Content-Type: text/plain; charset=us-ascii
Mime-Version: 1.0 (Mac OS X Mail 8.2 \(2104\))
Subject: Re: PATCH: Added Node Type & Catalog objects [pgAdmin4]
From: Murtuza Zabuawala <murtuza.zabuawala@enterprisedb.com>
In-Reply-To: 
 <CA+OCxoxyCifnk7KT9ZZGBNYtH6gjCoJ7j8tUwSoBxzgAtCPZFA@mail.gmail.com>
Date: Thu, 7 Apr 2016 17:11:58 +0530
Cc: pgadmin-hackers <pgadmin-hackers@postgresql.org>
Content-Transfer-Encoding: quoted-printable
Message-Id: <101BFDF3-B2FF-45A8-8990-20FD363613F6@enterprisedb.com>
References: 
 <CAKKotZSda6+wgyzckDJPXKK3dc8U-UXho9hmaaBp+LwZzdnhaQ@mail.gmail.com>
 <CA+OCxoxNx3ghQZ+YnWeNpyVs_BaJE+yL7C6H-PUFh6ddpr061Q@mail.gmail.com>
 <CA+OCxozEyaxygwg4KKNWKjKcLw=a-s6R64Tij9Z2JSVzzEKLrg@mail.gmail.com>
 <CAKKotZTGn2aUnkrotK5u=9hh+nD4gUC9mvKWo2FcDVuEhH5oyA@mail.gmail.com>
 <CA+OCxoyeSbbyKFm236agok7DNr4iRt16On9Kea+Z3tUj_iXa_g@mail.gmail.com>
 <CAKKotZQVJPRkHKzhix+s9zc+QnwUfWfOj+sVFfk6c7abKuyECg@mail.gmail.com>
 <CA+OCxowYVt2V-sr5YQm-YDxhmNiam-4iSudNSmydu_e2_T3eMA@mail.gmail.com>
 <CAKKotZSrb=YrxbaasHQF9DXA-roFmzmA7cVC3tVPMdhYwL58cw@mail.gmail.com>
 <CA+OCxoz_rEVOMa1nfJxH16-ZB6YXozm-DAAV8HOE5EORqAwQLQ@mail.gmail.com>
 <CAKKotZTFTsZ2fHwMQem6v9prw8Un_qa8SqRmSanvU5eMXmTW8A@mail.gmail.com>
 <CA+OCxowAevPwtNb1O6Shx1NndTjNJYPBd_ie-xnk4Hw4j-nKqA@mail.gmail.com>
 <CA+OCxoxyCifnk7KT9ZZGBNYtH6gjCoJ7j8tUwSoBxzgAtCPZFA@mail.gmail.com>
To: Dave Page <dpage@pgadmin.org>
Precedence: bulk
Sender: pgadmin-hackers-owner@postgresql.org

Thanks Dave.

Sure will look in it ASAP.

Regards,
Murtuza

> On 07-Apr-2016, at 5:06 pm, Dave Page <dpage@pgadmin.org> wrote:
>=20
> Hi Murtuza
>=20
> On Tue, Mar 22, 2016 at 9:22 AM, Dave Page <dpage@pgadmin.org> wrote:
>> Hi
>>=20
>> On Tue, Mar 22, 2016 at 8:14 AM, Murtuza Zabuawala
>> <murtuza.zabuawala@enterprisedb.com> wrote:
>>> Hi Dave,
>>>=20
>>> We can create new external type using below method, By running all of b=
elow
>>> queries at the same time , we can not create separate external type by =
only
>>> using create type statement.
>>>=20
>>> So as per my discussion with Ashesh, We should not allow user to create
>>> external type in pgAdmin4 but only show definition in edit mode.
>>=20
>> Hmm, would it not make sense to allow the user to create the shell
>> type as well (perhaps, with a new type of "SHELL")? Then they could do
>> what is needed (and that should be easy, as it's just CREATE TYPE
>> foo;)
>>=20
>> For example:
>>=20
>> CREATE TYPE box;
>>=20
>> CREATE FUNCTION my_box_in_function(cstring) RETURNS box AS ... ;
>> CREATE FUNCTION my_box_out_function(box) RETURNS cstring AS ... ;
>>=20
>> CREATE TYPE box (
>>    INTERNALLENGTH =3D 16,
>>    INPUT =3D my_box_in_function,
>>    OUTPUT =3D my_box_out_function
>> );
>>=20
>> CREATE TABLE myboxes (
>>    id integer,
>>    description box
>> );
>=20
> In the interests of making progress, I've committed the most recent
> patch, with a number of minor changes most significantly, the Postgres
> docs and system catalogs seem to have different ideas about what to
> call length, precision and scale. pgAdmin 3 followed the catalogs and
> used length and precision, however I've updated pgAdmin 4 to use
> "Length/precision" and "Scale" which is inline with the Postgres docs.
> That's only in the UI though - the code follows the catalogs.
>=20
> There are still a couple of issues - please provide fixes ASAP:
>=20
> 1) If you create a composite type that contains a sized type (e.g.
> numeric(5, 4), the precision and scale are not shown if you later open
> the properties dialogue, or in the reverse engineered SQL.
>=20
> E.g. what pgAdmin3 shows as:
>=20
> CREATE TYPE pem.blergh AS
>   (c1 text COLLATE pg_catalog."C",
>    c2 numeric(5));
>=20
> Is shown by pgAdmin4 as:
>=20
> CREATE TYPE pem.blergh AS
>    (c1 text COLLATE pg_catalog."C", c2 numeric);
>=20
> (adding the \n's would be good too).
>=20
> 2) If you select a different type of type in create mode, the new
> options are shown below those for the previously selected type,
> instead of replacing them. Please see the attached screenshot.
>=20
> 3) I would still like us to support External types. I believe the
> simple option here is to re-add the code you had previously, and to
> add a new type of type called "SHELL" as discussed in my previous
> email above. The user would then be able to create a SHELL type, add
> the required functions, then come back and create the EXTERNAL type.
>=20
> I'll add cards to our internal kanban chart for these issues.
>=20
> Thanks.
>=20
> --=20
> Dave Page
> Blog: http://pgsnake.blogspot.com
> Twitter: @pgsnake
>=20
> EnterpriseDB UK: http://www.enterprisedb.com
> The Enterprise PostgreSQL Company
> <Screen Shot 2016-04-07 at 12.33.11.png>


--=20
Sent via pgadmin-hackers mailing list (pgadmin-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgadmin-hackers