具有重复参数的PostgreSQL函数

我在pg_catalog.pg_stat_get_activity中偶然发现了一个好奇的函数签名：

CREATE OR REPLACE FUNCTION pg_stat_get_activity(
    IN pid integer,OUT datid oid,OUT pid integer,-- more parameters...)
  RETURNS SetoF record AS 'pg_stat_get_activity'
  LANGUAGE internal STABLE
  COST 1
  ROWS 100;

此函数两次声明相同的参数名称,也从information_schema报告.

select 
  parameter_mode,parameter_name
from information_schema.parameters
where specific_schema = 'pg_catalog'
and specific_name like 'pg_stat_get_activity%'
order by ordinal_position

以上收益率(另见本SQLFiddle)：

+--------------+----------------+
|parameter_mode|parameter_name  |
+--------------+----------------+
|IN            |pid             |
|OUT           |datid           |
|OUT           |pid             |
|...           |...             |
+--------------+----------------+

天真地,我尝试创建一个类似的功能,但没有用：

CREATE FUNCTION f_2647(p1 IN int,p1 OUT int)
AS $$
BEGIN
    p1 := p1;
END;
$$LANGUAGE plpgsql;

我的问题：

>为什么内部pg_stat_get_activity函数重新声明两次相同的参数名称？这是为了什么目的？例如.为什么不直接使用INOUT参数？
>内部pg_stat_get_activity函数和我的函数有什么不同？为什么我不能使用这种语法？

我知道这些是相当学术性的问题,但我需要正确理解这一点,以便在jOOQ 代码生成器中修复issue.

我注意到它出现在9.2中.在9.1版中,out字段名为procpid：

 parameter_mode |  parameter_name  
----------------+------------------
 IN             | pid
 OUT            | datid
 OUT            | procpid
 OUT            | usesysid
 ...

寻找postgres git历史的变化导致这个提交：

commit 4f42b546fd87a80be30c53a0f2c897acb826ad52
Author: Magnus Hagander 
Date:   Thu Jan 19 14:19:20 2012 +0100

    Separate state from query string in pg_stat_activity

    This separates the state (running/idle/idleintransaction etc) into
    it's own field ("state"),and leaves the query field containing just
    query text.

    The query text will Now mean "current query" when a query is running
    and "last query" in other states. Accordingly,the field has been
    renamed from current_query to query.

    Since backwards compatibility was broken anyway to make that,the procpid
    field has also been renamed to pid - along with the same field in
    pg_stat_replication for consistency.

    Scott Mead and Magnus Hagander,review work from Greg Smith

在改变的行中,这是感兴趣的行：

-DATA(insert OID = 2022 (  pg_stat_get_activity         PGNSP PGUID 12 1 100 0 0 f f f f t s 1 0 2249 "23" "{23,26,23,25,16,1184,869,23}" "{i,o,o}" "{pid,datid,procpid,usesysid,application_name,current_query,waiting,xact_start,query_start,backend_start,client_addr,client_hostname,client_port}" _null_ pg_stat_get_activity _null_ _null_ _null_ ));
+DATA(insert OID = 2022 (  pg_stat_get_activity         PGNSP PGUID 12 1 100 0 0 f f f f t s 1 0 2249 "23" "{23,pid,state,query,state_change,client_port}" _null_ pg_stat_get_activity _null_ _null_ _null_ ));

在这种预先消化的形式中,作者没有注意到pid的双重使用,或者他们并不关心,因为它在实践中是无害的,这似乎是合理的.

它是通过的,因为这些内部函数是由initdb在跳过普通用户函数的创建检查的快速路径中创建的.

具有重复参数的PostgreSQL函数

相关文章