[
https://issues.apache.org/jira/browse/OAK-3728?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Thomas Mueller updated OAK-3728:
--------------------------------
Attachment: OAK-3728.patch
I have used the following info for the Oak indexes:
* reference: Oak index for reference lookup.
* uuid: Oak index for UUID lookup (direct lookup of nodes with the mixin
'mix:referenceable').
* nodetype: Oak index for queries with node type, and possibly path
restrictions, for example "/jcr:root/content//element(*, mix:language)"
* counter: Oak index that allows to estimate how many nodes are stored below a
given path, to decide whether traversing or using an index is faster.
* acPrincipalName: Oak index used by authorization to quickly search a
principal by name.
* authorizableId: Oak index used by the user management to quickly search an
authorizable by id.
* principalName: Oak index used by the user management to quickly search a
prinipal by name.
[~anchela] could you check if those descriptions make sense (mainly for the
last 3 indexes)?
> Document indexes in the index itself
> ------------------------------------
>
> Key: OAK-3728
> URL: https://issues.apache.org/jira/browse/OAK-3728
> Project: Jackrabbit Oak
> Issue Type: Improvement
> Components: query
> Reporter: Thomas Mueller
> Assignee: Thomas Mueller
> Fix For: 1.3.12
>
> Attachments: OAK-3728.patch
>
>
> Out-of-the box indexes should be documented, so that people know why they are
> needed. This is mainly needed for Oak indexes; we don't need to enforce this
> for user defined indexes.
> To do that, for each index, a property "info" could be added. Example:
> {noformat}
> /oak:index/nodetype/info = "Oak index used for ..."
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
