Re: psql's describe command (for sequences) output improvement

Hi all,

I have a patch which tries to improve the '\d some_sequence_name' command output in psql utility. Before sending the patch to pgsql-patches I just want to know your opinion whether the new output of the command is OK or it needs to be modified before sending it to pgsql-patches.

For comparison purpose, here is the output of the '\d some_sequence_name' command:

-- Say we have created a sequence named 'test_seq' already as CREATE SEQUENCE test_seq.

--
-- Output: before applying the patch.
--
testdb=# \d test_seq;
Sequence "public.test_seq"
Column | Type
---------------+---------
sequence_name | name
last_value | bigint
increment_by | bigint
max_value | bigint
min_value | bigint
cache_value | bigint
log_cnt | bigint
is_cycled | boolean
is_called | boolean

--
-- Output: after applying the patch.
--
testdb=# \d test_seq;
Sequence "public.test_seq"
last_value | increment_by | max_value | min_value | cache_value | log_cnt | is_cycled | is_called
------------+--------------+---------------------+-----------+-------------+---------+-----------+-----------
1 | 1 | 9223372036854775807 | 1 | 1 | 1 | f | f

Thanks in advance for your feedback(s).

--
Aftab.

On 04/12/2007, Aftab Hussain <aftab(dot)hussain(at)enterprisedb(dot)com> wrote:
>
>
> Hi all,
>
> I have a patch which tries to improve the '\d some_sequence_name' command
> output in psql utility. Before sending the patch to pgsql-patches I just
> want to know your opinion whether the new output of the command is OK or it
> needs to be modified before sending it to pgsql-patches.
>
> For comparison purpose, here is the output of the '\d some_sequence_name'
> command:
>
> -- Say we have created a sequence named 'test_seq' already as CREATE
> SEQUENCE test_seq.
>
> --
> -- Output: before applying the patch.
> --
> testdb=# \d test_seq;
> Sequence "public.test_seq"
> Column | Type
> ---------------+---------
> sequence_name | name
> last_value | bigint
> increment_by | bigint
> max_value | bigint
> min_value | bigint
> cache_value | bigint
> log_cnt | bigint
> is_cycled | boolean
> is_called | boolean
>
> --
> -- Output: after applying the patch.
> --
> testdb=# \d test_seq;
> Sequence "public.test_seq"
> last_value | increment_by | max_value | min_value | cache_value
> | log_cnt | is_cycled | is_called
>
> ------------+--------------+---------------------+-----------+-------------+---------+-----------+-----------
> 1 | 1 | 9223372036854775807 | 1 | 1
> | 1 | f | f
>
>
>
> Thanks in advance for your feedback(s).
>
> --
> Aftab.
>
Why?

is that not what

select * from test_seq;

does already.

\d command return meta data not data.

Peter Childs

On Dec 4, 2007 7:47 PM, Peter Childs <peterachilds(at)gmail(dot)com> wrote:

>
>
> On 04/12/2007, Aftab Hussain <aftab(dot)hussain(at)enterprisedb(dot)com> wrote:
> >
> >
> > Hi all,
> >
> > I have a patch which tries to improve the '\d some_sequence_name'
> > command output in psql utility. Before sending the patch to pgsql-patches I
> > just want to know your opinion whether the new output of the command is OK
> > or it needs to be modified before sending it to pgsql-patches.
> >
> > For comparison purpose, here is the output of the '\d
> > some_sequence_name' command:
> >
> > -- Say we have created a sequence named 'test_seq' already as CREATE
> > SEQUENCE test_seq.
> >
> > --
> > -- Output: before applying the patch.
> > --
> > testdb=# \d test_seq;
> > Sequence "public.test_seq"
> > Column | Type
> > ---------------+---------
> > sequence_name | name
> > last_value | bigint
> > increment_by | bigint
> > max_value | bigint
> > min_value | bigint
> > cache_value | bigint
> > log_cnt | bigint
> > is_cycled | boolean
> > is_called | boolean
> >
> > --
> > -- Output: after applying the patch.
> > --
> > testdb=# \d test_seq;
> > Sequence "public.test_seq"
> > last_value | increment_by | max_value | min_value |
> > cache_value | log_cnt | is_cycled | is_called
> >
> > ------------+--------------+---------------------+-----------+-------------+---------+-----------+-----------
> > 1 | 1 | 9223372036854775807 | 1 |
> > 1 | 1 | f | f
> >
> >
> >
> > Thanks in advance for your feedback(s).
> >
> > --
> > Aftab.
> >
> Why?
>
> is that not what
>
> select * from test_seq;
>
> does already.

>
> \d command return meta data not data.
>
> Peter Childs
>

In general, \d command is working perfectly for database objects.

For sequences, I think the current \d some_sequence command's output is
displaying information which does not help the end user very much. Also
isn't the newly display information (same as information provided by 'select
* from test_seq;' statement) is/can-be-considered the metadata information
about the sequences queried about (since for the returning sequences data we
have nextval('...'), currval('...') functions)?

Please correct me if I am wrong or have missed something.

Aftab Hussain.
EnterpriseDB. www.enterprisedb.com

Aftab Hussain wrote:
> In general, \d command is working perfectly for database objects.
>
> For sequences, I think the current \d some_sequence command's output is
> displaying information which does not help the end user very much. Also
> isn't the newly display information (same as information provided by 'select
> * from test_seq;' statement) is/can-be-considered the metadata information
> about the sequences queried about (since for the returning sequences data we
> have nextval('...'), currval('...') functions)?

Yes, you are kind of right that \d on sequences provides unhelpful
output, but having it display the sequence values seems odd. TODO has:

o Have psql show current values for a sequence

Maybe \d+ sequence_name should add a column that shows the current
values:

test=> \d+ x
Sequence "public.x"
Column | Type | Value | Description
---------------+---------+----------------------
sequence_name | name | x
last_value | bigint | 1
increment_by | bigint |
max_value | bigint |
min_value | bigint |
cache_value | bigint |
log_cnt | bigint |
is_cycled | boolean |
is_called | boolean |

--
Bruce Momjian <bruce(at)momjian(dot)us> http://momjian.us
EnterpriseDB http://postgres.enterprisedb.com

+ If your life is a hard drive, Christ can be your backup. +