On Thu, 30 Jul 2020 at 12:24, Fujii Masao <masao.fu...@oss.nttdata.com> wrote: > > > > On 2020/07/30 10:46, Michael Paquier wrote: > > On Thu, Jul 30, 2020 at 08:44:26AM +0900, Fujii Masao wrote: > >> Isn't it better to add the comment explaining why toast tables are > >> excluded from the tab completion for vacuum while they are vacuumable? > > > > Sounds sensible, still it does not apply only to vacuum. I would go > > as far as just adding a comment at the beginning of the block for > > schema queries: > > Yes, that seems better.
Agreed. > BTW, one thing I think a bit strange is that indexes for toast tables > are included in tab-completion for REINDEX, for example. That is, > "REINDEX INDEX<tab>" displays "pg_toast.", and "REINDEX INDEX pg_toast.<tab>" > displays indexes for toast tables. Maybe it's better to exclude them, > too. But there seems no simple way to do that. > So I'm ok with this current situation. Yeah, that's the reason why I mentioned about toast tables. "VACUUM <tab>" displays "pg_toast." but, "VACUUM pg_to<tab>" doesn't complement anything. Regards, -- Masahiko Sawada http://www.2ndQuadrant.com/ PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
