db-derby-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Shreyas Kaushik <Shreyas.Kaus...@Sun.COM>
Subject Re: [jira] Commented: (DERBY-18) Exposed name matching has bugs when the column name is qualified with a schema name.
Date Wed, 09 Feb 2005 11:12:31 GMT
Dan,

Please see below.

~ Shreyas

Daniel John Debrunner wrote:

>><>Shreyas Kaushik wrote:
>>
>>Attached is the new patch and answesr are inline.
>>
>>
>> 
>>
>>    
>>
>Daniel John Debrunner wrote:
>
>  
>
>>I got the patch to apply and am running tests on it.
>>
>> 
>>
>>    
>>
>9 (possibly with one unrelated failure) tests fail or have changed
>output with this patch. Most of the differences in the output seem to be
>a change in the reporting of the failing column to include the implicit
>schema name, e.g. T1.C1 is replaced by APP.T1.C1 in an error message.
>However, there are a few troublesome diffs. The dml162 test from the
>nist suite fails to create a view that previously was created ok. This
>dml162 issue would need to be resolved before the patch could be accepted.
>  
>
With the changes the fully qualified column name should be given or it 
would take the default schema and the
default schema does not have what the test is looking for. Since 
everything is prefixed with "HU." so should the
column in question.  With that change it works fine. Sending a patch for 
this also.

>In addition there are no additional, new, test cases that show the
>change in behaviour to demostrate that Derby-18 is indeed fixed. Test
>cases can be added to existing tests.
>  
>
Sending a patch having new tests ( 2 tests ) that check the fix.

>Again, like the identifier change, if a patch from a contibutor causes
>the tests to "fail" due to modified but correct output, who should be
>fixing up the test masters? I'd assume the contributor.
>  
>
Sending the patch for the master files as well.

>Dan.
>
>  
>

Mime
View raw message