[
https://issues.apache.org/jira/browse/OAK-2559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14343178#comment-14343178
]
Chetan Mehrotra commented on OAK-2559:
--------------------------------------
Some more details. This is a regression introduced with changes done for
OAK-2261 where older index definition is migrated to newer index rules based
format. As per earlier format the property names in {{excludePropertyNames}}
were considered in case insensitive way. The migration logic took care of first
part where it created property definition with lowercase name but did not
handled the case when config for a specific property is lookedup. So if
property {{Foo}} is excluded then when config for {{Foo}} is looked up it is
not found and instead it matches the default regex and hence gets indexed.
Due to this all excluded property in previous format got included in newer
format (if they differ in case) and hence would start getting indexed which is
not the intention. The fix now ensure that config lookup is performed in case
insensitive way
> Lucene index rules should be case insensitive
> ---------------------------------------------
>
> Key: OAK-2559
> URL: https://issues.apache.org/jira/browse/OAK-2559
> Project: Jackrabbit Oak
> Issue Type: Bug
> Components: oak-lucene
> Affects Versions: 1.0.9, 1.0.10, 1.0.11
> Reporter: Alex Parvulescu
> Assignee: Chetan Mehrotra
> Fix For: 1.0.12, 1.1.7
>
>
> Following the lucene index definitions update, the ignored properties are
> upgraded as a lower case version, but the rest of the lucene bits (indexing)
> still take the case into account, resulting in the exclude rules being
> ignored, and properties being indexed.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
