And another thing, can't I do this:
create table s.a (blah); create table s.b (blah);
create index myindex on s.a(blah); create index myindex on s.b(blah);
? When I drop them I have to specify the schema name, so presumably it tracks them that way. Why can't I have the same index name be on different tables?
David Rysdam wrote:
I have a script that automatically creates my database objects. In order to automatically create indexes, it needs to first make sure they don't exist.
For things like tables, this is easy:
select * from information_schema.tables where table_schema = "<myschema>" and table_name = "<tablename>"
But for indexes it is hard for some reason. There's a catalog table "pg_index", but it doesn't have index, schema or table names. I eventually found them in pg_class but the table and schema names aren't there.
After some searching around, I came across this very strange (to me, anyway) "::regclass" thing that let me do this:
select * from pg_catalog.pg_index where indexrelid = 'schema.index'::regclass
I'm not really clear what's that doing, but in any case it still isn't what I want. That query returns information when the index exists but errors out when the index doesn't exist. Is there a way I can get a non-erroring query on either condition that will tell me if an index exists on a given table in a given schema?
---------------------------(end of broadcast)--------------------------- TIP 3: if posting/reading through Usenet, please send an appropriate subscribe-nomail command to majordomo@xxxxxxxxxxxxxx so that your message can get through to the mailing list cleanly
---------------------------(end of broadcast)--------------------------- TIP 8: explain analyze is your friend
