[
https://issues.apache.org/jira/browse/CONNECTORS-1234?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14717712#comment-14717712
]
Shinichiro Abe commented on CONNECTORS-1234:
--------------------------------------------
bq. they are streamed
Does this means home-grown ES httpclient? If yes, this is not constructing a
document in memory which differs from SolrJ client. I'll post very large files
later, I don't have a large file at the moment. Thanks.
> TikaExtractor based indexing on Elasticsearch connector
> -------------------------------------------------------
>
> Key: CONNECTORS-1234
> URL: https://issues.apache.org/jira/browse/CONNECTORS-1234
> Project: ManifoldCF
> Issue Type: Improvement
> Reporter: Shinichiro Abe
> Assignee: Shinichiro Abe
> Attachments: CONNECTORS-1234.patch
>
>
> We could add the use-mapper-attachments flag.
> Default to true, current spec which asks for mapper-attachments plugin on ES
> side.
> If false, it allows us to index the content and metadata that extracted from
> files through Tika transformer, which means there is no need to install that
> plugin and put base64 encoded content.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
