[jira] [Updated] (OAK-2306) Path restriction evaluation support within LucenIndex

["Chetan Mehrotra (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/OAK-2306?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Chetan Mehrotra updated OAK-2306:
---------------------------------
    Attachment: OAK-2306.patch

Patch which uses PathHierrachyTokenizer to tokenize the paths

* If {{evaluatePathRestrictions}} is configured then instead of default analyzer a {{PerFieldAnalyzerWrapper}} would be used where for {{:ancestors}} field a PathHierrachyTokenizer based analyzer would be configured. For all other fields default {{OakAnalyzer}} would be used
* To support {{DIRECT_CHILDREN}} case path depth would also be indexed
* While indexing the parent path would indexed in {{:ancestors}} field
* On query side 
** ALL_CHILDREN - A term query would be used for the given path for {{:ancestors}} field
** DIRECT_CHILDREN - Boolean query combining depth query and term query would be used
* Added a new {{TokenizerChain}} based on [Solr|https://github.com/apache/lucene-solr/blob/trunk/solr/core/src/java/org/apache/solr/analysis/TokenizerChain.java] impl. This would be later used customizing analyzer usecase

With this setting enabled the index size increased from 64.5 MB to 68.5 MB for index containing 238713 documents

[~alex.parvulescu] [~teofili] Can you have look primarily around the usage of PathHierarchyTokenizer

> Path restriction evaluation support within LucenIndex
> -----------------------------------------------------
>
>                 Key: OAK-2306
>                 URL: https://issues.apache.org/jira/browse/OAK-2306
>             Project: Jackrabbit Oak
>          Issue Type: Improvement
>          Components: oak-lucene
>            Reporter: Chetan Mehrotra
>            Assignee: Chetan Mehrotra
>             Fix For: 1.2
>
>         Attachments: OAK-2306.patch
>
>
> LucenePropertyIndex has some support for evaluating path restrictions. This has been restricted recently with OAK-2134. Currently some testcase make test with path based queries.
> Provide a more performant implemenation based on [PathHierarchyTokenizer|http://lucene.apache.org/core/4_6_1/analyzers-common/org/apache/lucene/analysis/path/PathHierarchyTokenizer.html]



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)