[
https://issues.apache.org/jira/browse/VXQUERY-32?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14378179#comment-14378179
]
Steven Jacobs commented on VXQUERY-32:
--------------------------------------
The code that has been implemented so far is as follows:
1. There is a new syntax which will create a full Lucene index on a given
collection's current state
2. There is a new syntax to run an arbitrary query on the index rather than the
collection itself
This code is currently slightly out of sync with the codebase (but will not be
very hard to reintegrate). After that the next steps would be to find ways of
optimizing query creation and usage, For example, dynamically deciding
when/where an index can/should be used for a query. You will need to
familiarize yourself with the VXQuery codebase.
> VXQuery integration with Apache Lucene
> --------------------------------------
>
> Key: VXQUERY-32
> URL: https://issues.apache.org/jira/browse/VXQUERY-32
> Project: VXQuery
> Issue Type: Improvement
> Reporter: Vinayak Borkar
> Assignee: Preston Carman
> Labels: gsoc, gsoc2015, java, mentor, xml, xquery
>
> XML documents contain textual content and the need to search large corpora of
> documents is a fairly common task. The intent of this project is to leverage
> Apache Lucene's indexing and search capability so that users of the VXQuery
> engine can express and run text-search queries.
> This project will have two parts.
> 1. Design and implement the ability for users to create and manage text
> indices on collections of XML documents.
> 2. Implement functions in VXQuery to exploit these text-indices to execute
> relevant queries efficiently.
> As a starting point, the system does not need to do automatic index selection
> (decide to use the text index when the query did not actually refer to an
> index). Instead the functions would be used directly in the query and the
> system would have to use the said indices.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
