On 06/18/2014 02:34 AM, Ian Barwick wrote:
> On 14/06/18 7:51, Andreas Karlsson wrote:
>> On 06/17/2014 01:36 PM, Ian Barwick wrote:
>>> One issue - the table's internal triggers will also be listed. which can
>>> result in
>>> something like this:
>>>
>>> This is a bit of an extreme case, but I don't think manually manipulating
>>> internal triggers (which can only be done as a superuser) is a common
>>> enough
>>> operation to justify their inclusion. I suggest adding
>>> 'AND tgisinternal is FALSE' to 'Query_for_trigger_of_table' to hide them.
>>
>> Good suggestion. I have attached a patch which filters out the internal triggers,
>> both for ALTER TABLE and DROP TRIGGER. I am not entirely sure about the DROP TRIGGER
>> case but I think I prefer no auto completion of RI triggers.
>
> Thanks, looks good. Another reason for not autocompleting RI triggers is that
> the names are all auto-generated; on the offchance you are manually manipulating
> them individually, you'd have to have a pretty good idea of which ones you're
> working with anyway.
Thanks, could you update the status of the patch in the commitfest app
to "Ready for Committer"?
> Personally I think this patch could go into 9.4, as it's not introducing any
> new features and doesn't depend on any 9.5 syntax.
I think the feature freeze is strict to avoid having to think about what
is an exception and what is not.
--
Andreas Karlsson
