dbcli/sqlglot
1
0
Fork 0
Code Issues Activity

Merging upstream version 19.0.1.

Browse source 
Signed-off-by: Daniel Baumann <daniel@debian.org>
This commit is contained in:
Daniel Baumann 2025-02-13 21:16:09 +01:00
parent 348b067e1b
commit 89acb78953
Signed by: daniel
GPG key ID: FBB4F0E80A80222F
91 changed files with 45416 additions and 43096 deletions
Download patch file Download diff file Expand all files Collapse all files

42
tests/dialects/test_dialect.py
View file

@ -839,10 +839,6 @@ class TestDialect(Validator):
)
self.validate_all(
"STR_TO_DATE(x, '%Y-%m-%dT%H:%M:%S')",
read={
"mysql": "STR_TO_DATE(x, '%Y-%m-%dT%H:%i:%S')",
"starrocks": "STR_TO_DATE(x, '%Y-%m-%dT%H:%i:%S')",
},
write={
"drill": "TO_DATE(x, 'yyyy-MM-dd''T''HH:mm:ss')",
"mysql": "STR_TO_DATE(x, '%Y-%m-%dT%T')",
@ -1861,3 +1857,41 @@ SELECT
"tsql": "SELECT * FROM (SELECT *, COUNT(*) OVER () AS _w FROM t) AS _t WHERE _w > 1",
},
)
def test_nested_ctes(self):
self.validate_all(
"SELECT * FROM (WITH t AS (SELECT 1 AS c) SELECT c FROM t) AS subq",
write={
"bigquery": "SELECT * FROM (WITH t AS (SELECT 1 AS c) SELECT c FROM t) AS subq",
"clickhouse": "SELECT * FROM (WITH t AS (SELECT 1 AS c) SELECT c FROM t) AS subq",
"databricks": "SELECT * FROM (WITH t AS (SELECT 1 AS c) SELECT c FROM t) AS subq",
"duckdb": "SELECT * FROM (WITH t AS (SELECT 1 AS c) SELECT c FROM t) AS subq",
"hive": "WITH t AS (SELECT 1 AS c) SELECT * FROM (SELECT c FROM t) AS subq",
"mysql": "SELECT * FROM (WITH t AS (SELECT 1 AS c) SELECT c FROM t) AS subq",
"postgres": "SELECT * FROM (WITH t AS (SELECT 1 AS c) SELECT c FROM t) AS subq",
"presto": "SELECT * FROM (WITH t AS (SELECT 1 AS c) SELECT c FROM t) AS subq",
"redshift": "SELECT * FROM (WITH t AS (SELECT 1 AS c) SELECT c FROM t) AS subq",
"snowflake": "SELECT * FROM (WITH t AS (SELECT 1 AS c) SELECT c FROM t) AS subq",
"spark": "SELECT * FROM (WITH t AS (SELECT 1 AS c) SELECT c FROM t) AS subq",
"spark2": "WITH t AS (SELECT 1 AS c) SELECT * FROM (SELECT c FROM t) AS subq",
"sqlite": "SELECT * FROM (WITH t AS (SELECT 1 AS c) SELECT c FROM t) AS subq",
"trino": "SELECT * FROM (WITH t AS (SELECT 1 AS c) SELECT c FROM t) AS subq",
"tsql": "WITH t AS (SELECT 1 AS c) SELECT * FROM (SELECT c FROM t) AS subq",
},
)
self.validate_all(
"SELECT * FROM (SELECT * FROM (WITH t AS (SELECT 1 AS c) SELECT c FROM t) AS subq1) AS subq2",
write={
"bigquery": "SELECT * FROM (SELECT * FROM (WITH t AS (SELECT 1 AS c) SELECT c FROM t) AS subq1) AS subq2",
"duckdb": "SELECT * FROM (SELECT * FROM (WITH t AS (SELECT 1 AS c) SELECT c FROM t) AS subq1) AS subq2",
"hive": "WITH t AS (SELECT 1 AS c) SELECT * FROM (SELECT * FROM (SELECT c FROM t) AS subq1) AS subq2",
"tsql": "WITH t AS (SELECT 1 AS c) SELECT * FROM (SELECT * FROM (SELECT c FROM t) AS subq1) AS subq2",
},
)
self.validate_all(
"WITH t1(x) AS (SELECT 1) SELECT * FROM (WITH t2(y) AS (SELECT 2) SELECT y FROM t2) AS subq",
write={
"duckdb": "WITH t1(x) AS (SELECT 1) SELECT * FROM (WITH t2(y) AS (SELECT 2) SELECT y FROM t2) AS subq",
"tsql": "WITH t1(x) AS (SELECT 1), t2(y) AS (SELECT 2) SELECT * FROM (SELECT y FROM t2) AS subq",
},
)

15
tests/dialects/test_hive.py
View file

@ -429,6 +429,21 @@ class TestHive(Validator):
self.validate_identity(
"SELECT key, value, GROUPING__ID, COUNT(*) FROM T1 GROUP BY key, value WITH ROLLUP"
)
self.validate_all(
"SET hiveconf:some_var = 5",
write={
"hive": "SET hiveconf:some_var = 5",
"spark": "SET hiveconf:some_var = 5",
},
)
self.validate_all(
"SELECT ${hiveconf:some_var}",
write={
"hive": "SELECT ${hiveconf:some_var}",
"spark": "SELECT ${hiveconf:some_var}",
},
)
self.validate_all(
"SELECT A.1a AS b FROM test_a AS A",
write={

14
tests/dialects/test_mysql.py
View file

@ -564,8 +564,22 @@ class TestMySQL(Validator):
"STR_TO_DATE(x, '%M')",
read={"": "TS_OR_DS_TO_DATE(x, '%B')"},
)
self.validate_all(
"STR_TO_DATE(x, '%Y-%m-%d')",
write={"presto": "CAST(DATE_PARSE(x, '%Y-%m-%d') AS DATE)"},
)
self.validate_all(
"STR_TO_DATE(x, '%Y-%m-%dT%T')", write={"presto": "DATE_PARSE(x, '%Y-%m-%dT%T')"}
)
def test_mysql(self):
self.validate_all(
# MySQL doesn't support FULL OUTER joins
"SELECT * FROM t1 LEFT OUTER JOIN t2 ON t1.x = t2.x UNION SELECT * FROM t1 RIGHT OUTER JOIN t2 ON t1.x = t2.x",
read={
"postgres": "SELECT * FROM t1 FULL OUTER JOIN t2 ON t1.x = t2.x",
},
)
self.validate_all(
"a XOR b",
read={

45
tests/dialects/test_postgres.py
View file

@ -9,22 +9,39 @@ class TestPostgres(Validator):
dialect = "postgres"
def test_ddl(self):
self.validate_identity(
"CREATE INDEX foo ON bar.baz USING btree(col1 varchar_pattern_ops ASC, col2)"
)
self.validate_identity(
"CREATE TABLE test (x TIMESTAMP WITHOUT TIME ZONE[][])",
"CREATE TABLE test (x TIMESTAMP[][])",
)
self.validate_identity("CREATE INDEX idx_x ON x USING BTREE(x, y) WHERE (NOT y IS NULL)")
self.validate_identity("CREATE TABLE test (elems JSONB[])")
self.validate_identity("CREATE TABLE public.y (x TSTZRANGE NOT NULL)")
self.validate_identity("CREATE TABLE test (foo HSTORE)")
self.validate_identity("CREATE TABLE test (foo JSONB)")
self.validate_identity("CREATE TABLE test (foo VARCHAR(64)[])")
self.validate_identity("CREATE TABLE test (foo INT) PARTITION BY HASH(foo)")
self.validate_identity("INSERT INTO x VALUES (1, 'a', 2.0) RETURNING a")
self.validate_identity("INSERT INTO x VALUES (1, 'a', 2.0) RETURNING a, b")
self.validate_identity("INSERT INTO x VALUES (1, 'a', 2.0) RETURNING *")
self.validate_identity("UPDATE tbl_name SET foo = 123 RETURNING a")
self.validate_identity("CREATE TABLE cities_partdef PARTITION OF cities DEFAULT")
self.validate_identity(
"CREATE TABLE cust_part3 PARTITION OF customers FOR VALUES WITH (MODULUS 3, REMAINDER 2)"
)
self.validate_identity(
"CREATE TABLE measurement_y2016m07 PARTITION OF measurement (unitsales DEFAULT 0) FOR VALUES FROM ('2016-07-01') TO ('2016-08-01')"
)
self.validate_identity(
"CREATE TABLE measurement_ym_older PARTITION OF measurement_year_month FOR VALUES FROM (MINVALUE, MINVALUE) TO (2016, 11)"
)
self.validate_identity(
"CREATE TABLE measurement_ym_y2016m11 PARTITION OF measurement_year_month FOR VALUES FROM (2016, 11) TO (2016, 12)"
)
self.validate_identity(
"CREATE TABLE cities_ab PARTITION OF cities (CONSTRAINT city_id_nonzero CHECK (city_id <> 0)) FOR VALUES IN ('a', 'b')"
)
self.validate_identity(
"CREATE TABLE cities_ab PARTITION OF cities (CONSTRAINT city_id_nonzero CHECK (city_id <> 0)) FOR VALUES IN ('a', 'b') PARTITION BY RANGE(population)"
)
self.validate_identity(
"CREATE INDEX foo ON bar.baz USING btree(col1 varchar_pattern_ops ASC, col2)"
)
self.validate_identity(
"INSERT INTO x VALUES (1, 'a', 2.0) ON CONFLICT (id) DO NOTHING RETURNING *"
)
@ -43,7 +60,10 @@ class TestPostgres(Validator):
self.validate_identity(
"DELETE FROM event USING sales AS s WHERE event.eventid = s.eventid RETURNING a"
)
self.validate_identity("UPDATE tbl_name SET foo = 123 RETURNING a")
self.validate_identity(
"CREATE TABLE test (x TIMESTAMP WITHOUT TIME ZONE[][])",
"CREATE TABLE test (x TIMESTAMP[][])",
)
self.validate_all(
"CREATE OR REPLACE FUNCTION function_name (input_a character varying DEFAULT NULL::character varying)",
@ -187,6 +207,15 @@ class TestPostgres(Validator):
self.validate_identity(
"""ALTER TABLE ONLY "Album" ADD CONSTRAINT "FK_AlbumArtistId" FOREIGN KEY ("ArtistId") REFERENCES "Artist" ("ArtistId") ON DELETE RESTRICT"""
)
self.validate_identity(
"SELECT * FROM JSON_ARRAY_ELEMENTS('[1,true, [2,false]]') WITH ORDINALITY"
)
self.validate_identity(
"SELECT * FROM JSON_ARRAY_ELEMENTS('[1,true, [2,false]]') WITH ORDINALITY AS kv_json"
)
self.validate_identity(
"SELECT * FROM JSON_ARRAY_ELEMENTS('[1,true, [2,false]]') WITH ORDINALITY AS kv_json(a, b)"
)
self.validate_identity("x @@ y")
self.validate_identity("CAST(x AS MONEY)")
self.validate_identity("CAST(x AS INT4RANGE)")

16
tests/dialects/test_redshift.py
View file

@ -6,10 +6,26 @@ class TestRedshift(Validator):
dialect = "redshift"
def test_redshift(self):
self.validate_identity(
"SELECT DATE_DIFF('month', CAST('2020-02-29 00:00:00' AS TIMESTAMP), CAST('2020-03-02 00:00:00' AS TIMESTAMP))",
"SELECT DATEDIFF(month, CAST(CAST('2020-02-29 00:00:00' AS TIMESTAMP) AS DATE), CAST(CAST('2020-03-02 00:00:00' AS TIMESTAMP) AS DATE))",
)
self.validate_identity(
"SELECT * FROM x WHERE y = DATEADD('month', -1, DATE_TRUNC('month', (SELECT y FROM #temp_table)))",
"SELECT * FROM x WHERE y = DATEADD(month, -1, CAST(DATE_TRUNC('month', (SELECT y FROM #temp_table)) AS DATE))",
)
self.validate_all(
"LISTAGG(sellerid, ', ')",
read={
"duckdb": "STRING_AGG(sellerid, ', ')",
},
write={
# GROUP_CONCAT and STRING_AGG are aliases in DuckDB
"duckdb": "GROUP_CONCAT(sellerid, ', ')",
"redshift": "LISTAGG(sellerid, ', ')",
},
)
self.validate_all(
"SELECT APPROXIMATE COUNT(DISTINCT y)",
read={

5
tests/dialects/test_snowflake.py
View file

@ -13,6 +13,7 @@ class TestSnowflake(Validator):
expr.selects[0].assert_is(exp.AggFunc)
self.assertEqual(expr.sql(dialect="snowflake"), "SELECT APPROX_TOP_K(C4, 3, 5) FROM t")
self.validate_identity("SELECT OBJECT_CONSTRUCT()")
self.validate_identity("SELECT DAYOFMONTH(CURRENT_TIMESTAMP())")
self.validate_identity("SELECT DAYOFYEAR(CURRENT_TIMESTAMP())")
self.validate_identity("LISTAGG(data['some_field'], ',')")
@ -77,6 +78,10 @@ class TestSnowflake(Validator):
"SELECT {fn CEILING(5.3)}",
"SELECT CEIL(5.3)",
)
self.validate_identity(
"SELECT TO_TIMESTAMP(x) FROM t",
"SELECT CAST(x AS TIMESTAMPNTZ) FROM t",
)
self.validate_all("CAST(x AS BYTEINT)", write={"snowflake": "CAST(x AS INT)"})
self.validate_all("CAST(x AS CHAR VARYING)", write={"snowflake": "CAST(x AS VARCHAR)"})

5
tests/dialects/test_spark.py
View file

@ -592,9 +592,10 @@ TBLPROPERTIES (
self.validate_all(
"INSERT OVERWRITE TABLE table WITH cte AS (SELECT cola FROM other_table) SELECT cola FROM cte",
write={
"spark": "WITH cte AS (SELECT cola FROM other_table) INSERT OVERWRITE TABLE table SELECT cola FROM cte",
"databricks": "INSERT OVERWRITE TABLE table WITH cte AS (SELECT cola FROM other_table) SELECT cola FROM cte",
"hive": "WITH cte AS (SELECT cola FROM other_table) INSERT OVERWRITE TABLE table SELECT cola FROM cte",
"spark": "INSERT OVERWRITE TABLE table WITH cte AS (SELECT cola FROM other_table) SELECT cola FROM cte",
"spark2": "WITH cte AS (SELECT cola FROM other_table) INSERT OVERWRITE TABLE table SELECT cola FROM cte",
"databricks": "WITH cte AS (SELECT cola FROM other_table) INSERT OVERWRITE TABLE table SELECT cola FROM cte",
},
)

26
tests/dialects/test_teradata.py
View file

@ -148,18 +148,21 @@ class TestTeradata(Validator):
self.validate_all("a MOD b", write={"teradata": "a MOD b", "mysql": "a % b"})
def test_abbrev(self):
self.validate_all("a LT b", write={"teradata": "a < b"})
self.validate_all("a LE b", write={"teradata": "a <= b"})
self.validate_all("a GT b", write={"teradata": "a > b"})
self.validate_all("a GE b", write={"teradata": "a >= b"})
self.validate_all("a ^= b", write={"teradata": "a <> b"})
self.validate_all("a NE b", write={"teradata": "a <> b"})
self.validate_all("a NOT= b", write={"teradata": "a <> b"})
self.validate_all(
"SEL a FROM b",
write={"teradata": "SELECT a FROM b"},
self.validate_identity("a LT b", "a < b")
self.validate_identity("a LE b", "a <= b")
self.validate_identity("a GT b", "a > b")
self.validate_identity("a GE b", "a >= b")
self.validate_identity("a ^= b", "a <> b")
self.validate_identity("a NE b", "a <> b")
self.validate_identity("a NOT= b", "a <> b")
self.validate_identity("a EQ b", "a = b")
self.validate_identity("SEL a FROM b", "SELECT a FROM b")
self.validate_identity(
"SELECT col1, col2 FROM dbc.table1 WHERE col1 EQ 'value1' MINUS SELECT col1, col2 FROM dbc.table2",
"SELECT col1, col2 FROM dbc.table1 WHERE col1 = 'value1' EXCEPT SELECT col1, col2 FROM dbc.table2",
)
self.validate_identity("UPD a SET b = 1", "UPDATE a SET b = 1")
self.validate_identity("DEL FROM a", "DELETE FROM a")
def test_datatype(self):
self.validate_all(
@ -187,3 +190,4 @@ class TestTeradata(Validator):
"": "STR_TO_DATE('1992-01', '%Y-%d')",
},
)
self.validate_identity("CAST('1992-01' AS FORMAT 'YYYY-DD')")

44
tests/dialects/test_tsql.py
View file

@ -6,6 +6,36 @@ class TestTSQL(Validator):
dialect = "tsql"
def test_tsql(self):
self.validate_all(
"WITH t(c) AS (SELECT 1) SELECT * INTO foo FROM (SELECT c FROM t) AS temp",
read={
"duckdb": "CREATE TABLE foo AS WITH t(c) AS (SELECT 1) SELECT c FROM t",
},
)
self.validate_all(
"WITH y AS (SELECT 2 AS c) INSERT INTO t SELECT * FROM y",
read={
"duckdb": "WITH y AS (SELECT 2 AS c) INSERT INTO t SELECT * FROM y",
},
)
self.validate_all(
"WITH t(c) AS (SELECT 1) SELECT 1 AS c UNION (SELECT c FROM t)",
read={
"duckdb": "SELECT 1 AS c UNION (WITH t(c) AS (SELECT 1) SELECT c FROM t)",
},
)
self.validate_all(
"WITH t(c) AS (SELECT 1) MERGE INTO x AS z USING (SELECT c FROM t) AS y ON a = b WHEN MATCHED THEN UPDATE SET a = y.b",
read={
"postgres": "MERGE INTO x AS z USING (WITH t(c) AS (SELECT 1) SELECT c FROM t) AS y ON a = b WHEN MATCHED THEN UPDATE SET a = y.b",
},
)
self.validate_all(
"WITH t(n) AS (SELECT 1 AS n UNION ALL SELECT n + 1 AS n FROM t WHERE n < 4) SELECT * FROM (SELECT SUM(n) AS s4 FROM t) AS subq",
read={
"duckdb": "SELECT * FROM (WITH RECURSIVE t(n) AS (SELECT 1 AS n UNION ALL SELECT n + 1 AS n FROM t WHERE n < 4) SELECT SUM(n) AS s4 FROM t) AS subq",
},
)
self.validate_all(
"CREATE TABLE #mytemptable (a INTEGER)",
read={
@ -825,12 +855,18 @@ WHERE
def test_datename(self):
self.validate_all(
"SELECT DATENAME(mm,'1970-01-01')",
write={"spark": "SELECT DATE_FORMAT(CAST('1970-01-01' AS TIMESTAMP), 'MMMM')"},
"SELECT DATENAME(mm, '1970-01-01')",
write={
"spark": "SELECT DATE_FORMAT(CAST('1970-01-01' AS TIMESTAMP), 'MMMM')",
"tsql": "SELECT FORMAT(CAST('1970-01-01' AS DATETIME2), 'MMMM')",
},
)
self.validate_all(
"SELECT DATENAME(dw,'1970-01-01')",
write={"spark": "SELECT DATE_FORMAT(CAST('1970-01-01' AS TIMESTAMP), 'EEEE')"},
"SELECT DATENAME(dw, '1970-01-01')",
write={
"spark": "SELECT DATE_FORMAT(CAST('1970-01-01' AS TIMESTAMP), 'EEEE')",
"tsql": "SELECT FORMAT(CAST('1970-01-01' AS DATETIME2), 'dddd')",
},
)
def test_datepart(self):

2
tests/fixtures/optimizer/normalize.sql vendored
View file

@ -44,4 +44,4 @@ SELECT * FROM x WHERE (A AND B) OR C;
SELECT * FROM x WHERE (A OR C) AND (B OR C);
dt2 between '2022-01-01 12:00:00' and '2022-12-31' and dt2 >= '2022-05-01 12:00:00' or dt2 = '2021-06-01 12:00:00';
(dt2 <= '2022-12-31' OR dt2 = '2021-06-01 12:00:00') AND (dt2 = '2021-06-01 12:00:00' OR dt2 >= '2022-01-01 12:00:00') AND (dt2 = '2021-06-01 12:00:00' OR dt2 >= '2022-05-01 12:00:00')
(dt2 <= '2022-12-31' OR dt2 = '2021-06-01 12:00:00') AND (dt2 = '2021-06-01 12:00:00' OR dt2 >= '2022-01-01 12:00:00') AND (dt2 = '2021-06-01 12:00:00' OR dt2 >= '2022-05-01 12:00:00');

3
tests/fixtures/optimizer/normalize_identifiers.sql vendored
View file

@ -70,3 +70,6 @@ SELECT a /* sqlglot.meta case_sensitive */, B FROM table /* sqlglot.meta case_se
# dialect: redshift
SELECT COALESCE(json_val.a /* sqlglot.meta case_sensitive */, json_val.A /* sqlglot.meta case_sensitive */) FROM table;
SELECT COALESCE(json_val.a /* sqlglot.meta case_sensitive */, json_val.A /* sqlglot.meta case_sensitive */) FROM table;
SELECT @X;
SELECT @X;

4
tests/fixtures/optimizer/optimizer.sql vendored
View file

@ -480,8 +480,8 @@ JOIN "company_table" AS "company_table_2"
LEFT JOIN "unlocked" AS "unlocked"
ON "company_table_2"."id" = "unlocked"."company_id"
WHERE
CASE WHEN "unlocked"."company_id" IS NULL THEN 0 ELSE 1 END = FALSE
AND NOT "company_table_2"."id" IS NULL;
NOT "company_table_2"."id" IS NULL
AND CASE WHEN "unlocked"."company_id" IS NULL THEN 0 ELSE 1 END = FALSE;
# title: db.table alias clash
# execute: false

8
tests/fixtures/optimizer/pushdown_predicates.sql vendored
View file

@ -11,7 +11,7 @@ SELECT x.a FROM (SELECT * FROM x) AS x JOIN y WHERE (x.a = y.a AND x.a = 1 AND x
SELECT x.a FROM (SELECT * FROM x) AS x JOIN y ON x.a = y.a WHERE TRUE;
SELECT x.a FROM (SELECT * FROM x) AS x JOIN y WHERE (x.a = y.a AND x.a = 1 AND x.b = 1) OR x.a = y.b;
SELECT x.a FROM (SELECT * FROM x) AS x JOIN y ON x.a = y.a OR x.a = y.b WHERE (x.a = y.a AND x.a = 1 AND x.b = 1) OR x.a = y.b;
SELECT x.a FROM (SELECT * FROM x) AS x JOIN y ON x.a = y.a OR x.a = y.b WHERE (x.a = 1 AND x.a = y.a AND x.b = 1) OR x.a = y.b;
SELECT x.a FROM (SELECT x.a AS a, x.b * 1 AS c FROM x) AS x WHERE x.c = 1;
SELECT x.a FROM (SELECT x.a AS a, x.b * 1 AS c FROM x WHERE x.b * 1 = 1) AS x WHERE TRUE;
@ -26,10 +26,10 @@ SELECT x.a FROM x AS x JOIN (SELECT y.a FROM y AS y) AS y ON y.a = 1 AND x.a = y
SELECT x.a FROM x AS x JOIN (SELECT y.a FROM y AS y WHERE y.a = 1) AS y ON x.a = y.a AND TRUE;
SELECT x.a AS a FROM x AS x JOIN (SELECT * FROM y AS y) AS y ON y.a = 1 WHERE x.a = 1 AND x.b = 1 AND y.a = x.a;
SELECT x.a AS a FROM x AS x JOIN (SELECT * FROM y AS y WHERE y.a = 1) AS y ON y.a = x.a AND TRUE WHERE x.a = 1 AND x.b = 1 AND TRUE;
SELECT x.a AS a FROM x AS x JOIN (SELECT * FROM y AS y WHERE y.a = 1) AS y ON TRUE AND y.a = x.a WHERE x.a = 1 AND x.b = 1 AND TRUE;
SELECT x.a AS a FROM x AS x CROSS JOIN (SELECT * FROM y AS y) AS y WHERE x.a = 1 AND x.b = 1 AND y.a = x.a AND y.a = 1;
SELECT x.a AS a FROM x AS x JOIN (SELECT * FROM y AS y WHERE y.a = 1) AS y ON y.a = x.a AND TRUE WHERE x.a = 1 AND x.b = 1 AND TRUE AND TRUE;
SELECT x.a AS a FROM x AS x JOIN (SELECT * FROM y AS y WHERE y.a = 1) AS y ON TRUE AND y.a = x.a WHERE x.a = 1 AND x.b = 1 AND TRUE AND TRUE;
with t1 as (SELECT x.a, x.b, ROW_NUMBER() OVER (PARTITION BY x.a ORDER BY x.a) as row_num FROM x) SELECT t1.a, t1.b FROM t1 WHERE row_num = 1;
WITH t1 AS (SELECT x.a, x.b, ROW_NUMBER() OVER (PARTITION BY x.a ORDER BY x.a) AS row_num FROM x) SELECT t1.a, t1.b FROM t1 WHERE row_num = 1;
@ -43,4 +43,4 @@ SELECT x.cnt AS cnt FROM (SELECT COUNT(1) AS cnt FROM x AS x HAVING COUNT(1) > 0
-- Pushdown predicate to HAVING (DNF)
SELECT x.cnt AS cnt FROM (SELECT COUNT(1) AS cnt, COUNT(x.a) AS cnt_a, COUNT(x.b) AS cnt_b FROM x AS x) AS x WHERE (x.cnt_a > 0 AND x.cnt_b > 0) OR x.cnt > 0;
SELECT x.cnt AS cnt FROM (SELECT COUNT(1) AS cnt, COUNT(x.a) AS cnt_a, COUNT(x.b) AS cnt_b FROM x AS x HAVING COUNT(1) > 0 OR (COUNT(x.a) > 0 AND COUNT(x.b) > 0)) AS x WHERE x.cnt > 0 OR (x.cnt_a > 0 AND x.cnt_b > 0);
SELECT x.cnt AS cnt FROM (SELECT COUNT(1) AS cnt, COUNT(x.a) AS cnt_a, COUNT(x.b) AS cnt_b FROM x AS x HAVING COUNT(1) > 0 OR (COUNT(x.a) > 0 AND COUNT(x.b) > 0)) AS x WHERE x.cnt > 0 OR (x.cnt_a > 0 AND x.cnt_b > 0);

4
tests/fixtures/optimizer/qualify_tables.sql vendored
View file

@ -117,3 +117,7 @@ SELECT x FROM c.db.t AS t, LATERAL UNNEST(t.xs) AS x;
# title: lateral unnest without alias
SELECT x FROM t, LATERAL UNNEST(t.xs);
SELECT x FROM c.db.t AS t, LATERAL UNNEST(t.xs) AS _q_0;
# title: table with ordinality
SELECT * FROM t CROSS JOIN JSON_ARRAY_ELEMENTS(t.response) WITH ORDINALITY AS kv_json;
SELECT * FROM c.db.t AS t CROSS JOIN JSON_ARRAY_ELEMENTS(t.response) WITH ORDINALITY AS kv_json;

3
tests/fixtures/optimizer/quote_identities.sql vendored
View file

@ -6,3 +6,6 @@ SELECT "a" FROM "x";
SELECT x.a AS a FROM db.x;
SELECT "x"."a" AS "a" FROM "db"."x";
SELECT @x;
SELECT @x;

20
tests/fixtures/optimizer/simplify.sql vendored
View file

@ -631,19 +631,19 @@ COALESCE(x);
x;
COALESCE(x, 1) = 2;
x = 2 AND NOT x IS NULL;
NOT x IS NULL AND x = 2;
2 = COALESCE(x, 1);
2 = x AND NOT x IS NULL;
COALESCE(x, 1, 1) = 1 + 1;
x = 2 AND NOT x IS NULL;
NOT x IS NULL AND x = 2;
COALESCE(x, 1, 2) = 2;
x = 2 AND NOT x IS NULL;
NOT x IS NULL AND x = 2;
COALESCE(x, 3) <= 2;
x <= 2 AND NOT x IS NULL;
NOT x IS NULL AND x <= 2;
COALESCE(x, 1) <> 2;
x <> 2 OR x IS NULL;
@ -872,7 +872,7 @@ x = 5 AND y = x;
x = 5 AND y = 5;
5 = x AND y = x;
y = 5 AND 5 = x;
5 = x AND y = 5;
x = 5 OR y = x;
x = 5 OR y = x;
@ -899,13 +899,13 @@ x = 5 AND x + 3 = 8;
x = 5;
x = 5 AND (SELECT x FROM t WHERE y = 1);
x = 5 AND (SELECT x FROM t WHERE y = 1);
(SELECT x FROM t WHERE y = 1) AND x = 5;
x = 1 AND y > 0 AND (SELECT z = 5 FROM t WHERE y = 1);
x = 1 AND y > 0 AND (SELECT z = 5 FROM t WHERE y = 1);
(SELECT z = 5 FROM t WHERE y = 1) AND x = 1 AND y > 0;
x = 1 AND x = y AND (SELECT z FROM t WHERE a AND (b OR c));
x = 1 AND (SELECT z FROM t WHERE a AND (b OR c)) AND 1 = y;
(SELECT z FROM t WHERE a AND (b OR c)) AND 1 = y AND x = 1;
t1.a = 39 AND t2.b = t1.a AND t3.c = t2.b;
t1.a = 39 AND t2.b = 39 AND t3.c = 39;
@ -920,10 +920,10 @@ x = 1 AND CASE x WHEN 5 THEN FALSE ELSE TRUE END;
x = 1;
x = y AND CASE WHEN x = 5 THEN FALSE ELSE TRUE END;
x = y AND CASE WHEN x = 5 THEN FALSE ELSE TRUE END;
CASE WHEN x = 5 THEN FALSE ELSE TRUE END AND x = y;
x = 1 AND CASE WHEN y = 5 THEN x = z END;
x = 1 AND CASE WHEN y = 5 THEN 1 = z END;
CASE WHEN y = 5 THEN 1 = z END AND x = 1;
--------------------------------------
-- Simplify Conditionals

4
tests/fixtures/optimizer/tpc-ds/tpc-ds.sql vendored
View file

@ -2547,8 +2547,8 @@ JOIN "date_dim" AS "date_dim"
)
WHERE
"_u_3"."_u_4" IS NULL
AND ARRAY_ANY("_u_0"."_u_2", "_x" -> "cs1"."cs_warehouse_sk" <> "_x")
AND NOT "_u_0"."_u_1" IS NULL
AND ARRAY_ANY("_u_0"."_u_2", "_x" -> "cs1"."cs_warehouse_sk" <> "_x")
ORDER BY
COUNT(DISTINCT "cs1"."cs_order_number")
LIMIT 100;
@ -12586,8 +12586,8 @@ JOIN "web_site" AS "web_site"
AND "ws1"."ws_web_site_sk" = "web_site"."web_site_sk"
WHERE
"_u_3"."_u_4" IS NULL
AND ARRAY_ANY("_u_0"."_u_2", "_x" -> "ws1"."ws_warehouse_sk" <> "_x")
AND NOT "_u_0"."_u_1" IS NULL
AND ARRAY_ANY("_u_0"."_u_2", "_x" -> "ws1"."ws_warehouse_sk" <> "_x")
ORDER BY
COUNT(DISTINCT "ws1"."ws_order_number")
LIMIT 100;

6
tests/fixtures/optimizer/tpc-h/tpc-h.sql vendored
View file

@ -249,9 +249,9 @@ FROM "orders" AS "orders"
LEFT JOIN "_u_0" AS "_u_0"
ON "_u_0"."l_orderkey" = "orders"."o_orderkey"
WHERE
CAST("orders"."o_orderdate" AS DATE) < CAST('1993-10-01' AS DATE)
NOT "_u_0"."l_orderkey" IS NULL
AND CAST("orders"."o_orderdate" AS DATE) < CAST('1993-10-01' AS DATE)
AND CAST("orders"."o_orderdate" AS DATE) >= CAST('1993-07-01' AS DATE)
AND NOT "_u_0"."l_orderkey" IS NULL
GROUP BY
"orders"."o_orderpriority"
ORDER BY
@ -1348,8 +1348,8 @@ WHERE
"_u_2"."l_orderkey" IS NULL
OR NOT ARRAY_ANY("_u_2"."_u_3", "_x" -> "_x" <> "l1"."l_suppkey")
)
AND ARRAY_ANY("_u_0"."_u_1", "_x" -> "_x" <> "l1"."l_suppkey")
AND NOT "_u_0"."l_orderkey" IS NULL
AND ARRAY_ANY("_u_0"."_u_1", "_x" -> "_x" <> "l1"."l_suppkey")
GROUP BY
"supplier"."s_name"
ORDER BY

2
tests/test_expressions.py
View file

@ -443,7 +443,7 @@ class TestExpressions(unittest.TestCase):
return None
return node
self.assertEqual(expression.transform(remove_non_list_arg).sql(), "CAST(x AS )")
self.assertEqual(expression.transform(remove_non_list_arg).sql(), "CAST(x AS)")
expression = parse_one("SELECT a, b FROM x")

18
tests/test_optimizer.py
View file

@ -246,6 +246,15 @@ class TestOptimizer(unittest.TestCase):
"CREATE FUNCTION `udfs`.`myTest`(`x` FLOAT64) AS (1)",
)
self.assertEqual(
optimizer.qualify.qualify(
parse_one("SELECT `bar_bazfoo_$id` FROM test", read="spark"),
schema={"test": {"bar_bazFoo_$id": "BIGINT"}},
dialect="spark",
).sql(dialect="spark"),
"SELECT `test`.`bar_bazfoo_$id` AS `bar_bazfoo_$id` FROM `test` AS `test`",
)
self.check_file(
"qualify_columns", qualify_columns, execute=True, schema=self.schema, set_dialect=True
)
@ -271,6 +280,8 @@ class TestOptimizer(unittest.TestCase):
set_dialect=True,
)
self.assertEqual(optimizer.normalize_identifiers.normalize_identifiers("a%").sql(), '"a%"')
def test_pushdown_projection(self):
self.check_file("pushdown_projections", pushdown_projections, schema=self.schema)
@ -978,3 +989,10 @@ FROM READ_CSV('tests/fixtures/optimizer/tpc-h/nation.csv.gz', 'delimiter', '|')
query = parse_one("select a.b:c from d", read="snowflake")
qualified = optimizer.qualify.qualify(query)
self.assertEqual(qualified.expressions[0].alias, "c")
def test_qualify_tables_no_schema(self):
query = parse_one("select a from b")
self.assertEqual(
optimizer.qualify_tables.qualify_tables(query, catalog="catalog").sql(),
"SELECT a FROM b AS b",
)