Flow control for T-SQL

This topic provides reference information comparing flow control constructs between Microsoft SQL Server and Amazon Aurora PostgreSQL. You can use this information to understand the similarities and differences in flow control mechanisms when migrating from SQL Server to Aurora PostgreSQL. The topic outlines various flow control commands available in both systems, highlighting where direct equivalents exist and suggesting alternatives where they don’t.

SQL Server Usage

Although SQL Server is a mostly declarative language, it does support flow control commands, which provide run time dynamic changes in script run paths.

Before SQL/PSM was introduced in SQL:1999, the ANSI standard didn’t include flow control constructs. Therefore, there are significant syntax differences among RDBMS engines.

SQL Server provides the following flow control keywords.

Examples

The following example demonstrates a solution for running different processes based on the number of items in an order.

Create and populate an OrderItems table.

CREATE TABLE OrderItems
(
OrderID INT NOT NULL,
Item VARCHAR(20) NOT NULL,
Quantity SMALLINT NOT NULL,
PRIMARY KEY(OrderID, Item)
);

INSERT INTO OrderItems (OrderID, Item, Quantity)
VALUES
(1, 'M8 Bolt', 100),
(2, 'M8 Nut', 100),
(3, 'M8 Washer', 200);

Declare a cursor for looping through all OrderItems and calculating the total quantity for each order.

DECLARE OrderItemCursor CURSOR FAST_FORWARD
FOR
SELECT OrderID,
  SUM(Quantity) AS NumItems
FROM OrderItems
GROUP BY OrderID
ORDER BY OrderID;

DECLARE @OrderID INT, @NumItems INT;

-- Instantiate the cursor and loop through all orders.
OPEN OrderItemCursor;

FETCH NEXT FROM OrderItemCursor
INTO @OrderID, @NumItems

WHILE @@Fetch_Status = 0
BEGIN;
  IF @NumItems > 100
    PRINT 'EXECUTING LogLargeOrder - '
    + CAST(@OrderID AS VARCHAR(5))
    + ' ' + CAST(@NumItems AS VARCHAR(5));
  ELSE
    PRINT 'EXECUTING LogSmallOrder - '
    + CAST(@OrderID AS VARCHAR(5))
    + ' ' + CAST(@NumItems AS VARCHAR(5));

FETCH NEXT FROM OrderItemCursor
INTO @OrderID, @NumItems;
END;

-- Close and deallocate the cursor.
CLOSE OrderItemCursor;
DEALLOCATE OrderItemCursor;

For the preceding example, the result looks as shown following.

EXECUTING LogSmallOrder - 1 100
EXECUTING LogSmallOrder - 2 100
EXECUTING LogLargeOrder - 3 200

For more information, see Control-of-Flow in the SQL Server documentation.

PostgreSQL Usage

Amazon Aurora PostgreSQL-Compatible Edition (Aurora PostgreSQL) provides the following flow control constructs:

Examples

The following example demonstrates a solution for running different logic based on the number of items in an order. It provides the same functionality as the example for SQL Server flow control. However, unlike the SQL Server example ran as a batch script, Aurora PostgreSQL variables can only be used in stored routines such as procedures and functions.

Create and populate an OrderItems table.

CREATE TABLE OrderItems
(
  OrderID INT NOT NULL,
  Item VARCHAR(20) NOT NULL,
  Quantity SMALLINT NOT NULL,
  PRIMARY KEY(OrderID, Item)
);

INSERT INTO OrderItems (OrderID, Item, Quantity)
VALUES
(1, 'M8 Bolt', 100),
(2, 'M8 Nut', 100),
(3, 'M8 Washer', 200);

Create a procedure to declare a cursor and loop through the order items.

CREATE OR REPLACE FUNCTION P()
  RETURNS numeric
  LANGUAGE plpgsql
AS $function$
DECLARE
  done int default false;
  var_OrderID int;
  var_NumItems int;
  OrderItemCursor CURSOR FOR SELECT OrderID, SUM(Quantity) AS NumItems
  FROM OrderItems
  GROUP BY OrderID
  ORDER BY OrderID;

  BEGIN
    OPEN OrderItemCursor;
    LOOP
      fetch from OrderItemCursor INTO var_OrderID, var_NumItems;
    EXIT WHEN NOT FOUND;
    IF var_NumItems > 100 THEN
      RAISE NOTICE 'EXECUTING LogLargeOrder - %s',var_OrderID;
      RAISE NOTICE 'Num Items: %s', var_NumItems;
    ELSE
      RAISE NOTICE 'EXECUTING LogSmallOrder - %s',var_OrderID;
      RAISE NOTICE 'Num Items: %s', var_NumItems;
    END IF;
    END LOOP;
done = TRUE;
CLOSE OrderItemCursor;
END; $function$

Summary

While there are some syntax differences between SQL Server and Aurora PostgreSQL flow control statements, most rewrites should be straightforward. The following table summarizes the differences and identifies how to modify T-SQL code to support similar functionality in Aurora PostgreSQL PL/pgSQL.

For more information, see Control Structures in the PostgreSQL documentation.