[ 
https://issues.apache.org/jira/browse/FLINK-13547?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16914052#comment-16914052
 ] 

Kevin Zhang edited comment on FLINK-13547 at 8/23/19 8:33 AM:
--------------------------------------------------------------

Hi [~docete], [~jark],is there substitute standard functions for the removed 
functions, like jsonvalue? Some of these removed functions are very useful and 
we are using them in our project, otherwise we could only add these functions 
by ourselves.


was (Author: kevinzwx):
Hi [~docete] [~jark],is there substitute standard functions for the removed 
functions, like jsonvalue? Some of these removed functions are very useful and 
we are using them in our project, otherwise we could only add these functions 
by ourselves.

> Verify and correct string function's semantic for Blink planner
> ---------------------------------------------------------------
>
>                 Key: FLINK-13547
>                 URL: https://issues.apache.org/jira/browse/FLINK-13547
>             Project: Flink
>          Issue Type: Sub-task
>          Components: Table SQL / Planner
>    Affects Versions: 1.9.0, 1.10.0
>            Reporter: Zhenghua Gao
>            Assignee: Zhenghua Gao
>            Priority: Critical
>              Labels: pull-request-available
>             Fix For: 1.9.0
>
>          Time Spent: 20m
>  Remaining Estimate: 0h
>
> Currently many string builtin functions in blink planner follow hive/spark 
> semantics, which should keep compatible with old planner. And some 
> non-standard functions(Blink planner intros) should be removed.
>  * concat/concat_ws function (null treatment)
>  * substring function (follow calcite/flink)
>  * from_base64 should return string not binary
>  * intro truncate function to blink planner
>  * uuid should be no-argument (remove the one-argument version)
>  * length/jsonvalue/keyvalue/substr (non-standard function should be removed)
>  * md5/sha1/sha2/sha224/sha256/sha384/sha512(remove the two-arguments version)
>  * ascii (operand type should beSqlTypeFamily.CHARACTER)



--
This message was sent by Atlassian Jira
(v8.3.2#803003)

Reply via email to