suppose I have an XML like this:
<body>
<nested attr="bla">
<name>foo</name>
</nested>
<nested attr="blub">
<name>bar</name>
</nested>
<nested attr="bli">
<name>baz</name>
</nested>
</body>
What I want now is a way to get the index of the nested tag with a certain value. So for example the index of bar would be 1 (or 2 if you count from 1).
I have already done this using Apache CachedXPathApi:
public int getIndex(String path, String value) throws TransformerException {
NodeIterator it = cachedXPathAPI.selectNodeIterator(document, path);
Node node;
int i = 0;
while((node = it.nextNode()) != null) {
if(node.getTextContent().equals(value)) {
return i;
}
++i;
}
return -1;
}
Which works fine for getting the index from the attribute attr like this:
getIndex("/body/nested/@attr", "blub")
But I don’t know how to do this for the nested values. If I use /body/nested/name then it will obviously only ever count the name tags within nested which is not what I want.
How can I solve this, either by changing the Java Code or maybe even with a special XPath expression?
Advertisement
Answer
Just evaluate this XPath expression:
count(/*/*[name='bar']/preceding-sibling::*) + 1
XSLT-based verification:
The following transformation simply evaluates the XPath expression and outputs the result of this evaluation:
<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
<xsl:output omit-xml-declaration="yes" indent="yes"/>
<xsl:template match="/">
<xsl:value-of select="count(/*/*[name='bar']/preceding-sibling::*) +1"/>
</xsl:template>
</xsl:stylesheet>
When applied on the provided XML document:
<body>
<nested attr="bla">
<name>foo</name>
</nested>
<nested attr="blub">
<name>bar</name>
</nested>
<nested attr="bli">
<name>baz</name>
</nested>
</body>
the wanted, correct result is produced:
2
II. Update
In a comment the OP described another case, where there is no element in the XML document that satisfies the filtering condition. In such case the previous expression (above) returns 1 and this is incorrect.
Here is an XPath expression that returns the correct result in all cases:
(boolean(/*/*[name='bar'])) * (count(/*/*[name='bar']/preceding-sibling::*) +1)
This is the previous XPath expression, multiplied by another XPath expression (on its left) which evaluates to 0 if there is no element that satisfies the filtering predicate, and to 1 otherwise.
The XSLT-based verification shows that this XPath expression evaluates to the correct 1-based index values in the case when a satisfying element exists, and to 0 when no such element exists. Here we take advantage of the fact that the implicit conversion number(false()) is 0 and number(true()) is 1 .
Here is the latter example:
<body>
<nested attr="bla">
<name>foo</name>
</nested>
<nested attr="blub">
<name>barr</name>
</nested>
<nested attr="bli">
<name>baz</name>
</nested>
</body>
Here no element has a <name> child with string value 'bar'. And when we apply the transformation:
<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
<xsl:output omit-xml-declaration="yes" indent="yes"/>
<xsl:template match="/">
<xsl:value-of select=
"(boolean(/*/*[name='bar'])) * (count(/*/*[name='bar']/preceding-sibling::*) +1)"/>
</xsl:template>
</xsl:stylesheet>
the correct, wanted result is produced:
0