Обсуждение: psql's \d functions broken for views in current sources

psql has some problems with views in current CVS: \d doesn't show views,
and if you do \d on a specific view, it doesn't identify it as a view
and doesn't show the view definition rule.

I assume this breakage is from the recent RELKIND_VIEW change;
probably psql didn't get updated to know about the new relkind.

Anyone care to work on this?
        regards, tom lane

Tom Lane wrote:

> psql has some problems with views in current CVS: \d doesn't show views,
> and if you do \d on a specific view, it doesn't identify it as a view
> and doesn't show the view definition rule.
>
> I assume this breakage is from the recent RELKIND_VIEW change;
> probably psql didn't get updated to know about the new relkind.
>

Probably psql uses pg_views though I don't remember correctly.
It seemd that pg_views(initdb) should be changed first.

Regards.

Hiroshi Inoue

Hiroshi Inoue <Inoue@tpf.co.jp> writes:
> Tom Lane wrote:
>> I assume this breakage is from the recent RELKIND_VIEW change;
>> probably psql didn't get updated to know about the new relkind.

> Probably psql uses pg_views though I don't remember correctly.
> It seemd that pg_views(initdb) should be changed first.

No, pg_views still works --- although it could be made far more
efficient (don't need the WHERE EXISTS(...) test anymore, just look
at relkind).  So I don't think that explains why psql is misbehaving.

You are right that we ought to change the definition of pg_views,
anyway.
        regards, tom lane

Was this addressed?


> Hiroshi Inoue <Inoue@tpf.co.jp> writes:
> > Tom Lane wrote:
> >> I assume this breakage is from the recent RELKIND_VIEW change;
> >> probably psql didn't get updated to know about the new relkind.
> 
> > Probably psql uses pg_views though I don't remember correctly.
> > It seemd that pg_views(initdb) should be changed first.
> 
> No, pg_views still works --- although it could be made far more
> efficient (don't need the WHERE EXISTS(...) test anymore, just look
> at relkind).  So I don't think that explains why psql is misbehaving.
> 
> You are right that we ought to change the definition of pg_views,
> anyway.
> 
>             regards, tom lane
> 


--  Bruce Momjian                        |  http://candle.pha.pa.us pgman@candle.pha.pa.us               |  (610)
853-3000+  If your life is a hard drive,     |  830 Blythe Avenue +  Christ can be your backup.        |  Drexel Hill,
Pennsylvania19026