Daniel

I can look at it, but unfortunatelly I have no spare time now and in near future. Feel free to submit pull request if you manage to fix it.

This `JOIN ndx_cols ON (a.attnum = ndx_cols.col_ndx)` is used to get the columns on which the index is defined. The expression index has col_ndx = 0, which filters out everything...

Following query works for me: ```sql WITH ndx_list AS ( SELECT pg_index.indexrelid, pg_class.oid FROM pg_index, pg_class WHERE pg_class.relname = 'a' AND pg_class.oid = pg_index.indrelid ), ndx_cols AS ( SELECT pg_class.relname,...

Good question. I tried briefly, but I just get a definition of whole index (CREATE INDEX...). However DBeaver somehow manages to get it, so I'll look into it more. ...

```sql WITH ndx_list AS ( SELECT pg_index.indexrelid, pg_class.oid FROM pg_index, pg_class WHERE pg_class.relname = 'a' AND pg_class.oid = pg_index.indrelid ) SELECT pg_class.relname AS constraint_name, CASE i.indisprimary WHEN TRUE THEN 'PRIMARY'...

Same query as above, but using ROW_NUMBER instead of WITH ORDINALITY: ```sql WITH ndx_list AS ( SELECT pg_index.indexrelid, pg_class.oid FROM pg_index, pg_class WHERE pg_class.relname = 'a' AND pg_class.oid = pg_index.indrelid...