[
https://issues.apache.org/jira/browse/IOTDB-3723?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17568380#comment-17568380
]
liaolanyu commented on IOTDB-3723:
----------------------------------
Design doc: https://apache-iotdb.feishu.cn/docx/doxcnXETf9GowOrSiedcMr3oNlg
> [IOTDB-3723] Replace FilterOperator with FilterAndProjectOperator and use
> batch processing for better performance
> -----------------------------------------------------------------------------------------------------------------
>
> Key: IOTDB-3723
> URL: https://issues.apache.org/jira/browse/IOTDB-3723
> Project: Apache IoTDB
> Issue Type: Task
> Reporter: Yuan Tian
> Assignee: liaolanyu
> Priority: Major
> Labels: pull-request-available
>
> Now, query logics in TransformOperator and FilterOperator are too complex. It
> is caused by some cases related to udf.
> If we split pure ValueFilter query and expression calculaton with UDF, we can
> simplify and accelerate valuefilter query.
> And futhermore, we can use code gen to generate code to do the calculation
> more effectively.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
