          <details key="inverse" value="subProcesses"/>
          <details key="tabGroup" value="Process Details"/>
          <details key="tab" value="Requirements"/>
        </eAnnotations>
      </eStructuralFeatures>

      <eStructuralFeatures xsi:type="ecore:EReference" name="gap" upperBound="-1"
          eType="#//blueharmonypilot/Gap">
        <eAnnotations source="sednaProperty">
          <details key="inverse" value="subProcesses"/>
          <details key="tabGroup" value="Design Considerations"/>
          <details key="tab" value="Gaps"/>
        </eAnnotations>
      </eStructuralFeatures>
<eClassifiers>

Above is the input xml, it has two "eClassifiers" and the values (@value) for tabGroups (@tabGroup) are "Process Details", "Design Considerations", "Process Details" (for the first eClassifier), "Process Details", Design Considerations" (for the second eClassifier). I would like to avoid duplicates only within an "eClassifier" not across. As I mentioned in the earlier post, usng the following, I get duplicates avoided across the eClassifiers.

<xsl:for-each select="//eSubpackages/eClassifiers">
<xsl:for-each select="eStructuralFeatures/eAnnotations/details[@key='tabGroup'][not(@value=preceding::eStructuralFeatures/eAnnotations/details[@key='tabGroup']/@value)]">
<xsl:value-of select="@value"/>
</xsl:for-each>
</xsl:for-each>

When I try to restict the duplicate checking using the following, I get duplicates infact.
<xsl:for-each select="//eSubpackages/eClassifiers">
<xsl:variable name="classifier" select="@name"/>
<xsl:for-each select="eStructuralFeatures/eAnnotations/details[@key='tabGroup'][not(@value=preceding::eClassifiers[@name=$classifier]/eStructuralFeatures/eAnnotations/details[@key='tabGroup']/@value)]">
</xsl:for-each>
</xsl:for-each>

>What I could not get right was duplicate checking within a particular subtree rather than the whole document

I think that for your particular data you should be using the preceding-sibling axis rather than the preceding axis. Something like

<xsl:for-each select="eStructuralFeatures[not(eAnnotations/details[@key='tabGroup']/@value=preceding-sibling::eStructuralFeatures/eAnnotations/details[@key='tabGroup']/@value)]/eAnnotations/details[@key='tabGroup']/@value)">

Michael Kay
http://www.saxonica.com/
Author, XSLT 2.0 and XPath 2.0 Programmer's Reference

Precise outputs
--------------
When I use the following,

<xsl:for-each select="//eSubpackages/eClassifiers">
<xsl:variable name="classifier" select="@name"/>
<xsl:for-each select="eStructuralFeatures/eAnnotations/details[@key='tabGroup'][not(@value=preceding::eStructuralFeatures/eAnnotations/details[@key='tabGroup']/@value)]">


eClassifier name <xsl:value-of select="../../../@name"/>

current tab group <xsl:value-of select="@value"/>

</xsl:for-each>
</xsl:for-each>




output is
---------
eClassifier name BusinessProcess
current tab group Process Details

eClassifier name BusinessProcess
current tab group Design Considerations

When I use the following,

<xsl:for-each select="//eSubpackages/eClassifiers">
<xsl:variable name="classifier" select="@name"/>
<xsl:for-each select="eStructuralFeatures/eAnnotations/details[@key='tabGroup'][not(@value=preceding::eClassifiers[@name=$classifier]/eStructuralFeatures/eAnnotations/details[@key='tabGroup']/@value)]">


eClassifier name <xsl:value-of select="../../../@name"/>

current tab group <xsl:value-of select="@value"/>
</xsl:for-each>
</xsl:for-each>


the output is
-------------

eClassifier name BusinessProcess
current tab group Process Details


eClassifier name BusinessProcess
current tab group Design Considerations

eClassifier name BusinessProcess
current tab group Process Details

eClassifier name SubProcess
current tab group Process Details

eClassifier name SubProcess
current tab group Design Considerations

Thanks Michael!