[ https://issues.apache.org/jira/browse/OAK-637?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jukka Zitting resolved OAK-637.
-------------------------------
Resolution: Fixed
Fix Version/s: 0.7
Added an optimization in 1447649 (and a related integration test fix in revision 1447660).
The result for login() calls is pretty impressive:
{noformat}
# Login min 10% 50% 90% max
before 2040 2042 2055 2169 2576
after 947 951 986 1082 1201
{noformat}
> Optimize SelectorImpl.evaluateTypeMatch()
> -----------------------------------------
>
> Key: OAK-637
> URL: https://issues.apache.org/jira/browse/OAK-637
> Project: Jackrabbit Oak
> Issue Type: Improvement
> Components: core
> Reporter: Jukka Zitting
> Assignee: Jukka Zitting
> Fix For: 0.7
>
>
> Node type evaluation in {{SelectorImpl}} is quite expensive especially for things like UUID lookups that produce just a single result and use {{nt:base}} as the node type. In such cases the current implementation ends up reading and caching the entire set of node types just to determine that actually any node matches nt:base.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira