Hi,
i’m a beginner whit xslt then i need help.
I want to delete all elementos son of the node Article that they aren’t except ‘PubModel’ and ‘ArticleDate’. I used the following code. It works but it doesn’t keep attributes in the resulting XML. Why? How can I do?

<xsl:template match="*|/">
    <xsl:copy>
        <xsl:apply-templates/>
    </xsl:copy>
</xsl:template>

<xsl:template match="//PubmedArticle/MedlineCitation/Article/*[name()!='PubModel'][name()!='ArticleDate']

You want the first bit to be the 'identity transform':

<xsl:template match="node()|@*">
   <xsl:copy>
   <xsl:apply-templates select="@*"/>
   <xsl:apply-templates/>
   </xsl:copy>
 </xsl:template>


/- Sam Judson : Wrox Technical Editor -/

Tank you. It work but it keep all attributes.
I wanted a little bit different thing.
For example, if i want to keep MedlineCitation node's child Article, that is an element and MedlineCitation node's attributes Owner and status,
i use the following expression, but i keep also all other attributes. Why?


<xsl:template match="//PubmedArticle/MedlineCitation/*[name()!='Article'][name()!='Owner'][name()!='Status']" />

Tanks a lot again.

I think there may be a terminology problem here. You are referring to "Owner" and "Status" as attributes, but your XPath expression is treating them as child elements. It might be useful if you show us your source XML and desired output XML so we can work out what you are talking about.

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

For example.
If I have this tree structure:

<PubmedArticle>
    <MedlineCitation Owner="NLM" Status="Publisher">
      <PMID>15688454</PMID>
      <MedlineJournalInfo>
        <MedlineTA>J Morphol</MedlineTA>
        <NlmUniqueID>0406125</NlmUniqueID>
      </MedlineJournalInfo>
    </MedlineCitation>
  </PubmedArticle>

And I want this output:

<MedlineCitation Owner="NLM" >
      <MedlineJournalInfo>
        <MedlineTA>J Morphol</MedlineTA>
        <NlmUniqueID>0406125</NlmUniqueID>
      </MedlineJournalInfo>
    </MedlineCitation>
  </PubmedArticle>

I had then delete red elements.
If i use the following template:

<xsl:template match="//PubmedArticle/MedlineCitation/*[name()!='MedlineJournalInfo'][name()!='Owner'] />

It doesn't work. Maybe because Owner isn't a child of MedlineCitation. Then How can i say to xslt processor delete all
MedlineCitation attributes except 'Owner'attribute?

If you want to DELETE the red elements then you simply need the identity template like so:

<xsl:template match="node()|@*">
   <xsl:copy>
   <xsl:apply-templates select="@*"/>
   <xsl:apply-templates/>
   </xsl:copy>
 </xsl:template>

Followed by the matches for the ones you want to delete (i.e. not copy):

<xsl:template match="MedlineCitation/@Status" Priority="2" />
<xsl:template match="MedlineCitation/PMID" Priority="2" />

The priority isn't actually required but it highlights what you are trying to do.

/- Sam Judson : Wrox Technical Editor -/

You seem confused by the difference between attributes and child elements. MedLineJournalInfo is an element, Owner is an attribute. So you want:

<xsl:template match="//PubmedArticle/MedlineCitation/*[name()!='MedlineJournalInfo'] />

<xsl:template match="//PubmedArticle/MedlineCitation/@*[name()!='Owner'] />

Except that this is far more complicated than it needs to be. Better to use the identity template (as given by Sam) and then define empty template rules for the nodes you want to delete, namely match="PMID" and match="@Status".



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

I tried this template,
and now all work properly.


<xsl:template match="//PubmedArticle/MedlineCitation/@*[name()!='Owner'] />


Tanks a lot Michael Kay!
Tanks a lot Sam Judson!