[
https://issues.apache.org/jira/browse/LUCENE-6371?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14512953#comment-14512953
]
Paul Elschot commented on LUCENE-6371:
--------------------------------------
I just investigated this a little by collecting these:
{code}
public class IndexedPayload {
public int position;
public Term term;
public BytesRef bytesRef;
}
{code}
But now I'm realizing that the problem here is that we don't know in advance
which indexed information is available and should be provided for each term:
the term itself, its position, payload and/or offsets.
This is similar to a TokenStream and its attributes, so why not provide a
TokenStream instead?
This would boil down to providing a subset (matching the query) of the
TokenStream used to index the document.
The PrefillTokenStream of LUCENE-5687 could maybe also be used for this.
The question is: could that work well?
Would such a TokenStream be good for highlighting, too?
> Improve Spans payload collection
> --------------------------------
>
> Key: LUCENE-6371
> URL: https://issues.apache.org/jira/browse/LUCENE-6371
> Project: Lucene - Core
> Issue Type: Improvement
> Reporter: Paul Elschot
> Priority: Minor
>
> Spin off from LUCENE-6308, see the comments there from around 23 March 2015.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
