PostgreSQL 函数/触发器为每次调用的每一行分配相同的 ID

Question

I want to insert into a child table every time the parent table is updated. But when this happens, all of the new records inserted into the child table should have the same id. The ids will only increment if the parent table is separately updated another time. How can I do this?

In this case, I want to insert into the child table every new My_Date field from when the parent table is updated. Below is an example of what this would look like.

When parent table gains two new rows...

My_Date
old
old
new
new

Child table gains two new rows, both assigned to same ID (the ID autoincrements in table definition)

My_Date ID
...
new     4
new     4

When parent table gains two new rows again...

My_Date
old
old
old
old
new
new

Child table gains two new rows, both assigned to same new ID

My_Date ID
...
old     4  
old     4
new     5
new     5

Here is what I have so far.

CREATE or replace FUNCTION update_child() RETURNS trigger AS
  $BODY$
BEGIN
  INSERT INTO child
  SET My_Date = NEW.My_Date /*Not sure if this is correct*/
  /*Give every row the same ID*/
  RETURN NEW;
END
$BODY$
LANGUAGE plpgsql;

CREATE TRIGGER update_child_after_update
AFTER UPDATE
ON parent
FOR EACH ROW
EXECUTE PROCEDURE update_child(); 

Answer 1

I believe what you want is a combination of a sequence and a statement-level (as opposed to row level) trigger:

https://www.postgresql.org/docs/current/trigger-definition.html#:~:text=PostgreSQL%20offers%20both%20per%2Drow,statement%20that%20fired%20the%20trigger .

PostgreSQL offers both per-row triggers and per-statement triggers. With a per-row trigger, the trigger function is invoked once for each row that is affected by the statement that fired the trigger. In contrast, a per-statement trigger is invoked only once when an appropriate statement is executed, regardless of the number of rows affected by that statement. In particular, a statement that affects zero rows will still result in the execution of any applicable per-statement triggers. These two types of triggers are sometimes called row-level triggers and statement-level triggers, respectively.

This is really bare bones, but I think it will demonstrate the desired behavior you described in your question.

create sequence child_id;

CREATE or replace FUNCTION update_child()
RETURNS trigger AS
$BODY$
DECLARE
  r1 record;
  new_id int;
BEGIN
  new_id := nextval ('child_id');
  FOR r1 IN SELECT * FROM new_table
  LOOP
    insert into child
    select r1.my_date, new_id;
  END loop;
  RETURN NEW;
END
$BODY$
LANGUAGE plpgsql;

CREATE TRIGGER update_child_after_update
AFTER UPDATE
ON parent
REFERENCING OLD TABLE AS old_table NEW TABLE AS new_table
FOR EACH statement
EXECUTE PROCEDURE update_child(); 

CREATE TRIGGER update_child_after_insert
AFTER INSERT
ON parent
REFERENCING NEW TABLE AS new_table
FOR EACH statement
EXECUTE PROCEDURE update_child(); 

There are a lot of restrictions on statement-level triggers, and it's worth reading up on them. For example, only the "after" is supported.

Presumably your parent table also has some form of PK, which you would also be passing to the child, but I'm hopeful that's easy enough to see where you would insert that in the code example above.