[
https://issues.apache.org/jira/browse/LUCENE-9203?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17030005#comment-17030005
]
juan camilo rodriguez duran commented on LUCENE-9203:
-----------------------------------------------------
Hi, [~jpountz]
Currently I'm working on a refactor for the doc values API based on a
CompoundDocValuesFormat the main idea behind is to break the code in
Lucene80DVConsumer/Producer in proper classes with single responsibility. All
this in order to leverage the existing code and just plug an extension for
sorted and sorted set terms dictionary.
This "new" format uses a single file, so the meta data for a given field is
held in memory until all "data" finishes then the meta data (NumericEntry,
SortedEntry, etc) is flushed at the end. BTW I would like to know the reason
to dont reuse the same class when writing.
I'm looking forward for your feedback.
> Make DocValuesIterator public
> -----------------------------
>
> Key: LUCENE-9203
> URL: https://issues.apache.org/jira/browse/LUCENE-9203
> Project: Lucene - Core
> Issue Type: Improvement
> Components: core/index
> Affects Versions: 8.4
> Reporter: juan camilo rodriguez duran
> Priority: Trivial
> Labels: docValues
>
> By doing this, we improve extensibility for new formats. Additionally this
> will improve coherence with the public method already existent in the class.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org
For additional commands, e-mail: issues-h...@lucene.apache.org
