This example shows the use of the HASH BY clause for a table UDF. The table function add2int takes two integer values as input and returns both of them and their sum.
The query in this example selects all columns from add2int, which specifies that its input, dt, is hashed by dt.y1. The specified hashing might not be relevant to the add2int function and is only used for illustration.
The expected processing of the query is that dt is first spooled and then hashed by y1 among the AMPs. The final hashed spool is used as the input to add2int.
CREATE TABLE t1 ( a1 INTEGER, b1 INTEGER); CREATE TABLE t2 ( a2 INTEGER, b2 INTEGER); CREATE FUNCTION add2int ( a INTEGER, b INTEGER) RETURNS TABLE ( addend1 INTEGER, addend2 INTEGER, mysum INTEGER) SPECIFIC add2int LANGUAGE C NO SQL PARAMETER STYLE SQL NOT DETERMINISTIC CALLED ON NULL INPUT EXTERNAL NAME 'CS!add3int!add2int.c'; WITH dt(x1,y1) AS (SELECT a1,b1 FROM t1) SELECT * FROM TABLE (add2int(dt.x1,dt.y1) HASH BY y1) AS aa;