Re: Query to get the "next available" unique suffix for a name

On Tue, Sep 28, 2010 at 12:44 AM, Arjen Nienhuis <a.g.nienhuis@gmail.com> wrote:
>> 4) Do a SELECT on each row that starts with "MikeChristensen" and then
>> trying to append the row count to the end, this might not be exact but
>> it's somewhat intelligent as a starting point.  However, this might
>> require some special indexes on this table to quickly scan rows that
>> start with a certain name.  However, if I get to the point where this
>> becomes slow then I can worry about it at that point since this would
>> only be run on new account creation and not super speed critical.
>>
>
> CREATE TABLE test (a text PRIMARY KEY);
>
> INSERT INTO test (a)
> SELECT * FROM (SELECT 'MikeChristensen' || generate_series AS t FROM
> generate_series(1,10000)) x
> WHERE t NOT in (SELECT a FROM test) LIMIT 1 RETURNING *;

Now that's an interesting approach, I will play around with that - thanks!!