Re: Strange bug...

From: "Dave Page" <dpage(at)vale-housing(dot)co(dot)uk>
To: "Donald Fraser" <demolish(at)cwgsy(dot)net>, <pgadmin-support(at)postgresql(dot)org>
Subject: Re: Strange bug...
Date: 2003-01-21 13:28:07
Message-ID: 03AF4E498C591348A42FC93DEA9661B88582@mail.vale-housing.co.uk
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgadmin-support

It's a bit bizarre I know, but it comes from some old code that had to
figure out if an index is system generated. The rules are:

1) Unique = True AND Name = TableName_FirstColName_key (i.e. created as
a unique constraint)
2) Primary = True (i.e. created as a primary key on a table).

In the second example you sent, you missed the s_ prefix from the name,
so rule 1 doesn't catch it as a system index. It's not a perfect
soilution, but until PostgreSQL 7.3 came along, it was the best I could
come up with. The changes to support 7.3 using a new better method are
extensive, so I will only be implementing them in pgAdmin III.

Regards, Dave.

-----Original Message-----
From: Donald Fraser [mailto:demolish(at)cwgsy(dot)net]
Sent: 21 January 2003 12:59
To: pgadmin-support(at)postgresql(dot)org
Subject: Re: [pgadmin-support] Strange bug...


Thanks for that Dave,
I don't mind it not displaying the constraint as an index but
the one thing I would ask for is consistency...

For example if I create the following table via SQL:

CREATE TABLE public.tbl_cmpycat (
id int4 NOT NULL,
s_desc text NOT NULL,
id_editedby int4,
dt_edited timestamp,
CONSTRAINT tbl_cmpycat_desc_key UNIQUE (s_desc),
CONSTRAINT tbl_cmpycat_pkey PRIMARY KEY (id)
) WITHOUT OIDS;

When I view it under pgAdminII it lists one index named
tbl_cmpycat_desc_key...?

This then leads on to another bug...
When I try to drop the above table (not the index) pgAdminII
attempts to drop the index first and then I get the following message as
recorded in the log.
21-01-2003 12:52:17 - SQL (Bugs): DROP INDEX
public.tbl_cmpycat_desc_key
21-01-2003 12:52:17 - Error in pgAdmin II:basActions.Drop:
-2147467259 - ERROR: Cannot drop index tbl_cmpycat_desc_key because
constraint tbl_cmpycat_desc_key on table tbl_cmpycat requires it You
may drop constraint tbl_cmpycat_desc_key on table tbl_cmpycat instead

Regards
Donald Fraser

----- Original Message -----
From: Dave Page <mailto:dpage(at)vale-housing(dot)co(dot)uk>
To: Donald Fraser <mailto:demolish(at)cwgsy(dot)net> ;
pgadmin-support(at)postgresql(dot)org
Sent: Tuesday, January 21, 2003 12:28 PM
Subject: RE: [pgadmin-support] Strange bug...


Hi Donald,

This is a feature, not a bug :-)

Seriously, pgAdmin figures out that the index is part of a
constraint and classes it as a system object, therefore hiding it. If
you switch on View System Objects on the View menu, you will see both
indexes under the table. My guess is that pg_dump is not quite so clever
and misses the UNIQUE constraint from the table definition, adding the
index manually instead.

Regards, Dave.

-----Original Message-----
From: Donald Fraser [mailto:demolish(at)cwgsy(dot)net]
Sent: 17 January 2003 20:05
To: pgadmin-support(at)postgresql(dot)org
Subject: [pgadmin-support] Strange bug...



OS W2K SP2
pgAdmin II 1.4.12
PostgreSQL 7.3 on i686-pc-linux-gnu, compiled by GCC
2.96

Another strange bug...

I create the following table using an SQL statement:

CREATE TABLE "tbl_usertype" (
"id" int4 NOT NULL,
"s_desc" varchar(60) NOT NULL,
CONSTRAINT "tbl_usertype_s_desc_key" UNIQUE ("s_desc"),
CONSTRAINT "tbl_usertype_pkey" PRIMARY KEY ("id")
) WITHOUT OIDS;
GRANT SELECT ON "tbl_usertype" TO PUBLIC;

pgAdmin then reports the following as the SQL
statements:

-- Table: public.tbl_usertype
CREATE TABLE public.tbl_usertype (
id int4 NOT NULL,
s_desc varchar(60) NOT NULL,
CONSTRAINT tbl_usertype_s_desc_key UNIQUE (s_desc),
CONSTRAINT tbl_usertype_pkey PRIMARY KEY (id)
) WITHOUT OIDS;
GRANT SELECT ON TABLE public.tbl_usertype TO PUBLIC;
GRANT ALL ON TABLE public.tbl_usertype TO postgres;

Now one would expect to see under pgAdmin one Index
named "tbl_usertype_s_desc_key".
pgAdmin reports zero Indexes....?
Again I checked the output from pg_dumpall and it
definitely exists.
pg_dump displays the following lines.
-- Name: tbl_usertype_s_desc_key; Type: INDEX; Schema:
public; Owner: postgres
CREATE UNIQUE INDEX tbl_usertype_s_desc_key ON
tbl_usertype USING btree (s_desc);

Regards
Donald Fraser.

Browse pgadmin-support by date

  From Date Subject
Next Message Dave Page 2003-01-21 13:38:57 Re: Anomolous schemas
Previous Message Donald Fraser 2003-01-21 12:58:37 Re: Strange bug...