From 927622154b73e40945e9581ab5456864b4fdc5e6 Mon Sep 17 00:00:00 2001 From: Corey Huinker Date: Tue, 4 Nov 2025 15:24:35 -0500 Subject: [PATCH v8 1/7] Refactor output format of pg_ndistinct. The existing format of pg_ndistinct uses a single-object JSON structure where each key is itself a comma-separated list of attnums. While this is a very compact format, it's confusing to read and is difficult to manipulate values within the object. This wasn't a concern until statistics import functions were introduced, enabling users to inject hypothetical statistics into an object to observe their effect on the query planner. The new format is an array of objects, each object must have the keys "attributes", which must contain an array of attnums, and "ndistinct", which must be an integer. This is a quirk because the underlying internal storage is a double, but the value stored was always an integer. The change in format is adequately described from the changes to src/test/regress/expected/stats_ext.out so description here is redundant. --- src/backend/statistics/mvdistinct.c | 19 +-- src/test/regress/expected/stats_ext.out | 156 +++++++++++++++++++++--- src/test/regress/sql/stats_ext.sql | 12 +- doc/src/sgml/perform.sgml | 34 +++++- 4 files changed, 186 insertions(+), 35 deletions(-) diff --git a/src/backend/statistics/mvdistinct.c b/src/backend/statistics/mvdistinct.c index 7e7a63405c8b..ca60841b813e 100644 --- a/src/backend/statistics/mvdistinct.c +++ b/src/backend/statistics/mvdistinct.c @@ -360,26 +360,27 @@ pg_ndistinct_out(PG_FUNCTION_ARGS) StringInfoData str; initStringInfo(&str); - appendStringInfoChar(&str, '{'); + appendStringInfoChar(&str, '['); for (i = 0; i < ndist->nitems; i++) { - int j; MVNDistinctItem item = ndist->items[i]; if (i > 0) appendStringInfoString(&str, ", "); - for (j = 0; j < item.nattributes; j++) - { - AttrNumber attnum = item.attributes[j]; + if (item.nattributes <= 0) + elog(ERROR, "invalid zero-length attribute array in MVNDistinct"); - appendStringInfo(&str, "%s%d", (j == 0) ? "\"" : ", ", attnum); - } - appendStringInfo(&str, "\": %d", (int) item.ndistinct); + appendStringInfo(&str, "{\"attributes\": [%d", item.attributes[0]); + + for (int j = 1; j < item.nattributes; j++) + appendStringInfo(&str, ", %d", item.attributes[j]); + + appendStringInfo(&str, "], \"ndistinct\": %d}", (int) item.ndistinct); } - appendStringInfoChar(&str, '}'); + appendStringInfoChar(&str, ']'); PG_RETURN_CSTRING(str.data); } diff --git a/src/test/regress/expected/stats_ext.out b/src/test/regress/expected/stats_ext.out index 73a7ef973559..2dc771369e52 100644 --- a/src/test/regress/expected/stats_ext.out +++ b/src/test/regress/expected/stats_ext.out @@ -196,7 +196,7 @@ Statistics objects: "public.ab1_a_b_stats" ON a, b FROM ab1; STATISTICS 0 ANALYZE ab1; -SELECT stxname, stxdndistinct, stxddependencies, stxdmcv, stxdinherit +SELECT stxname, jsonb_pretty(d.stxdndistinct::text::jsonb) AS stxdndistinct, stxddependencies, stxdmcv, stxdinherit FROM pg_statistic_ext s LEFT JOIN pg_statistic_ext_data d ON (d.stxoid = s.oid) WHERE s.stxname = 'ab1_a_b_stats'; stxname | stxdndistinct | stxddependencies | stxdmcv | stxdinherit @@ -476,13 +476,43 @@ SELECT * FROM check_estimated_rows('SELECT COUNT(*) FROM ndistinct GROUP BY a, ( -- correct command CREATE STATISTICS s10 ON a, b, c FROM ndistinct; ANALYZE ndistinct; -SELECT s.stxkind, d.stxdndistinct +SELECT s.stxkind, jsonb_pretty(d.stxdndistinct::text::jsonb) AS stxdndistinct FROM pg_statistic_ext s, pg_statistic_ext_data d WHERE s.stxrelid = 'ndistinct'::regclass AND d.stxoid = s.oid; - stxkind | stxdndistinct ----------+----------------------------------------------------- - {d,f,m} | {"3, 4": 11, "3, 6": 11, "4, 6": 11, "3, 4, 6": 11} + stxkind | stxdndistinct +---------+-------------------------- + {d,f,m} | [ + + | { + + | "ndistinct": 11,+ + | "attributes": [ + + | 3, + + | 4 + + | ] + + | }, + + | { + + | "ndistinct": 11,+ + | "attributes": [ + + | 3, + + | 6 + + | ] + + | }, + + | { + + | "ndistinct": 11,+ + | "attributes": [ + + | 4, + + | 6 + + | ] + + | }, + + | { + + | "ndistinct": 11,+ + | "attributes": [ + + | 3, + + | 4, + + | 6 + + | ] + + | } + + | ] (1 row) -- minor improvement, make sure the ctid does not break the matching @@ -558,13 +588,43 @@ INSERT INTO ndistinct (a, b, c, filler1) mod(i,23) || ' dollars and zero cents' FROM generate_series(1,1000) s(i); ANALYZE ndistinct; -SELECT s.stxkind, d.stxdndistinct +SELECT s.stxkind, jsonb_pretty(d.stxdndistinct::text::jsonb) AS stxdndistinct FROM pg_statistic_ext s, pg_statistic_ext_data d WHERE s.stxrelid = 'ndistinct'::regclass AND d.stxoid = s.oid; - stxkind | stxdndistinct ----------+---------------------------------------------------------- - {d,f,m} | {"3, 4": 221, "3, 6": 247, "4, 6": 323, "3, 4, 6": 1000} + stxkind | stxdndistinct +---------+---------------------------- + {d,f,m} | [ + + | { + + | "ndistinct": 221, + + | "attributes": [ + + | 3, + + | 4 + + | ] + + | }, + + | { + + | "ndistinct": 247, + + | "attributes": [ + + | 3, + + | 6 + + | ] + + | }, + + | { + + | "ndistinct": 323, + + | "attributes": [ + + | 4, + + | 6 + + | ] + + | }, + + | { + + | "ndistinct": 1000,+ + | "attributes": [ + + | 3, + + | 4, + + | 6 + + | ] + + | } + + | ] (1 row) -- correct estimates @@ -623,7 +683,7 @@ SELECT * FROM check_estimated_rows('SELECT COUNT(*) FROM ndistinct GROUP BY a, ( (1 row) DROP STATISTICS s10; -SELECT s.stxkind, d.stxdndistinct +SELECT s.stxkind, jsonb_pretty(d.stxdndistinct::text::jsonb) AS stxdndistinct FROM pg_statistic_ext s, pg_statistic_ext_data d WHERE s.stxrelid = 'ndistinct'::regclass AND d.stxoid = s.oid; @@ -707,13 +767,43 @@ SELECT * FROM check_estimated_rows('SELECT COUNT(*) FROM ndistinct GROUP BY a, ( CREATE STATISTICS s10 (ndistinct) ON (a+1), (b+100), (2*c) FROM ndistinct; ANALYZE ndistinct; -SELECT s.stxkind, d.stxdndistinct +SELECT s.stxkind, jsonb_pretty(d.stxdndistinct::text::jsonb) AS stxdndistinct FROM pg_statistic_ext s, pg_statistic_ext_data d WHERE s.stxrelid = 'ndistinct'::regclass AND d.stxoid = s.oid; - stxkind | stxdndistinct ----------+------------------------------------------------------------------- - {d,e} | {"-1, -2": 221, "-1, -3": 247, "-2, -3": 323, "-1, -2, -3": 1000} + stxkind | stxdndistinct +---------+---------------------------- + {d,e} | [ + + | { + + | "ndistinct": 221, + + | "attributes": [ + + | -1, + + | -2 + + | ] + + | }, + + | { + + | "ndistinct": 247, + + | "attributes": [ + + | -1, + + | -3 + + | ] + + | }, + + | { + + | "ndistinct": 323, + + | "attributes": [ + + | -2, + + | -3 + + | ] + + | }, + + | { + + | "ndistinct": 1000,+ + | "attributes": [ + + | -1, + + | -2, + + | -3 + + | ] + + | } + + | ] (1 row) SELECT * FROM check_estimated_rows('SELECT COUNT(*) FROM ndistinct GROUP BY (a+1), (b+100)'); @@ -756,13 +846,43 @@ SELECT * FROM check_estimated_rows('SELECT COUNT(*) FROM ndistinct GROUP BY a, b CREATE STATISTICS s10 (ndistinct) ON a, b, (2*c) FROM ndistinct; ANALYZE ndistinct; -SELECT s.stxkind, d.stxdndistinct +SELECT s.stxkind, jsonb_pretty(d.stxdndistinct::text::jsonb) AS stxdndistinct FROM pg_statistic_ext s, pg_statistic_ext_data d WHERE s.stxrelid = 'ndistinct'::regclass AND d.stxoid = s.oid; - stxkind | stxdndistinct ----------+------------------------------------------------------------- - {d,e} | {"3, 4": 221, "3, -1": 247, "4, -1": 323, "3, 4, -1": 1000} + stxkind | stxdndistinct +---------+---------------------------- + {d,e} | [ + + | { + + | "ndistinct": 221, + + | "attributes": [ + + | 3, + + | 4 + + | ] + + | }, + + | { + + | "ndistinct": 247, + + | "attributes": [ + + | 3, + + | -1 + + | ] + + | }, + + | { + + | "ndistinct": 323, + + | "attributes": [ + + | 4, + + | -1 + + | ] + + | }, + + | { + + | "ndistinct": 1000,+ + | "attributes": [ + + | 3, + + | 4, + + | -1 + + | ] + + | } + + | ] (1 row) SELECT * FROM check_estimated_rows('SELECT COUNT(*) FROM ndistinct GROUP BY a, b'); diff --git a/src/test/regress/sql/stats_ext.sql b/src/test/regress/sql/stats_ext.sql index 96771600d578..207c431e68c9 100644 --- a/src/test/regress/sql/stats_ext.sql +++ b/src/test/regress/sql/stats_ext.sql @@ -125,7 +125,7 @@ ALTER TABLE ab1 ALTER a SET STATISTICS -1; ALTER STATISTICS ab1_a_b_stats SET STATISTICS 0; \d ab1 ANALYZE ab1; -SELECT stxname, stxdndistinct, stxddependencies, stxdmcv, stxdinherit +SELECT stxname, jsonb_pretty(d.stxdndistinct::text::jsonb) AS stxdndistinct, stxddependencies, stxdmcv, stxdinherit FROM pg_statistic_ext s LEFT JOIN pg_statistic_ext_data d ON (d.stxoid = s.oid) WHERE s.stxname = 'ab1_a_b_stats'; ALTER STATISTICS ab1_a_b_stats SET STATISTICS -1; @@ -297,7 +297,7 @@ CREATE STATISTICS s10 ON a, b, c FROM ndistinct; ANALYZE ndistinct; -SELECT s.stxkind, d.stxdndistinct +SELECT s.stxkind, jsonb_pretty(d.stxdndistinct::text::jsonb) AS stxdndistinct FROM pg_statistic_ext s, pg_statistic_ext_data d WHERE s.stxrelid = 'ndistinct'::regclass AND d.stxoid = s.oid; @@ -338,7 +338,7 @@ INSERT INTO ndistinct (a, b, c, filler1) ANALYZE ndistinct; -SELECT s.stxkind, d.stxdndistinct +SELECT s.stxkind, jsonb_pretty(d.stxdndistinct::text::jsonb) AS stxdndistinct FROM pg_statistic_ext s, pg_statistic_ext_data d WHERE s.stxrelid = 'ndistinct'::regclass AND d.stxoid = s.oid; @@ -364,7 +364,7 @@ SELECT * FROM check_estimated_rows('SELECT COUNT(*) FROM ndistinct GROUP BY a, ( DROP STATISTICS s10; -SELECT s.stxkind, d.stxdndistinct +SELECT s.stxkind, jsonb_pretty(d.stxdndistinct::text::jsonb) AS stxdndistinct FROM pg_statistic_ext s, pg_statistic_ext_data d WHERE s.stxrelid = 'ndistinct'::regclass AND d.stxoid = s.oid; @@ -399,7 +399,7 @@ CREATE STATISTICS s10 (ndistinct) ON (a+1), (b+100), (2*c) FROM ndistinct; ANALYZE ndistinct; -SELECT s.stxkind, d.stxdndistinct +SELECT s.stxkind, jsonb_pretty(d.stxdndistinct::text::jsonb) AS stxdndistinct FROM pg_statistic_ext s, pg_statistic_ext_data d WHERE s.stxrelid = 'ndistinct'::regclass AND d.stxoid = s.oid; @@ -423,7 +423,7 @@ CREATE STATISTICS s10 (ndistinct) ON a, b, (2*c) FROM ndistinct; ANALYZE ndistinct; -SELECT s.stxkind, d.stxdndistinct +SELECT s.stxkind, jsonb_pretty(d.stxdndistinct::text::jsonb) AS stxdndistinct FROM pg_statistic_ext s, pg_statistic_ext_data d WHERE s.stxrelid = 'ndistinct'::regclass AND d.stxoid = s.oid; diff --git a/doc/src/sgml/perform.sgml b/doc/src/sgml/perform.sgml index 106583fb2965..e0e9f0468cb8 100644 --- a/doc/src/sgml/perform.sgml +++ b/doc/src/sgml/perform.sgml @@ -1579,9 +1579,39 @@ ANALYZE zipcodes; SELECT stxkeys AS k, stxdndistinct AS nd FROM pg_statistic_ext join pg_statistic_ext_data on (oid = stxoid) WHERE stxname = 'stts2'; --[ RECORD 1 ]------------------------------------------------------&zwsp;-- +-[ RECORD 1 ]------------------- k | 1 2 5 -nd | {"1, 2": 33178, "1, 5": 33178, "2, 5": 27435, "1, 2, 5": 33178} +nd | [ + + | { + + | "ndistinct": 33178,+ + | "attributes": [ + + | 1, + + | 2 + + | ] + + | }, + + | { + + | "ndistinct": 33178,+ + | "attributes": [ + + | 1, + + | 5 + + | ] + + | }, + + | { + + | "ndistinct": 27435,+ + | "attributes": [ + + | 2, + + | 5 + + | ] + + | }, + + | { + + | "ndistinct": 33178,+ + | "attributes": [ + + | 1, + + | 2, + + | 5 + + | ] + + | } + + | ] (1 row) This indicates that there are three combinations of columns that -- 2.51.0