[ 
https://issues.apache.org/jira/browse/CASSGO-118?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18073391#comment-18073391
 ] 

Stefan Miklosovic commented on CASSGO-118:
------------------------------------------

I agree this is a bug and should be fixed. Once it is null then it should be 
null in the database too. 

I do not think there is somebody out there who is relying on the current buggy 
behavior, so I don't think we should enable legacy one. For them to depend on 
legacy behavior, how would that look like in practice? Their checking for 
nullity would need to explicitly go over each UDT field and check if it is 
null? That is hardly happening, imho.

I am not sure release-wise where this should be put. 

> Nil map for UDT column inserts zero-valued UDT instead of NULL
> --------------------------------------------------------------
>
>                 Key: CASSGO-118
>                 URL: https://issues.apache.org/jira/browse/CASSGO-118
>             Project: Apache Cassandra Go driver
>          Issue Type: Bug
>            Reporter: Bohdan Siryk
>            Priority: Normal
>          Time Spent: 20m
>  Remaining Estimate: 0h
>
> Inserting a row with a nil map for a UDT-typed column results in a non-NULL 
> value whose fields behave like defaults / zero values, instead of storing 
> NULL for the whole UDT.
> *Expected*
> The UDT column is NULL in DB.
> *Actual*
> The UDT column is present (not NULL) and fields appear as type defaults 
> (empty string, 0, false, etc.).
> To reproduce it, please use this code snippet on github  gist: 
> https://gist.github.com/worryg0d/22810cd6609cae985ca5c53725c3e3ae
> You will have the following in your C*:
> {code:java}
> cqlsh> SELECT * from gocqltest.persons ;
>  id | address                          | name
> ----+----------------------------------+---------
>   1 | {street: 'Street', zip: '12345'} | person1
>   2 |        {street: null, zip: null} | person2
> {code}



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to