Re: Filling Missing Primary Key Values
От | Steve Atkins |
---|---|
Тема | Re: Filling Missing Primary Key Values |
Дата | |
Msg-id | 232F46CB-CFB0-4991-AD70-4F7606D5B71D@blighty.com обсуждение исходный текст |
Ответ на | Re: Filling Missing Primary Key Values (Rich Shepard <rshepard@appl-ecosys.com>) |
Ответы |
Re: Filling Missing Primary Key Values
|
Список | pgsql-general |
On Aug 11, 2011, at 12:34 PM, Rich Shepard wrote: > On Thu, 11 Aug 2011, Chris Travers wrote: > >> The simplest seems to me to be a sequence and use nextval() to populate >> the null values. The major advantage would be that the sequence could stay >> around in case you need it again. So for example: >> >> create sequence my_varchar_values; > >> UPDATE my_table set my_varchar = >> nextval('my_varchar_values')::varchar(12) where my_varchar IS NULL; > > Chris, > > I was wondering if this was the best approach since I have new data to add > to the table. Don't need a starting value, eh? This will fail if any of the existing values are integers in the range that you're inserting - and it may fail in the future, as you add new records if they clash with existing entries. It's still a good way to go, but might need some care or some tweaking - adding a prefix, maybe. Cheers, Steve
В списке pgsql-general по дате отправления: