[CDBI] Working with tables without primary keys?

Left to My Own Devices nazareth at eye-of-newt.com
Tue Nov 8 03:12:31 GMT 2005


At 7:55 PM -0800 11/7/05, Mark Schoonover wrote:
>Left to My Own Devices <mailto:nazareth at eye-of-newt.com> tapped at Monday,
>November 07, 2005 6:49 PM:
>
>>>   > Given a table like this:
>>>>
>>>>   +--------+----------+
>>>>   | foo_id | foo_name |
>>>>   +--------+----------+
>>>>   |      0 | bar-1    |
>>>>   |      8 | bar-2    |
>>>>   |      3 | bar-3    |
>>>>   |      8 | bar-4    |
>>>>   |      7 | bar-5    |
>>>>   |      4 | bar-6    |
>>>>   |      7 | bar-7    |
>>>>   |      4 | bar-8    |
>>>>   |      0 | bar-9    |
>>>>   +--------+----------+
>>>
>>>  The primary key of that table seems to be (foo_id, foo_name).  At
>>>  least there are no dupes in this sample data.  Just set that as your
>>>  primary key.
>>
>>  foo_name by itself is unique in this sample. If it's guaranteed to be
>>  unique, you could use it as the key.
>
>It doesn't look unique to me, there's two eights, and two sevens, two fours,
>two zeros.... Three is the only unique number... You can go to three, but
>should not stop unless going to four, unless three is you final number, then
>you should throw the grenade...

You're looking at foo_id. I said that foo_name was unique.


-- 
Aneel Nazareth -- http://eye-of-newt.com/nazareth --




More information about the ClassDBI mailing list