-- Generic extended statistics support

-- We will be checking execution plans without/with statistics, so
-- let's make sure we get simple non-parallel plans. Also set the
-- work_mem low so that we can use small amounts of data.
SET max_parallel_workers = 0;
SET max_parallel_workers_per_gather = 0;
SET work_mem = '128kB';

-- Ensure stats are dropped sanely
CREATE TABLE ab1 (a INTEGER, b INTEGER, c INTEGER);
CREATE STATISTICS ab1_a_b_stats ON (a, b) FROM ab1;
DROP STATISTICS ab1_a_b_stats;

CREATE SCHEMA regress_schema_2;
CREATE STATISTICS regress_schema_2.ab1_a_b_stats ON (a, b) FROM ab1;

-- Let's also verify the pg_get_statisticsextdef output looks sane.
SELECT pg_get_statisticsextdef(oid) FROM pg_statistic_ext WHERE staname = 'ab1_a_b_stats';

DROP STATISTICS regress_schema_2.ab1_a_b_stats;

-- Ensure statistics are dropped when columns are
CREATE STATISTICS ab1_b_c_stats ON (b, c) FROM ab1;
CREATE STATISTICS ab1_a_b_c_stats ON (a, b, c) FROM ab1;
CREATE STATISTICS ab1_a_b_stats ON (a, b) FROM ab1;
ALTER TABLE ab1 DROP COLUMN a;
\d ab1
DROP TABLE ab1;

-- Ensure things work sanely with SET STATISTICS 0
CREATE TABLE ab1 (a INTEGER, b INTEGER);
ALTER TABLE ab1 ALTER a SET STATISTICS 0;
INSERT INTO ab1 SELECT a, a%23 FROM generate_series(1, 1000) a;
CREATE STATISTICS ab1_a_b_stats ON (a, b) FROM ab1;
ANALYZE ab1;
ALTER TABLE ab1 ALTER a SET STATISTICS -1;
ANALYZE ab1;
DROP TABLE ab1;


-- n-distinct tests
CREATE TABLE ndistinct (
    filler1 TEXT,
    filler2 NUMERIC,
    a INT,
    b INT,
    filler3 DATE,
    c INT,
    d INT
);

-- over-estimates when using only per-column statistics
INSERT INTO ndistinct (a, b, c, filler1)
     SELECT i/100, i/100, i/100, cash_words((i/100)::money)
       FROM generate_series(1,30000) s(i);

ANALYZE ndistinct;

-- Group Aggregate, due to over-estimate of the number of groups
EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY a, b;

EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY b, c;

EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY a, b, c;

EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY a, b, c, d;

EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY b, c, d;

-- unknown column
CREATE STATISTICS s10 ON (unknown_column) FROM ndistinct;

-- single column
CREATE STATISTICS s10 ON (a) FROM ndistinct;

-- single column, duplicated
CREATE STATISTICS s10 ON (a,a) FROM ndistinct;

-- two columns, one duplicated
CREATE STATISTICS s10 ON (a, a, b) FROM ndistinct;

-- correct command
CREATE STATISTICS s10 ON (a, b, c) FROM ndistinct;

ANALYZE ndistinct;

SELECT staenabled, standistinct
  FROM pg_statistic_ext WHERE starelid = 'ndistinct'::regclass;

-- Hash Aggregate, thanks to estimates improved by the statistic
EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY a, b;

EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY b, c;

EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY a, b, c;

-- last two plans keep using Group Aggregate, because 'd' is not covered
-- by the statistic and while it's NULL-only we assume 200 values for it
EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY a, b, c, d;

EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY b, c, d;

TRUNCATE TABLE ndistinct;

-- under-estimates when using only per-column statistics
INSERT INTO ndistinct (a, b, c, filler1)
     SELECT mod(i,50), mod(i,51), mod(i,32),
            cash_words(mod(i,33)::int::money)
       FROM generate_series(1,10000) s(i);

ANALYZE ndistinct;

SELECT staenabled, standistinct
  FROM pg_statistic_ext WHERE starelid = 'ndistinct'::regclass;

-- plans using Group Aggregate, thanks to using correct esimates
EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY a, b;

EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY a, b, c;

EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY a, b, c, d;

EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY b, c, d;

EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY a, d;

DROP STATISTICS s10;

SELECT staenabled, standistinct
  FROM pg_statistic_ext WHERE starelid = 'ndistinct'::regclass;

-- dropping the statistics switches the plans to Hash Aggregate,
-- due to under-estimates
EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY a, b;

EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY a, b, c;

EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY a, b, c, d;

EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY b, c, d;

EXPLAIN (COSTS off)
 SELECT COUNT(*) FROM ndistinct GROUP BY a, d;

DROP TABLE ndistinct;