--
-- SELECT_DISTINCT
--
--
-- awk '{print $3;}' onek.data | sort -n | uniq
--
SELECT DISTINCT two FROM onek ORDER BY 1;
 two 
-----
   0
   1
(2 rows)

--
-- awk '{print $5;}' onek.data | sort -n | uniq
--
SELECT DISTINCT ten FROM onek ORDER BY 1;
 ten 
-----
   0
   1
   2
   3
   4
   5
   6
   7
   8
   9
(10 rows)

--
-- awk '{print $16;}' onek.data | sort -d | uniq
--
SELECT DISTINCT string4 FROM onek ORDER BY 1;
 string4 
---------
 AAAAxx
 HHHHxx
 OOOOxx
 VVVVxx
(4 rows)

--
-- awk '{print $3,$16,$5;}' onek.data | sort -d | uniq |
-- sort +0n -1 +1d -2 +2n -3
--
SELECT DISTINCT two, string4, ten
   FROM onek
   ORDER BY two using <, string4 using <, ten using <;
 two | string4 | ten 
-----+---------+-----
   0 | AAAAxx  |   0
   0 | AAAAxx  |   2
   0 | AAAAxx  |   4
   0 | AAAAxx  |   6
   0 | AAAAxx  |   8
   0 | HHHHxx  |   0
   0 | HHHHxx  |   2
   0 | HHHHxx  |   4
   0 | HHHHxx  |   6
   0 | HHHHxx  |   8
   0 | OOOOxx  |   0
   0 | OOOOxx  |   2
   0 | OOOOxx  |   4
   0 | OOOOxx  |   6
   0 | OOOOxx  |   8
   0 | VVVVxx  |   0
   0 | VVVVxx  |   2
   0 | VVVVxx  |   4
   0 | VVVVxx  |   6
   0 | VVVVxx  |   8
   1 | AAAAxx  |   1
   1 | AAAAxx  |   3
   1 | AAAAxx  |   5
   1 | AAAAxx  |   7
   1 | AAAAxx  |   9
   1 | HHHHxx  |   1
   1 | HHHHxx  |   3
   1 | HHHHxx  |   5
   1 | HHHHxx  |   7
   1 | HHHHxx  |   9
   1 | OOOOxx  |   1
   1 | OOOOxx  |   3
   1 | OOOOxx  |   5
   1 | OOOOxx  |   7
   1 | OOOOxx  |   9
   1 | VVVVxx  |   1
   1 | VVVVxx  |   3
   1 | VVVVxx  |   5
   1 | VVVVxx  |   7
   1 | VVVVxx  |   9
(40 rows)

--
-- awk '{print $2;}' person.data |
-- awk '{if(NF!=1){print $2;}else{print;}}' - emp.data |
-- awk '{if(NF!=1){print $2;}else{print;}}' - student.data |
-- awk 'BEGIN{FS="      ";}{if(NF!=1){print $5;}else{print;}}' - stud_emp.data |
-- sort -n -r | uniq
--
SELECT DISTINCT p.age FROM person* p ORDER BY age using >;
 age 
-----
  98
  88
  78
  68
  60
  58
  50
  48
  40
  38
  34
  30
  28
  25
  24
  23
  20
  19
  18
   8
(20 rows)

--
-- Check mentioning same column more than once
--
EXPLAIN (VERBOSE, COSTS OFF)
SELECT count(*) FROM
  (SELECT DISTINCT two, four, two FROM tenk1) ss;
                       QUERY PLAN                       
--------------------------------------------------------
 Aggregate
   Output: count(*)
   ->  HashAggregate
         Output: tenk1.two, tenk1.four, tenk1.two
         Group Key: tenk1.two, tenk1.four
         ->  Seq Scan on public.tenk1
               Output: tenk1.two, tenk1.four, tenk1.two
(7 rows)

SELECT count(*) FROM
  (SELECT DISTINCT two, four, two FROM tenk1) ss;
 count 
-------
     4
(1 row)

--
-- Compare results between plans using sorting and plans using hash
-- aggregation. Force spilling in both cases by setting work_mem low.
--
SET work_mem='64kB';
-- Produce results with sorting.
SET enable_hashagg=FALSE;
SET jit_above_cost=0;
EXPLAIN (costs off)
SELECT DISTINCT g%1000 FROM generate_series(0,9999) g;
                   QUERY PLAN                   
------------------------------------------------
 Unique
   ->  Sort
         Sort Key: ((g % 1000))
         ->  Function Scan on generate_series g
(4 rows)

CREATE TABLE distinct_group_1 AS
SELECT DISTINCT g%1000 FROM generate_series(0,9999) g;
SET jit_above_cost TO DEFAULT;
CREATE TABLE distinct_group_2 AS
SELECT DISTINCT (g%1000)::text FROM generate_series(0,9999) g;
SET enable_seqscan = 0;
-- Check to see we get an incremental sort plan
EXPLAIN (costs off)
SELECT DISTINCT hundred, two FROM tenk1;
                     QUERY PLAN                      
-----------------------------------------------------
 Unique
   ->  Incremental Sort
         Sort Key: hundred, two
         Presorted Key: hundred
         ->  Index Scan using tenk1_hundred on tenk1
(5 rows)

RESET enable_seqscan;
SET enable_hashagg=TRUE;
-- Produce results with hash aggregation.
SET enable_sort=FALSE;
SET jit_above_cost=0;
EXPLAIN (costs off)
SELECT DISTINCT g%1000 FROM generate_series(0,9999) g;
                QUERY PLAN                
------------------------------------------
 HashAggregate
   Group Key: (g % 1000)
   ->  Function Scan on generate_series g
(3 rows)

CREATE TABLE distinct_hash_1 AS
SELECT DISTINCT g%1000 FROM generate_series(0,9999) g;
SET jit_above_cost TO DEFAULT;
CREATE TABLE distinct_hash_2 AS
SELECT DISTINCT (g%1000)::text FROM generate_series(0,9999) g;
SET enable_sort=TRUE;
SET work_mem TO DEFAULT;
-- Compare results
(SELECT * FROM distinct_hash_1 EXCEPT SELECT * FROM distinct_group_1)
  UNION ALL
(SELECT * FROM distinct_group_1 EXCEPT SELECT * FROM distinct_hash_1);
 ?column? 
----------
(0 rows)

(SELECT * FROM distinct_hash_1 EXCEPT SELECT * FROM distinct_group_1)
  UNION ALL
(SELECT * FROM distinct_group_1 EXCEPT SELECT * FROM distinct_hash_1);
 ?column? 
----------
(0 rows)

DROP TABLE distinct_hash_1;
DROP TABLE distinct_hash_2;
DROP TABLE distinct_group_1;
DROP TABLE distinct_group_2;
-- Test parallel DISTINCT
SET parallel_tuple_cost=0;
SET parallel_setup_cost=0;
SET min_parallel_table_scan_size=0;
SET max_parallel_workers_per_gather=2;
-- Ensure we get a parallel plan
EXPLAIN (costs off)
SELECT DISTINCT four FROM tenk1;
                     QUERY PLAN                     
----------------------------------------------------
 Unique
   ->  Sort
         Sort Key: four
         ->  Gather
               Workers Planned: 2
               ->  HashAggregate
                     Group Key: four
                     ->  Parallel Seq Scan on tenk1
(8 rows)

-- Ensure the parallel plan produces the correct results
SELECT DISTINCT four FROM tenk1;
 four 
------
    0
    1
    2
    3
(4 rows)

CREATE OR REPLACE FUNCTION distinct_func(a INT) RETURNS INT AS $$
  BEGIN
    RETURN a;
  END;
$$ LANGUAGE plpgsql PARALLEL UNSAFE;
-- Ensure we don't do parallel distinct with a parallel unsafe function
EXPLAIN (COSTS OFF)
SELECT DISTINCT distinct_func(1) FROM tenk1;
                        QUERY PLAN                        
----------------------------------------------------------
 Unique
   ->  Sort
         Sort Key: (distinct_func(1))
         ->  Index Only Scan using tenk1_hundred on tenk1
(4 rows)

-- make the function parallel safe
CREATE OR REPLACE FUNCTION distinct_func(a INT) RETURNS INT AS $$
  BEGIN
    RETURN a;
  END;
$$ LANGUAGE plpgsql PARALLEL SAFE;
-- Ensure we do parallel distinct now that the function is parallel safe
EXPLAIN (COSTS OFF)
SELECT DISTINCT distinct_func(1) FROM tenk1;
                     QUERY PLAN                     
----------------------------------------------------
 Unique
   ->  Gather Merge
         Workers Planned: 2
         ->  Unique
               ->  Sort
                     Sort Key: (distinct_func(1))
                     ->  Parallel Seq Scan on tenk1
(7 rows)

RESET max_parallel_workers_per_gather;
RESET min_parallel_table_scan_size;
RESET parallel_setup_cost;
RESET parallel_tuple_cost;
--
-- Test the planner's ability to use a LIMIT 1 instead of a Unique node when
-- all of the distinct_pathkeys have been marked as redundant
--
-- Ensure we get a plan with a Limit 1
EXPLAIN (COSTS OFF)
SELECT DISTINCT four FROM tenk1 WHERE four = 0;
         QUERY PLAN         
----------------------------
 Limit
   ->  Seq Scan on tenk1
         Filter: (four = 0)
(3 rows)

-- Ensure the above gives us the correct result
SELECT DISTINCT four FROM tenk1 WHERE four = 0;
 four 
------
    0
(1 row)

-- Ensure we get a plan with a Limit 1
EXPLAIN (COSTS OFF)
SELECT DISTINCT four FROM tenk1 WHERE four = 0 AND two <> 0;
                 QUERY PLAN                  
---------------------------------------------
 Limit
   ->  Seq Scan on tenk1
         Filter: ((two <> 0) AND (four = 0))
(3 rows)

-- Ensure no rows are returned
SELECT DISTINCT four FROM tenk1 WHERE four = 0 AND two <> 0;
 four 
------
(0 rows)

-- Ensure we get a plan with a Limit 1 when the SELECT list contains constants
EXPLAIN (COSTS OFF)
SELECT DISTINCT four,1,2,3 FROM tenk1 WHERE four = 0;
         QUERY PLAN         
----------------------------
 Limit
   ->  Seq Scan on tenk1
         Filter: (four = 0)
(3 rows)

-- Ensure we only get 1 row
SELECT DISTINCT four,1,2,3 FROM tenk1 WHERE four = 0;
 four | ?column? | ?column? | ?column? 
------+----------+----------+----------
    0 |        1 |        2 |        3
(1 row)

--
-- Also, some tests of IS DISTINCT FROM, which doesn't quite deserve its
-- very own regression file.
--
CREATE TEMP TABLE disttable (f1 integer);
INSERT INTO DISTTABLE VALUES(1);
INSERT INTO DISTTABLE VALUES(2);
INSERT INTO DISTTABLE VALUES(3);
INSERT INTO DISTTABLE VALUES(NULL);
-- basic cases
SELECT f1, f1 IS DISTINCT FROM 2 as "not 2" FROM disttable;
 f1 | not 2 
----+-------
  1 | t
  2 | f
  3 | t
    | t
(4 rows)

SELECT f1, f1 IS DISTINCT FROM NULL as "not null" FROM disttable;
 f1 | not null 
----+----------
  1 | t
  2 | t
  3 | t
    | f
(4 rows)

SELECT f1, f1 IS DISTINCT FROM f1 as "false" FROM disttable;
 f1 | false 
----+-------
  1 | f
  2 | f
  3 | f
    | f
(4 rows)

SELECT f1, f1 IS DISTINCT FROM f1+1 as "not null" FROM disttable;
 f1 | not null 
----+----------
  1 | t
  2 | t
  3 | t
    | f
(4 rows)

-- check that optimizer constant-folds it properly
SELECT 1 IS DISTINCT FROM 2 as "yes";
 yes 
-----
 t
(1 row)

SELECT 2 IS DISTINCT FROM 2 as "no";
 no 
----
 f
(1 row)

SELECT 2 IS DISTINCT FROM null as "yes";
 yes 
-----
 t
(1 row)

SELECT null IS DISTINCT FROM null as "no";
 no 
----
 f
(1 row)

-- negated form
SELECT 1 IS NOT DISTINCT FROM 2 as "no";
 no 
----
 f
(1 row)

SELECT 2 IS NOT DISTINCT FROM 2 as "yes";
 yes 
-----
 t
(1 row)

SELECT 2 IS NOT DISTINCT FROM null as "no";
 no 
----
 f
(1 row)

SELECT null IS NOT DISTINCT FROM null as "yes";
 yes 
-----
 t
(1 row)