Обсуждение: inconsitent tab completion in psql

Disclaimer that this was found under RC1 so it might have been fixed in final
release, though I don't recall it being reported, anyway, the problem is that
tab completion does not seem to work for "\h alter operator class".

template1=# \h ALTER
AGGREGATE   DOMAIN      INDEX       SCHEMA      TABLESPACE  USER
CONVERSION  FUNCTION    LANGUAGE    SEQUENCE    TRIGGER
DATABASE    GROUP       OPERATOR    TABLE       TYPE
template1=# \h ALTER OPERATOR

template1=# \h ALTER OPERATOR CL

template1=# \h ALTER OPERATOR CLASS
Command:     ALTER OPERATOR CLASS
Description: change the definition of an operator class
Syntax:
ALTER OPERATOR CLASS name USING index_method RENAME TO newname
ALTER OPERATOR CLASS name USING index_method OWNER TO newowner

template1=#


I noticed another other issues too, like CREATE CONSTRAINT TRIGGER doesnt show
up when you do \h CREATE<tab><tab> and when you do \h CREATE CON<tab><tab>
you get CREATE CONVERSION instead.

--
Robert Treat
Build A Brighter Lamp :: Linux Apache {middleware} PostgreSQL

Robert Treat wrote:
> Disclaimer that this was found under RC1 so it might have been fixed in final
> release, though I don't recall it being reported, anyway, the problem is that
> tab completion does not seem to work for "\h alter operator class".
>
> template1=# \h ALTER
> AGGREGATE   DOMAIN      INDEX       SCHEMA      TABLESPACE  USER
> CONVERSION  FUNCTION    LANGUAGE    SEQUENCE    TRIGGER
> DATABASE    GROUP       OPERATOR    TABLE       TYPE
> template1=# \h ALTER OPERATOR
>
> template1=# \h ALTER OPERATOR CL
>
> template1=# \h ALTER OPERATOR CLASS
> Command:     ALTER OPERATOR CLASS
> Description: change the definition of an operator class
> Syntax:
> ALTER OPERATOR CLASS name USING index_method RENAME TO newname
> ALTER OPERATOR CLASS name USING index_method OWNER TO newowner
>
> template1=#


The reason this happens is because both ALTER OPERATOR and ALTER
OPERATOR CLASS are supported so for the third word you would have to do
tab complete only when the C-L-A-S-S is unique.  I suppose you could say
as soon as they type 'C' it can't match an operator name but it hardly
seems worth the complexity.

> I noticed another other issues too, like CREATE CONSTRAINT TRIGGER doesnt show
> up when you do \h CREATE<tab><tab> and when you do \h CREATE CON<tab><tab>
> you get CREATE CONVERSION instead.

CREATE CONSTRAINT TRIGGER is not supported because it is designed to be
used only by pg_dump and is not for general use.  I added a source file
comment mentioning why we don't support tab completion for it.

--
  Bruce Momjian                        |  http://candle.pha.pa.us
  pgman@candle.pha.pa.us               |  (610) 359-1001
  +  If your life is a hard drive,     |  13 Roberts Road
  +  Christ can be your backup.        |  Newtown Square, Pennsylvania 19073

On Sun, Jan 23, 2005 at 11:01:14AM -0500, Bruce Momjian wrote:

> The reason this happens is because both ALTER OPERATOR and ALTER
> OPERATOR CLASS are supported so for the third word you would have to do
> tab complete only when the C-L-A-S-S is unique.  I suppose you could say
> as soon as they type 'C' it can't match an operator name but it hardly
> seems worth the complexity.

Maybe you could add, to the query that extract operator names to
complete, a UNION clause with the constant CLASS.  So tab completion for
ALTER OPERATOR would show both the operators and the CLASS constant.
But beware of

ALTER OPERATOR ~<tab>

you'd have to add a LIKE clause to the CLASS arm of the union to prevent
that, I think.

--
Alvaro Herrera (<alvherre[@]dcc.uchile.cl>)
"Para tener más hay que desear menos"