Assignment Statements

Variable Syntax

Figure 1 shows the syntax diagram for assigning a value to a variable.

Figure 1 assignment_value::=
Click to enlarge

The syntax is described as follows:

variable_name: specifies the name of a variable.
value: can be a value or an expression. The type of value must be compatible with the type of variable_name.

Example:

    
         DECLARE
    emp_id  INTEGER := 7788; -- Assign a value.
BEGIN
    emp_id := 5; -- Assign a value.
    DBE_OUTPUT.PRINT_LINE(emp_id);
    emp_id := 5*7784;
    DBE_OUTPUT.PRINT_LINE(emp_id);
END;
/
-- The result is as follows:
5
38920
ANONYMOUS BLOCK EXECUTE

Nested Value Assignment

Figure 2 shows the syntax diagram for assigning a nested value to a variable.

Figure 2 nested_assignment_value::=
Click to enlarge

The syntax in Figure 2 is described as follows:

variable_name: specifies the name of a variable.
col_name: specifies the column name.
subscript: index, which is used for an array variable. The value can be a value or an expression and must be of the INT type.
value: can be a value or an expression. The type of value must be compatible with the type of variable_name.

Example:

    
         gaussdb=# CREATE TYPE o1 AS (a int, b int);
CREATE TYPE
gaussdb=# DECLARE
    TYPE r1 is VARRAY(10) of o1;
    emp_id  r1;
BEGIN
    emp_id(1).a := 5;-- Assign a value.
    emp_id(1).b := 5*7784;
END;
/
ANONYMOUS BLOCK EXECUTE

In INTO mode, values can be assigned only to the columns at the first layer. Two-dimensional or above arrays are not supported.

INTO/BULK COLLECT INTO

Values returned by statements in a stored procedure are stored in variables. BULK COLLECT INTO allows some or all returned values to be temporarily stored in an array.
An empty result set can be returned.

Syntax

SELECT select_expressions INTO [STRICT] target FROM ...
SELECT INTO [STRICT] target [FROM ..]
EXECUTE [IMMEDIATE] select_expressions BULK COLLECT INTO target ...

The syntax is described as follows:

select_expressions: SQL statement for query. You can use basic SQL commands and INTO clauses to assign the result of a single row or multiple columns to a variable (such as record, row, or scalar variable list).
target: target variable, which can be a record variable, a row variable, or a comma-separated list of simple variables and record/row fields.
STRICT (Optional): If behavior_compat_options is set to 'select_into_return_null' (disabled by default) and the STRICT option is specified, the query must return a non-empty result set. Otherwise, an error is reported: "NO_DATA_FOUND", "TOO_MANY_ROWS" or "QUERY_RETURNED_NO_ROWS". If the option is not specified, the empty result set can be returned.

BULK COLLECT INTO can only assign values to arrays or collections in batches. The collection type uses LIMIT properly to prevent performance deterioration caused by excessive operations on data.
INTO and BULK COLLECT INTO support only direct nesting of record type value with less than four layers.
To return an empty result set, you need to enable PG-compatible mode during database initialization. Set behavior_compat_options to 'select_into_return_null', the compatibility mode is enabled. If the GUC parameter behavior_compat_options is not set, the compatibility mode is disabled.
For array variables, elements in parentheses () are preferentially identified as index sets. Therefore, expressions with parentheses cannot be written after array variables. For example, select (1+3) into va(5) cannot be written as select into va(5) (1+3) or select into va[5] (1+3).
INSERT INTO, UPDATE INTO, DELETE INTO, and EXECUTION INTO cannot return an empty result set.
Multiple variables are not assigned values because the syntax of the subsequent variables is incorrect.
BULK COLLECT INTO can be used only in the A-compatible database.
The IMMEDIATE keyword is used only for syntax compatibility and has no actual meaning.

Example:

gaussdb=# DROP TABLE IF EXISTS customers;
NOTICE:  table "customers" does not exist, skipping
DROP TABLE
gaussdb=# CREATE TABLE customers(id int,name varchar);
CREATE TABLE
gaussdb=# INSERT INTO customers VALUES(1,'ab');
gaussdb=# DECLARE
    my_id integer;
BEGIN
    select id into my_id from customers limit 1; -- Assign a value.
END;
/
ANONYMOUS BLOCK EXECUTE
gaussdb=# DECLARE
    type id_list is varray(6) of customers.id%type;
    id_arr id_list;
BEGIN
    select id bulk collect into id_arr from customers order by id DESC limit 20; -- Assign values in batches.
END;
/
ANONYMOUS BLOCK EXECUTE
gaussdb=# CREATE TABLE test(a integer);
CREATE TABLE
gaussdb=# INSERT INTO test VALUES(1);
INSERT 0 1
gaussdb=# CREATE OR REPLACE FUNCTION check_test() RETURNS integer
    language plpgsql
    AS $function$
    DECLARE
	    b integer;
    BEGIN
	    SELECT INTO b FROM test WHERE a=1; -- Return an empty result set.
            RETURN b;
    END;
    $function$;
CREATE FUNCTION
gaussdb=# SELECT check_test();
 check_test
------------
          1
(1 row)
gaussdb=#  DECLARE
    TYPE id_list IS varray(6) OF customers.id%type;
    id_arr id_list;
    sql_qry varchar2(150);
BEGIN
    sql_qry := 'SELECT id FROM customers ORDER BY id DESC LIMIT 20';
    EXECUTE IMMEDIATE sql_qry BULK COLLECT INTO id_arr; -- Assign values in batches.
END;
/
ANONYMOUS BLOCK EXECUTE