Re: Improve readability by using designated initializers when possible - Mailing list pgsql-hackers

From Peter Eisentraut
Subject Re: Improve readability by using designated initializers when possible
Date
Msg-id 26345df1-24d6-4691-824d-a70641fb5e62@eisentraut.org
Whole thread Raw
In response to Re: Improve readability by using designated initializers when possible  (jian he <jian.universality@gmail.com>)
Responses Re: Improve readability by using designated initializers when possible
List pgsql-hackers
On 18.03.24 11:01, jian he wrote:
> select relname from pg_class where relisshared and relkind = 'r';
>          relname
> -----------------------
>   pg_authid
>   pg_subscription
>   pg_database
>   pg_db_role_setting
>   pg_tablespace
>   pg_auth_members
>   pg_shdepend
>   pg_shdescription
>   pg_replication_origin
>   pg_shseclabel
>   pg_parameter_acl
> (11 rows)
> 
> EventTriggerSupportsObject should return false for the following:
> SharedSecLabelRelationId
> SharedDescriptionRelationId
> DbRoleSettingRelationId
> SharedDependRelationId
> 
> but I am not sure ReplicationOriginRelationId.

EventTriggerSupportsObject() (currently named 
EventTriggerSupportsObjectClass()) is only used by the deletion code, 
and these additional classes are not supported there anyway.  Also, if 
they happen to show up there for some reason, then 
EventTriggerSQLDropAddObject() would error out in 
getObjectIdentityParts() or getObjectTypeDescription().  So you wouldn't 
get an event trigger firing on a previously unsupported class by 
accident.  So I think this is robust enough.



pgsql-hackers by date:

Previous
From: Daniel Gustafsson
Date:
Subject: Re: sslinfo extension - add notbefore and notafter timestamps
Next
From: Jakub Wartak
Date:
Subject: Re: pg_upgrade --copy-file-range