Postgresql Postresql:组合类型并选择到
有人能解释一下吗?(所有内容都在下面的示例中!)。我使用postgresql 9.4。 警告在下面的代码中,我删除了“object type”复合类型和“example_data”临时表,请确保数据库中没有这些表。。。以防万一,我对drop命令进行了注释,如果需要,取消注释它们Postgresql Postresql:组合类型并选择到,postgresql,postgresql-9.4,composite-types,Postgresql,Postgresql 9.4,Composite Types,有人能解释一下吗?(所有内容都在下面的示例中!)。我使用postgresql 9.4。 警告在下面的代码中,我删除了“object type”复合类型和“example_data”临时表,请确保数据库中没有这些表。。。以防万一,我对drop命令进行了注释,如果需要,取消注释它们 --drop type if exists object_type; --drop table if exists example_data; create type object_type as ( i
--drop type if exists object_type;
--drop table if exists example_data;
create type object_type as (
id text
,value text);
create temporary table example_data as (
select
'id1'::text as id,
'example value'::text as value);
do $$
declare
my_object object_type;
begin
-- The fiolowing doesn't work, we can't use INTO with the whole object
-- Some how, it tries to put the whole object into the first attribute:
select (id,value)::object_type
into my_object
from example_data;
raise warning 'ID: %, VALUE: %', my_object.id, my_object.value;
-- What a shame! It would have been so much more convenient than the following:
-- to feed the object we need to repeat each one of the attribute in the INTO section:
select id, value
into my_object.id, my_object.value
from example_data;
raise warning 'ID: %, VALUE: %', my_object.id, my_object.value;
/*
In that example it is not so bad, but when you have very large object, it is very ugly to repeat each one
of the attribute, for example:
select (att1, att2, att3, att4, att5, att6, att7, att8, att9, att10)::object_type2 into my_object2
and, (very heavy):
select att1, att2, att3, att4, att5, att6, att7, att8, att9, att10
into my_object2.att1, my_object2.att2, my_object2.att3, my_object2.att4, my_object2.att5, my_object2.att6,
my_object2.att7, my_object2.att8, my_object2.att9, my_object2.att10
*/
end $$;
--drop type if exists object_type;
--drop table if exists example_data;
跟随
...
select id, value
into my_object
from example_data;
...