Re: Triggers with DO functionality
От | Tom Lane |
---|---|
Тема | Re: Triggers with DO functionality |
Дата | |
Msg-id | 4761.1329515190@sss.pgh.pa.us обсуждение исходный текст |
Ответ на | Re: Triggers with DO functionality (Andrew Dunstan <andrew@dunslane.net>) |
Ответы |
Re: Triggers with DO functionality
Re: Triggers with DO functionality Re: Triggers with DO functionality |
Список | pgsql-hackers |
Andrew Dunstan <andrew@dunslane.net> writes: > On 02/17/2012 03:58 PM, Thom Brown wrote: >> On 17 February 2012 20:40, Dimitri Fontaine<dimitri@2ndquadrant.fr> wrote: >>> I'm not so sure about other function properties (SET, COST, ROWS, >>> SECURITY DEFINER etc) because applying default and punting users to go >>> use the full CREATE FUNCTION syntax would be a practical answer here. > This is going to be pretty much a piece of syntactic sugar. Would it > matter that much if the trigger functions made thus are all volatile? If > someone wants the full function feature set they can always use CREATE > FUNCTION first. I think I'm with Dimitri - let's keep it simple. Volatility is a complete no-op for a trigger function anyway, as are other planner parameters such as cost/rows, because there is no planning involved in trigger calls. Of the existing CREATE FUNCTION options, I think only LANGUAGE, SECURITY DEFINER, and SET are of any possible interest for a trigger function. And I don't have any problem deeming SET a second-order thing that you should have to go use CREATE FUNCTION for. But perhaps SECURITY DEFINER is a common enough need to justify including in this shorthand form. Has anybody stopped to look at the SQL standard for this? In-line trigger definitions are actually what they intend, IIRC. regards, tom lane
В списке pgsql-hackers по дате отправления: