mirror of
https://git.postgresql.org/git/postgresql.git
synced 2026-02-13 18:07:05 +08:00
0a6ea4001a
Up to now we've not worried much about joins where the join key is a relation's CTID column, reasoning that storing a table's CTIDs in some other table would be pretty useless. However, there are use-cases for this sort of query involving self-joins, so that argument doesn't really hold water. With larger relations, a merge or hash join is desirable. We had a btree opclass for type "tid", allowing merge joins on CTID, but no hash opclass so that hash joins weren't possible. Add the missing infrastructure. This also potentially enables hash aggregation on "tid", though the use-cases for that aren't too clear. Discussion: https://postgr.es/m/1853.1545453106@sss.pgh.pa.us
98 lines
3.3 KiB
PL/PgSQL
98 lines
3.3 KiB
PL/PgSQL
-- tests for tidscans
|
|
|
|
CREATE TABLE tidscan(id integer);
|
|
|
|
-- only insert a few rows, we don't want to spill onto a second table page
|
|
INSERT INTO tidscan VALUES (1), (2), (3);
|
|
|
|
-- show ctids
|
|
SELECT ctid, * FROM tidscan;
|
|
|
|
-- ctid equality - implemented as tidscan
|
|
EXPLAIN (COSTS OFF)
|
|
SELECT ctid, * FROM tidscan WHERE ctid = '(0,1)';
|
|
SELECT ctid, * FROM tidscan WHERE ctid = '(0,1)';
|
|
|
|
EXPLAIN (COSTS OFF)
|
|
SELECT ctid, * FROM tidscan WHERE '(0,1)' = ctid;
|
|
SELECT ctid, * FROM tidscan WHERE '(0,1)' = ctid;
|
|
|
|
-- OR'd clauses
|
|
EXPLAIN (COSTS OFF)
|
|
SELECT ctid, * FROM tidscan WHERE ctid = '(0,2)' OR '(0,1)' = ctid;
|
|
SELECT ctid, * FROM tidscan WHERE ctid = '(0,2)' OR '(0,1)' = ctid;
|
|
|
|
-- ctid = ScalarArrayOp - implemented as tidscan
|
|
EXPLAIN (COSTS OFF)
|
|
SELECT ctid, * FROM tidscan WHERE ctid = ANY(ARRAY['(0,1)', '(0,2)']::tid[]);
|
|
SELECT ctid, * FROM tidscan WHERE ctid = ANY(ARRAY['(0,1)', '(0,2)']::tid[]);
|
|
|
|
-- ctid != ScalarArrayOp - can't be implemented as tidscan
|
|
EXPLAIN (COSTS OFF)
|
|
SELECT ctid, * FROM tidscan WHERE ctid != ANY(ARRAY['(0,1)', '(0,2)']::tid[]);
|
|
SELECT ctid, * FROM tidscan WHERE ctid != ANY(ARRAY['(0,1)', '(0,2)']::tid[]);
|
|
|
|
-- tid equality extracted from sub-AND clauses
|
|
EXPLAIN (COSTS OFF)
|
|
SELECT ctid, * FROM tidscan
|
|
WHERE (id = 3 AND ctid IN ('(0,2)', '(0,3)')) OR (ctid = '(0,1)' AND id = 1);
|
|
SELECT ctid, * FROM tidscan
|
|
WHERE (id = 3 AND ctid IN ('(0,2)', '(0,3)')) OR (ctid = '(0,1)' AND id = 1);
|
|
|
|
-- nestloop-with-inner-tidscan joins on tid
|
|
SET enable_hashjoin TO off; -- otherwise hash join might win
|
|
EXPLAIN (COSTS OFF)
|
|
SELECT t1.ctid, t1.*, t2.ctid, t2.*
|
|
FROM tidscan t1 JOIN tidscan t2 ON t1.ctid = t2.ctid WHERE t1.id = 1;
|
|
SELECT t1.ctid, t1.*, t2.ctid, t2.*
|
|
FROM tidscan t1 JOIN tidscan t2 ON t1.ctid = t2.ctid WHERE t1.id = 1;
|
|
EXPLAIN (COSTS OFF)
|
|
SELECT t1.ctid, t1.*, t2.ctid, t2.*
|
|
FROM tidscan t1 LEFT JOIN tidscan t2 ON t1.ctid = t2.ctid WHERE t1.id = 1;
|
|
SELECT t1.ctid, t1.*, t2.ctid, t2.*
|
|
FROM tidscan t1 LEFT JOIN tidscan t2 ON t1.ctid = t2.ctid WHERE t1.id = 1;
|
|
RESET enable_hashjoin;
|
|
|
|
-- exercise backward scan and rewind
|
|
BEGIN;
|
|
DECLARE c CURSOR FOR
|
|
SELECT ctid, * FROM tidscan WHERE ctid = ANY(ARRAY['(0,1)', '(0,2)']::tid[]);
|
|
FETCH ALL FROM c;
|
|
FETCH BACKWARD 1 FROM c;
|
|
FETCH FIRST FROM c;
|
|
ROLLBACK;
|
|
|
|
-- tidscan via CURRENT OF
|
|
BEGIN;
|
|
DECLARE c CURSOR FOR SELECT ctid, * FROM tidscan;
|
|
FETCH NEXT FROM c; -- skip one row
|
|
FETCH NEXT FROM c;
|
|
-- perform update
|
|
EXPLAIN (ANALYZE, COSTS OFF, SUMMARY OFF, TIMING OFF)
|
|
UPDATE tidscan SET id = -id WHERE CURRENT OF c RETURNING *;
|
|
FETCH NEXT FROM c;
|
|
-- perform update
|
|
EXPLAIN (ANALYZE, COSTS OFF, SUMMARY OFF, TIMING OFF)
|
|
UPDATE tidscan SET id = -id WHERE CURRENT OF c RETURNING *;
|
|
SELECT * FROM tidscan;
|
|
-- position cursor past any rows
|
|
FETCH NEXT FROM c;
|
|
-- should error out
|
|
EXPLAIN (ANALYZE, COSTS OFF, SUMMARY OFF, TIMING OFF)
|
|
UPDATE tidscan SET id = -id WHERE CURRENT OF c RETURNING *;
|
|
ROLLBACK;
|
|
|
|
-- bulk joins on CTID
|
|
-- (these plans don't use TID scans, but this still seems like an
|
|
-- appropriate place for these tests)
|
|
EXPLAIN (COSTS OFF)
|
|
SELECT count(*) FROM tenk1 t1 JOIN tenk1 t2 ON t1.ctid = t2.ctid;
|
|
SELECT count(*) FROM tenk1 t1 JOIN tenk1 t2 ON t1.ctid = t2.ctid;
|
|
SET enable_hashjoin TO off;
|
|
EXPLAIN (COSTS OFF)
|
|
SELECT count(*) FROM tenk1 t1 JOIN tenk1 t2 ON t1.ctid = t2.ctid;
|
|
SELECT count(*) FROM tenk1 t1 JOIN tenk1 t2 ON t1.ctid = t2.ctid;
|
|
RESET enable_hashjoin;
|
|
|
|
DROP TABLE tidscan;
|