Re: DatabaseMetaData.getImportedKeys returning duplicate

On Monday, March 24, 2003, at 09:27 PM, Dave Cramer wrote:
> Patch applied, to 7.2.3, and HEAD

thanks!!
-pete

> On Mon, 2003-03-17 at 10:53, Peter Royal wrote:
>> On Monday, March 17, 2003, at 04:07 AM, Dave Cramer wrote:
>>> I didn't see your first message ??
>>
>> I got a copy back when I sent it.. but it was the first message I had
>> sent since subscribing. o well.
>>
>>> Yes this is a valid fix.
>>
>> Thanks. Should I submit a patch for this, since it is a trivial
>> one-line change.
>> -pete
>>
>>>
>>> Dave
>>> On Sun, 2003-03-16 at 21:39, Peter Royal wrote:
>>>> Anyone? Is a specific testcase needed? Is no-one else using this
>>>> metadata?
>>>> thanks.
>>>> -pete
>>>>
>>>> On Monday, March 3, 2003, at 09:27 PM, Peter Royal wrote:
>>>>
>>>>> I am having a problem with the DatabaseMetaData.getImportedKeys()
>>>>> function. I am asking for the keys with respect to a specific
>>>>> table.
>>>>> All is fine, except when I have two columns in a table that are
>>>>> both
>>>>> foreign keys to another table, ie:
>>>>>
>>>>> Table A Table B
>>>>> ------- -------
>>>>> col1 -----------pk
>>>>> col2 -----------/
>>>>>
>>>>> The ResultSet returned from getImportedKeys is structured:
>>>>>
>>>>> col1
>>>>> col2
>>>>> col1
>>>>> col2
>>>>>
>>>>> If I execute the query that is generated behind the scenes in the
>>>>> psql
>>>>> command-line program, adding the DISTINCT keyword to the query
>>>>> causes
>>>>> it to return the proper number of rows. Is that a valid fix, or
>>>>> will
>>>>> that break something else? Should the duplicate-reduction be done
>>>>> inside the java code that creates the ResultSet? I have no problem
>>>>> making the fix, I just don't know enough about the jdbc internals
>>>>> (nor
>>>>> postgres itself, I'm a new user) to be confident enough to know
>>>>> exactly *what* to fix.
>>>>>
>>>>> Anyways great work thus far, I look forward to helping improve the
>>>>> driver!
>>>>> -pete