El 04/05/12 09:28, Bill House escribió:
> On 05/04/2012 07:13 AM, Ricardo Araoz wrote:
>>
>> I see. And you want the longest path for a given md5 is what you want.
>> Are the "path" fields unique for a given md5?
>> If they are, then you could use
>>
>> select *
>> from theTable TT
>> inner join  (select md5, max(path) as 'path'
>>                   from theTable
>>                   group by md5
>>                   ) lastOnes
>>     on lastOnes.md5 = TT.md5
>>         and lastOnes.path = TT.path
>>
>> to get the subset you want.
>> If they are not, then how do you decide between two tuples with the same
>> md5-path?
> If the md5sum is identical, the file is identical and I want to keep
> only one instance of the file.  It does not matter which one.  File
> names may be changed later anyway.
>

All right then! Then the select will give you the set you want to keep
without further hassle.
If you want to delete the duplicates then you do a delete based on this
select (NOT in this select actually).
No need for any other complication, just one plain sql sentence sent to
the server.


_______________________________________________
Post Messages to: [email protected]
Subscription Maintenance: http://leafe.com/mailman/listinfo/dabo-users
Searchable Archives: http://leafe.com/archives/search/dabo-users
This message: http://leafe.com/archives/byMID/[email protected]

Reply via email to