Assign to NEW by key in a Postgres trigger

All we need is an easy explanation of the problem, so here it is.

In the trigger body, how can I assign a value to NEW by it’s field name?
That’s what I want to do:

some_key = "some_column";
NEW[some_key] = 5;

How to solve :

I know you bored from this bug, So we are here to help you! Take a deep breath and look at the explanation of your problem. We have many solutions to this problem, But we recommend you to use the first method because it is tested & true method that will 100% work for you.

Method 1

First of all, there is no "trigger body" (unlike Oracle). In Postgres you have a trigger function (also, misleadingly, called "procedure") with a function body and 0-n triggers (without body) calling this function.

The special variable NEW in plpgsql trigger functions is neither a map nor an array; it’s a record holding the new row:

NEW

Data type RECORD; variable holding the new database row for
INSERT/UPDATE operations in row-level triggers. This variable is
unassigned in statement-level triggers and for DELETE operations.

Assigning to a field (or column) of NEW is simple. The documented assignment operator is :=. (Since Postgres 9.4 also =.)

NEW.some_key := 5;

What you seem to be looking for is to parameterize the column name, which isn’t quite as simple.
The additional module hstore provides the #= operator. (It’s included in pretty much all standard distributions.) Install the module once per database with:

CREATE EXTENSION hstore;

Then you can:

NEW := NEW #= '"some_key"=>"5"'::hstore;

Sets the column some_key to ‘5’ – if the column exists.

  • An explicit cast to hstore is optional. The operator #= coerces a string literal to the right data type automatically.
  • hstore only stores text strings, so a given literal for the value may have to be cast twice – a very minor drawback compared to alternative solutions.
  • The given string literal has to fit the data type of the column, or an exception is raised.
  • If no column with the given name exists, nothing is changed, no exception raised.

Related answer with details and an alternative solution:

Code example

CREATE OR REPLACE FUNCTION trg_tbl_ins_bef()
  RETURNS trigger AS
$func$
BEGIN
   NEW := NEW #= '"some_key"=>"5"';
   RETURN NEW;
END
$func$ LANGUAGE plpgsql;

CREATE TRIGGER ins_bef
BEFORE INSERT ON tbl
FOR EACH ROW EXECUTE PROCEDURE trg_tbl_ins_bef();

Postgres 11 or later allows to replace the misleding term PROCEDURE (the old syntax will keep working for the foreseeable future):

...
FOR EACH ROW EXECUTE FUNCTION trg_tbl_ins_bef();

Method 2

I must admit that this is no easy way of solving it, but at least it’s a way. I created the below example as a standalone one, to avoid all clutter with trigger creation and such. If you use it in a trigger, you can remove the declaration and initialization of p and replace the remaining use with NEW.

DO $$
DECLARE p members_test; 
BEGIN
    p := (1,2,3);
    CREATE TEMP TABLE t ON COMMIT DROP AS SELECT p.*; -- a one row table holding 
                                                      -- the values of the variable

    EXECUTE format($e$UPDATE t SET %s = 43$e$, 'b'); -- this way you can access 
                                                     -- the columns dynamically

    SELECT * INTO p FROM t; -- assign the new values back to the variable
    RAISE INFO 'p: %', p;
END;
$$;

INFO:  p: (1,43,3)

Note: Use and implement method 1 because this method fully tested our system.
Thank you 🙂

All methods was sourced from stackoverflow.com or stackexchange.com, is licensed under cc by-sa 2.5, cc by-sa 3.0 and cc by-sa 4.0

Leave a Reply