PL/pgSQL Record Types

Summary: in this tutorial, you will learn about the PL/pgSQL record types, which enables you to define variables that can hold a single row from a result set.

Introduction to PL/pgSQL record types

PostgreSQL provides a “type” called the record that is similar to the row-type.

It’s important to note that a record isn’t a true type but rather a placeholder. Furthermore, the structure of a record variable will change when you reassign it to another value.

To declare a record variable, you simply use a variable name followed by the record keyword like this:

variable_name record;Code language: PostgreSQL SQL dialect and PL/pgSQL (pgsql)

A record variable is similar to a row-type variable, which can hold only one row of a result set.

Unlike a row-type variable, a record variable lacks a predefined structure. Instead, the structure of a record variable is determined when an actual row is assigned to it via the select or for statement.

To access a field in the record, you use the dot notation (.) syntax like this:

record_variable.field_name;Code language: PostgreSQL SQL dialect and PL/pgSQL (pgsql)

If you attempt to access a field in a record variable before it’s assigned, you’ll encounter an error.

PL/pgSQL record examples

Let’s take some examples of using the record variables.

1) Using record with the select into statement

The following example illustrates how to use the record variable with the select into statement:

do
$$
declare
	rec record;
begin
	-- select the film 
	select film_id, title, length 
	into rec
	from film
	where film_id = 200;
	
	raise notice '% % %', rec.film_id, rec.title, rec.length;   
	
end;
$$
language plpgsql;Code language: PostgreSQL SQL dialect and PL/pgSQL (pgsql)

How it works.

  • First, declare a record variable called rec in the declaration section.
  • Second use the select into statement to select a row whose film_id is 200 into the rec variable
  • Third, print out the information of the film via the record variable.

2) Using record variables in the for loop statement

The following shows how to use a record variable in a for loop statement:

do
$$
declare
	rec record;
begin
	for rec in select title, length 
			from film 
			where length > 50
			order by length
	loop
		raise notice '% (%)', rec.title, rec.length;	
	end loop;
end;
$$;Code language: PostgreSQL SQL dialect and PL/pgSQL (pgsql)

Here is the partial output:

NOTICE:  Hall Cassidy (51)
NOTICE:  Champion Flatliners (51)
NOTICE:  Deep Crusade (51)
NOTICE:  Simon North (51)
NOTICE:  English Bulworth (51)
...Code language: Shell Session (shell)

Note that you will learn more about the for loop statement in the for-loop tutorial.

How it works:

  • First, declare a variable named r with the type record.
  • Second, use the for loop statement to fetch rows from the film table (in the sample database). The for loop statement assigns the row that consists of title and length to the rec variable in each iteration.
  • Third, show the contents of the fields of the record variable by using the dot notation (rec.title and rec.length)

Summary

  • A record is a placeholder that can hold a single row of a result set.
  • A record does not have a predefined structure like a row variable. Its structure is determined when you assign a row to it.
Was this tutorial helpful ?