db-derby-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Mamta A. Satoor (JIRA)" <j...@apache.org>
Subject [jira] Commented: (DERBY-2972) Update or select with function in the where clause causes with TERRITORY_BASED collation fails with ERROR 42818: Comparisons between 'VARCHAR' and 'VARCHAR' are not supported.
Date Wed, 05 Sep 2007 19:49:34 GMT

    [ https://issues.apache.org/jira/browse/DERBY-2972?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12525196
] 

Mamta A. Satoor commented on DERBY-2972:
----------------------------------------

Kathey, I can see where you suggestion is coming from when I look at the examples that you
provided. But I also feel that when something comes from a specific schema, which in these
examples is SYSCS_UTIL, it seems like it should get collation of that schema. This is similar
to character columns coming from system schema vs user schema. String literals do not belong
to any schema and hence it's seems reasonable to have them take the collation for the current
compilation schema. But for stored objects, like user-defined functions, columns belonging
to tables, it seems like they should take the collation of their schema.

I am on the border line on this issue and would be willing to go either way. I wonder if anyone
else in the community thinks strongly one way or the other.

> Update or select with function in the where clause causes with TERRITORY_BASED collation
fails with ERROR 42818: Comparisons between 'VARCHAR' and 'VARCHAR' are not supported.
> -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>
>                 Key: DERBY-2972
>                 URL: https://issues.apache.org/jira/browse/DERBY-2972
>             Project: Derby
>          Issue Type: Bug
>          Components: SQL
>    Affects Versions: 10.4.0.0
>            Reporter: Kathey Marsden
>            Assignee: Kathey Marsden
>         Attachments: derby-2972_diff.txt, derby-2972_stat.txt
>
>
> The following update fails with ERROR 42818
> ij> update testing set a = PADSTRING('aa',2024) where a = PADSTRING('a',2024);
> ERROR 42818: Comparisons between 'VARCHAR' and 'VARCHAR' are not supported.
> See full script below 
> onnect 'jdbc:derby:nordb;territory=no_NO;collation=TERRITORY_BASED';
> CREATE FUNCTION  PADSTRING (DATA VARCHAR(32000), LENGTH INTEGER) RETURNS VARCHAR(32000)
EXTERNAL NAME 'org.apache.derbyTesting.functionTests.util.Formatters.padString' LANGUAGE JAVA
PARAMETER STYLE JAVA;
> create table testing 
> 	(a varchar(2024), b varchar(1024), c varchar(1024), d varchar(2048), e varchar(300))
;
> -- insert 9 rows into the table
> insert into testing values (PADSTRING('1',2024),  PADSTRING('2',1024), 
>        PADSTRING('3',1024), PADSTRING('4',2048),  PADSTRING('5',300));
> insert into testing values (PADSTRING('10',2024),  
>        PADSTRING('20',1024), PADSTRING('30',1024), 
>        PADSTRING('40',2048), PADSTRING('50',300));
> insert into testing values (PADSTRING('100',2024),  
>        PADSTRING('200',1024), PADSTRING('300',1024), 
>        PADSTRING('400',2048), PADSTRING('500',300));
> insert into testing values (PADSTRING('1000',2024),  
>        PADSTRING('2000',1024), PADSTRING('3000',1024), 
>        PADSTRING('4000',2048), PADSTRING('5000',300));
> insert into testing values (PADSTRING('10000',2024),  
>        PADSTRING('20000',1024),	PADSTRING('30000',1024), 
>        PADSTRING('40000',2048), PADSTRING('50000',300));
> insert into testing values (PADSTRING('100000',2024), 
>        PADSTRING('200000',1024), PADSTRING('300000',1024), 
>        PADSTRING('400000',2048), PADSTRING('500000',300));
> insert into testing values (PADSTRING('1000000',2024), 
>        PADSTRING('2000000',1024), PADSTRING('3000000',1024), 
>        PADSTRING('4000000',2048), PADSTRING('5000000',300));
> insert into testing values (PADSTRING('10000000',2024), 
>        PADSTRING('20000000',1024), PADSTRING('30000000',1024), 
>        PADSTRING('40000000',2048), PADSTRING('50000000',300));
> insert into testing values (PADSTRING('100000000',2024), 
>        PADSTRING('200000000',1024), PADSTRING('300000000',1024), 
>        PADSTRING('400000000',2048), PADSTRING('500000000',300));
> update testing set a = PADSTRING('aa',2024) where a = PADSTRING('a',2024);

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message