On Mon, Mar 19, 2012 at 1:01 PM, Alvaro Herrera
<alvhe...@commandprompt.com> wrote:

> I'm rather of the contrary opinion -- surely if we're going to complete
> function names, we should only complete those that are in schemas in the
> path; similarly for column names.

I think it makes sense to only include currently-visible functions,
but *not* only columns from currently visible tables, since we won't
know yet whether the user intends to schema-qualify the table name.

>  (BTW I didn't check but does this
> completion work if I schema-qualify a column name?)

Peter's proposed tab-completion only kicks in for the column-name
itself. Keep in mind, the user might be trying to enter:
  SELECT  schema.table.column ...
  SELECT  table.column ...
  SELECT  table_alias.column ...
  SELECT  column ...

and presumably want to tab-complete the second token somehow. I'm a
bit leery about trying to tab-complete those first two, and the third
is right out. Just having the fourth would make me happy.


Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:

Reply via email to