Difference between revisions of "XPath queries"

Revision as of 10:22, 3 March 2010

Can be found more information about XPATH in http://www.w3.org/TR/xpath20/

Examples

locked documents

//element(*, okm:document)[@jcr:lockOwner]

locked documents by user pavila

//element(*, okm:document)[@okm:author='pavila']

search documents by content

//element(*, okm:document)[jcr:contains(okm:content, 'linux')]

folders created by user pavila

//element(*, okm:folder)[@okm:author='pavila']

example of complex query, documents with content linux and keyword ubuntu

//element(*,okm:document)[jcr:contains(okm:content,'linux') and jcr:contains(@okm:keywords,'ubuntu')]

dereference

//element(*, nt:linkedFile)/jcr:deref(@jcr:content, '*')[jcr:contains(., 'foo')]

Note in the examples we're making queries from repository root node, but really the queries might be done from /jcr:root/okm:root/ - that's the taxonomy node - for example the locked documents might be done as

/jcr:root/okm:root//element(*, okm:document)[@jcr:lockOwner]

Type restrictions

We can specify the type of node that query returns. Restrictions include inheritage all subtype are included as possible results. The function element is used to evaluate a node type.

for example the node folders

//element(*, okm:folder)

Property restrictions

A query can have property restrictions. The operators used are (=, !=, <, <=, >, >=) in addtion there's some extra operators like jcr:like() and jcr:contains().

getting folders with author pavila

//element(*, okm:folder)[@okm:author = 'pavila']

documents name starting with linux

//element(*, okm:document)[jcr:like(@okm:name, 'linux%')]

documents with indexed content jackrabbit

//element(*, okm:resource)[jcr:contains(., 'jackrabbit')]/@jcr:path

Order

At ends query can be indicated the order of results.

ordering by name

 //element(*, okm:document)[@okm:keywords != ] order by @okm:name

Normally is used jcr:score to ordering, that indicates de ranking of the results.