AnsweredAssumed Answered

Lucene PATH wildcard query performance

Question asked by aspa on Nov 4, 2010
Latest reply on Nov 6, 2010 by smicyk
Hi,

I'm developing an application that will use Alfresco as a backend content repository.
I have a use case I'm implementing currently, where I would need to run frequent PATH wildcard queries
on the repository and I'm trying to figure out if this type of query will pose performance problems.

The PATH queries contain wildcards in the middle of the path and are of the form:

PATH:"/app:company_home/cm:foo/cm:bar/*/*/cm:baz/*"

How does Alfresco implement queries of this form?

With Lucene query language this type of query is implemented using the Lucene
index but the index seems to only store information about the immediate parent of a node
(not ancestors or full path). Is this correct?
This seems to imply that the Alfresco query engine can't get the results using a simple
index search but needs to process each level of wildcards in the middle of the path separetely, right?

In our case the hierarchy levels in which the 2 wildcards in the middle appear would
contain a fairly large number of nodes (~1000 each).

I'm using Alfresco v3.3 and the Alfresco Web Service client API for running the queries.

Outcomes