Selecting nodes depending on sibling node criteria using XSLT

Question

Suppose I have XML like this:

<child_metadata>
    <metadata>
        <attributes>
                <metadata_key value="include"/>
                <metadata_value value="value1"/>
        </attributes>
    </metadata>
    <metadata>
        <attributes>
                <metadata_key value="dont_include"/>
                <metadata_value value="value2"/>
        </attributes>
    </metadata>
    <metadata>
        <attributes>
                <metadata_key value="include"/>
                <metadata_value value="value3"/>
        </attributes>
    </metadata>
</child_metadata>

Using XSLT (without any extensions), I want to select only those values for which the metadata_key is "include". So, for this example, I want to select value1 and value3.

How do I do this?

Answer 1

The XPath expression to use (for a template or in a for-each loop) would be:

//metadata_value[../metadata_key/@value='include']/@value

Since it isn't clear what you mean by "select" I cannot post a complete XSLT sample. XSLT is for transformation, not query ("selection" of data); the query is done with XPath expressions such as the one I provided you above.

Answer 2

For the metadata_value node set, use

//metadata/attributes[metadata_key/@value='include']/metadata_value

add @value to get the values themselves.

Answer 3

I think you want

//atributes[metadata_key/@value='include']/metadata_vale/@value

or (as Lucero post)

//metadata_value[../metadata_key/@value='include']/@value

But, depending on what you want to do, you should use keys because here is a cross-reference (metadata_key/@value is used as the key to metadata_value/@value). It is also good to avoid the descendant axis.

Answer 4

Use:

/child_metadata/metadata/attributes[metadata_key/@value='include']/metadata_value/@value

Using the // abbreviation often has poor efficiency as this causes the whole (sub) tree to be traversed.