Re: Conditional INSERT - Mailing list pgsql-general

From Ken Tanzer
Subject Re: Conditional INSERT
Date
Msg-id CAD3a31U5YB8vd4CyKWrbuZoqgiE58SCdOUck6ju9b_yvrZ9Daw@mail.gmail.com
Whole thread Raw
In response to Re: Conditional INSERT  (Adrian Klaver <adrian.klaver@aklaver.com>)
Responses Re: Conditional INSERT
List pgsql-general
On Fri, Mar 15, 2019 at 11:59 AM Adrian Klaver <adrian.klaver@aklaver.com> wrote:
On 3/15/19 11:54 AM, basti wrote:
> this is a dns database, and the client is update the _acme-challenge for
> LE certificates. I don't want that the client can insert "any" txt record.
> the client should only insert data if the hostname start with
> _acme-challenge. i have no control on client.
>
> i have try this rule but the server reject this with a endless loop:

To borrow a quote:

"I had a problem so I decided to use a rule, now I have two problems."

Do not use a rule. As suggested upstream use a BEFORE INSERT trigger,
you will be a lot happier.

>
> CREATE RULE insert_acme AS ON INSERT TO t_dnsadmin_records_txt
>      WHERE NEW.hostname like '_acme-challenge%'
>      DO INSERT INTO t_dnsadmin_records_txt VALUES (
>                                      NEW.domainid,
>                                      NEW.hostname,
>                                      NEW.txtdata
>                                  );
>
>

Just curious, but wanted to follow up on whether rules are across-the-board discouraged?  I've seen disparaging comments about them, but I don't see any indication of that on the create rule page.

The other suggestion in this thread--a foreign key--will throw an error.  Your suggestion of a before trigger might well be better (and if so, why?), but is there anything particularly wrong or bad about using a rule that would actually work?  Something along these lines:  

CREATE RULE insert_acme AS ON INSERT TO t_dnsadmin_records_txt
WHERE NOT NEW.hostname like '_acme-challenge%'
DO INSTEAD NOTHING;

Thanks,
Ken



-- 
AGENCY Software  
A Free Software data system
By and for non-profits
(253) 245-3801

learn more about AGENCY or
follow the discussion.

pgsql-general by date:

Previous
From: Rob Sargent
Date:
Subject: Re: Camel case identifiers and folding
Next
From: Morris de Oryx
Date:
Subject: Re: Camel case identifiers and folding