JCR query by path

java jackrabbit jcr jcr-sql2
light_303 picture light_303 · Apr 2, 2013 · Viewed 8.8k times · Source

i am trying to query a subtree for a given node type by a this query:

SELECT * FROM [my:Type] AS my WHERE PATH(my) LIKE '/content/subtree/%'

somehow it does not give any results - using ISDESCENDANTNODE works - but query performance is terrible as soon as i have >5k elements total - no matter if inside or outside my subtree.

Answer

Thomas Mueller picture Thomas Mueller · Apr 3, 2013

I don't think that PATH(my) is valid JCR SQL or JCR 2.0 SQL-2 grammar. You should use SQL-2:

select * from [my:Type] where isdescendantnode('/content/subtree')

or XPath

/jcr:root/content//element(*, my:Type)

This may be slow because it will only use the index on the node type (the path is not indexed). If you need it to be faster, you could (for example) store the path as a property and then add the corresponding condition; however this will prevent fast move operations.

Related questions

How to run query on Apache Jackrabbit, explain with example

I am using Apache Jackrabbit as a database. In my case, root node has numbers of child nodes(only at depth 1). All child node has unique name, i.e., some Integer. Each child Node have some properties that I have …

Read More
What does ModeShape offer that JackRabbit doesn't?

I just familiarized myself with Apache JackRabbit. I've done a little multi-user repository for document management. If anybody used both of them, could you please answer these questions ? Is ModeShape somehow linked to JBoss ? I don't have much experience with …

Read More
How to access remote jackrabbit repository?

I need to work with remote jackrabbit repository. I use following code to connect to the local repository: Repository repository = new TransientRepository(); Session session = repository.login(new SimpleCredentials("username", "password".toCharArray())); and this works for the local repository but what …

Read More