summaryrefslogtreecommitdiffstats
path: root/tests
diff options
context:
space:
mode:
authorDaniel Baumann <daniel.baumann@progress-linux.org>2023-04-07 12:35:01 +0000
committerDaniel Baumann <daniel.baumann@progress-linux.org>2023-04-07 12:35:01 +0000
commit58c11f70074708344e433080e47621091a6dcd84 (patch)
tree2589166e0e58be4947e07a956d26efa497bccaf2 /tests
parentAdding upstream version 11.4.5. (diff)
downloadsqlglot-58c11f70074708344e433080e47621091a6dcd84.tar.xz
sqlglot-58c11f70074708344e433080e47621091a6dcd84.zip
Adding upstream version 11.5.2.upstream/11.5.2
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to '')
-rw-r--r--tests/dialects/test_clickhouse.py1
-rw-r--r--tests/dialects/test_dialect.py54
-rw-r--r--tests/dialects/test_duckdb.py23
-rw-r--r--tests/dialects/test_hive.py8
-rw-r--r--tests/dialects/test_mysql.py35
-rw-r--r--tests/dialects/test_presto.py10
-rw-r--r--tests/dialects/test_redshift.py11
-rw-r--r--tests/dialects/test_snowflake.py15
-rw-r--r--tests/dialects/test_spark.py10
-rw-r--r--tests/dialects/test_sqlite.py2
-rw-r--r--tests/dialects/test_tsql.py14
-rw-r--r--tests/fixtures/identity.sql3
-rw-r--r--tests/fixtures/optimizer/optimizer.sql1
-rw-r--r--tests/fixtures/optimizer/qualify_columns.sql22
-rw-r--r--tests/fixtures/optimizer/simplify.sql16
-rw-r--r--tests/fixtures/optimizer/tpc-ds/tpc-ds.sql16
-rw-r--r--tests/test_executor.py23
-rw-r--r--tests/test_expressions.py40
-rw-r--r--tests/test_optimizer.py9
-rw-r--r--tests/test_tokens.py1
20 files changed, 255 insertions, 59 deletions
diff --git a/tests/dialects/test_clickhouse.py b/tests/dialects/test_clickhouse.py
index d206bb1..40a3a04 100644
--- a/tests/dialects/test_clickhouse.py
+++ b/tests/dialects/test_clickhouse.py
@@ -5,6 +5,7 @@ class TestClickhouse(Validator):
dialect = "clickhouse"
def test_clickhouse(self):
+ self.validate_identity("SELECT match('abc', '([a-z]+)')")
self.validate_identity("dictGet(x, 'y')")
self.validate_identity("SELECT * FROM x FINAL")
self.validate_identity("SELECT * FROM x AS y FINAL")
diff --git a/tests/dialects/test_dialect.py b/tests/dialects/test_dialect.py
index 0805e9c..3558d62 100644
--- a/tests/dialects/test_dialect.py
+++ b/tests/dialects/test_dialect.py
@@ -290,6 +290,60 @@ class TestDialect(Validator):
read={"postgres": "INET '127.0.0.1/32'"},
)
+ def test_decode(self):
+ self.validate_identity("DECODE(bin, charset)")
+
+ self.validate_all(
+ "SELECT DECODE(a, 1, 'one')",
+ write={
+ "": "SELECT CASE WHEN a = 1 THEN 'one' END",
+ "oracle": "SELECT CASE WHEN a = 1 THEN 'one' END",
+ "redshift": "SELECT CASE WHEN a = 1 THEN 'one' END",
+ "snowflake": "SELECT CASE WHEN a = 1 THEN 'one' END",
+ "spark": "SELECT CASE WHEN a = 1 THEN 'one' END",
+ },
+ )
+ self.validate_all(
+ "SELECT DECODE(a, 1, 'one', 'default')",
+ write={
+ "": "SELECT CASE WHEN a = 1 THEN 'one' ELSE 'default' END",
+ "oracle": "SELECT CASE WHEN a = 1 THEN 'one' ELSE 'default' END",
+ "redshift": "SELECT CASE WHEN a = 1 THEN 'one' ELSE 'default' END",
+ "snowflake": "SELECT CASE WHEN a = 1 THEN 'one' ELSE 'default' END",
+ "spark": "SELECT CASE WHEN a = 1 THEN 'one' ELSE 'default' END",
+ },
+ )
+ self.validate_all(
+ "SELECT DECODE(a, NULL, 'null')",
+ write={
+ "": "SELECT CASE WHEN a IS NULL THEN 'null' END",
+ "oracle": "SELECT CASE WHEN a IS NULL THEN 'null' END",
+ "redshift": "SELECT CASE WHEN a IS NULL THEN 'null' END",
+ "snowflake": "SELECT CASE WHEN a IS NULL THEN 'null' END",
+ "spark": "SELECT CASE WHEN a IS NULL THEN 'null' END",
+ },
+ )
+ self.validate_all(
+ "SELECT DECODE(a, b, c)",
+ write={
+ "": "SELECT CASE WHEN a = b OR (a IS NULL AND b IS NULL) THEN c END",
+ "oracle": "SELECT CASE WHEN a = b OR (a IS NULL AND b IS NULL) THEN c END",
+ "redshift": "SELECT CASE WHEN a = b OR (a IS NULL AND b IS NULL) THEN c END",
+ "snowflake": "SELECT CASE WHEN a = b OR (a IS NULL AND b IS NULL) THEN c END",
+ "spark": "SELECT CASE WHEN a = b OR (a IS NULL AND b IS NULL) THEN c END",
+ },
+ )
+ self.validate_all(
+ "SELECT DECODE(tbl.col, 'some_string', 'foo')",
+ write={
+ "": "SELECT CASE WHEN tbl.col = 'some_string' THEN 'foo' END",
+ "oracle": "SELECT CASE WHEN tbl.col = 'some_string' THEN 'foo' END",
+ "redshift": "SELECT CASE WHEN tbl.col = 'some_string' THEN 'foo' END",
+ "snowflake": "SELECT CASE WHEN tbl.col = 'some_string' THEN 'foo' END",
+ "spark": "SELECT CASE WHEN tbl.col = 'some_string' THEN 'foo' END",
+ },
+ )
+
def test_if_null(self):
self.validate_all(
"SELECT IFNULL(1, NULL) FROM foo",
diff --git a/tests/dialects/test_duckdb.py b/tests/dialects/test_duckdb.py
index a15e6b4..245d82a 100644
--- a/tests/dialects/test_duckdb.py
+++ b/tests/dialects/test_duckdb.py
@@ -74,6 +74,17 @@ class TestDuckDB(Validator):
"spark": "TO_TIMESTAMP(x, 'M/d/yy h:mm a')",
},
)
+ self.validate_all(
+ "CAST(start AS TIMESTAMPTZ) AT TIME ZONE 'America/New_York'",
+ read={
+ "snowflake": "CONVERT_TIMEZONE('America/New_York', CAST(start AS TIMESTAMPTZ))",
+ },
+ write={
+ "bigquery": "TIMESTAMP(DATETIME(CAST(start AS TIMESTAMPTZ), 'America/New_York'))",
+ "duckdb": "CAST(start AS TIMESTAMPTZ) AT TIME ZONE 'America/New_York'",
+ "snowflake": "CONVERT_TIMEZONE('America/New_York', CAST(start AS TIMESTAMPTZ))",
+ },
+ )
def test_sample(self):
self.validate_all(
@@ -421,6 +432,18 @@ class TestDuckDB(Validator):
"snowflake": "CAST(COL AS ARRAY)",
},
)
+ self.validate_all(
+ "CAST([STRUCT_PACK(a := 1)] AS STRUCT(a BIGINT)[])",
+ write={
+ "duckdb": "CAST(LIST_VALUE({'a': 1}) AS STRUCT(a BIGINT)[])",
+ },
+ )
+ self.validate_all(
+ "CAST([[STRUCT_PACK(a := 1)]] AS STRUCT(a BIGINT)[][])",
+ write={
+ "duckdb": "CAST(LIST_VALUE(LIST_VALUE({'a': 1})) AS STRUCT(a BIGINT)[][])",
+ },
+ )
def test_bool_or(self):
self.validate_all(
diff --git a/tests/dialects/test_hive.py b/tests/dialects/test_hive.py
index 0161f1e..1a83575 100644
--- a/tests/dialects/test_hive.py
+++ b/tests/dialects/test_hive.py
@@ -689,3 +689,11 @@ class TestHive(Validator):
self.validate_identity("'\\z'")
self.validate_identity("'\\\z'")
self.validate_identity("'\\\\z'")
+
+ def test_data_type(self):
+ self.validate_all(
+ "CAST(a AS BIT)",
+ write={
+ "hive": "CAST(a AS BOOLEAN)",
+ },
+ )
diff --git a/tests/dialects/test_mysql.py b/tests/dialects/test_mysql.py
index 5059d05..f618728 100644
--- a/tests/dialects/test_mysql.py
+++ b/tests/dialects/test_mysql.py
@@ -204,6 +204,41 @@ class TestMySQL(Validator):
},
)
+ def test_match_against(self):
+ self.validate_all(
+ "MATCH(col1, col2, col3) AGAINST('abc')",
+ read={
+ "": "MATCH(col1, col2, col3) AGAINST('abc')",
+ "mysql": "MATCH(col1, col2, col3) AGAINST('abc')",
+ },
+ write={
+ "": "MATCH(col1, col2, col3) AGAINST('abc')",
+ "mysql": "MATCH(col1, col2, col3) AGAINST('abc')",
+ },
+ )
+ self.validate_all(
+ "MATCH(col1, col2) AGAINST('abc' IN NATURAL LANGUAGE MODE)",
+ write={"mysql": "MATCH(col1, col2) AGAINST('abc' IN NATURAL LANGUAGE MODE)"},
+ )
+ self.validate_all(
+ "MATCH(col1, col2) AGAINST('abc' IN NATURAL LANGUAGE MODE WITH QUERY EXPANSION)",
+ write={
+ "mysql": "MATCH(col1, col2) AGAINST('abc' IN NATURAL LANGUAGE MODE WITH QUERY EXPANSION)"
+ },
+ )
+ self.validate_all(
+ "MATCH(col1, col2) AGAINST('abc' IN BOOLEAN MODE)",
+ write={"mysql": "MATCH(col1, col2) AGAINST('abc' IN BOOLEAN MODE)"},
+ )
+ self.validate_all(
+ "MATCH(col1, col2) AGAINST('abc' WITH QUERY EXPANSION)",
+ write={"mysql": "MATCH(col1, col2) AGAINST('abc' WITH QUERY EXPANSION)"},
+ )
+ self.validate_all(
+ "MATCH(a.b) AGAINST('abc')",
+ write={"mysql": "MATCH(a.b) AGAINST('abc')"},
+ )
+
def test_mysql(self):
self.validate_all(
"SELECT a FROM tbl FOR UPDATE",
diff --git a/tests/dialects/test_presto.py b/tests/dialects/test_presto.py
index 1762e7a..1007899 100644
--- a/tests/dialects/test_presto.py
+++ b/tests/dialects/test_presto.py
@@ -539,6 +539,16 @@ class TestPresto(Validator):
"hive": "SELECT a, b, c, d, SUM(y) FROM z GROUP BY d, GROUPING SETS ((b, c)), CUBE (a), ROLLUP (a)",
},
)
+ self.validate_all(
+ "JSON_FORMAT(x)",
+ read={
+ "spark": "TO_JSON(x)",
+ },
+ write={
+ "presto": "JSON_FORMAT(x)",
+ "spark": "TO_JSON(x)",
+ },
+ )
def test_encode_decode(self):
self.validate_all(
diff --git a/tests/dialects/test_redshift.py b/tests/dialects/test_redshift.py
index ff730f8..0933051 100644
--- a/tests/dialects/test_redshift.py
+++ b/tests/dialects/test_redshift.py
@@ -69,9 +69,11 @@ class TestRedshift(Validator):
self.validate_all(
"DECODE(x, a, b, c, d)",
write={
- "": "MATCHES(x, a, b, c, d)",
- "oracle": "DECODE(x, a, b, c, d)",
- "snowflake": "DECODE(x, a, b, c, d)",
+ "": "CASE WHEN x = a OR (x IS NULL AND a IS NULL) THEN b WHEN x = c OR (x IS NULL AND c IS NULL) THEN d END",
+ "oracle": "CASE WHEN x = a OR (x IS NULL AND a IS NULL) THEN b WHEN x = c OR (x IS NULL AND c IS NULL) THEN d END",
+ "redshift": "CASE WHEN x = a OR (x IS NULL AND a IS NULL) THEN b WHEN x = c OR (x IS NULL AND c IS NULL) THEN d END",
+ "snowflake": "CASE WHEN x = a OR (x IS NULL AND a IS NULL) THEN b WHEN x = c OR (x IS NULL AND c IS NULL) THEN d END",
+ "spark": "CASE WHEN x = a OR (x IS NULL AND a IS NULL) THEN b WHEN x = c OR (x IS NULL AND c IS NULL) THEN d END",
},
)
self.validate_all(
@@ -97,9 +99,6 @@ class TestRedshift(Validator):
)
def test_identity(self):
- self.validate_identity(
- "SELECT DECODE(COL1, 'replace_this', 'with_this', 'replace_that', 'with_that')"
- )
self.validate_identity("CAST('bla' AS SUPER)")
self.validate_identity("CREATE TABLE real1 (realcol REAL)")
self.validate_identity("CAST('foo' AS HLLSKETCH)")
diff --git a/tests/dialects/test_snowflake.py b/tests/dialects/test_snowflake.py
index 940fa50..eb423a5 100644
--- a/tests/dialects/test_snowflake.py
+++ b/tests/dialects/test_snowflake.py
@@ -6,6 +6,11 @@ class TestSnowflake(Validator):
dialect = "snowflake"
def test_snowflake(self):
+ self.validate_identity("SELECT HLL(*)")
+ self.validate_identity("SELECT HLL(a)")
+ self.validate_identity("SELECT HLL(DISTINCT t.a)")
+ self.validate_identity("SELECT HLL(a, b, c)")
+ self.validate_identity("SELECT HLL(DISTINCT a, b, c)")
self.validate_identity("$x")
self.validate_identity("SELECT REGEXP_LIKE(a, b, c)")
self.validate_identity("PUT file:///dir/tmp.csv @%table")
@@ -376,14 +381,10 @@ class TestSnowflake(Validator):
},
)
self.validate_all(
- "DECODE(x, a, b, c, d)",
- read={
- "": "MATCHES(x, a, b, c, d)",
- },
+ "DECODE(x, a, b, c, d, e)",
write={
- "": "MATCHES(x, a, b, c, d)",
- "oracle": "DECODE(x, a, b, c, d)",
- "snowflake": "DECODE(x, a, b, c, d)",
+ "": "CASE WHEN x = a OR (x IS NULL AND a IS NULL) THEN b WHEN x = c OR (x IS NULL AND c IS NULL) THEN d ELSE e END",
+ "snowflake": "CASE WHEN x = a OR (x IS NULL AND a IS NULL) THEN b WHEN x = c OR (x IS NULL AND c IS NULL) THEN d ELSE e END",
},
)
diff --git a/tests/dialects/test_spark.py b/tests/dialects/test_spark.py
index b12f272..0da2931 100644
--- a/tests/dialects/test_spark.py
+++ b/tests/dialects/test_spark.py
@@ -345,3 +345,13 @@ TBLPROPERTIES (
"SELECT a, LOGICAL_OR(b) FROM table GROUP BY a",
write={"spark": "SELECT a, BOOL_OR(b) FROM table GROUP BY a"},
)
+
+ def test_current_user(self):
+ self.validate_all(
+ "CURRENT_USER",
+ write={"spark": "CURRENT_USER()"},
+ )
+ self.validate_all(
+ "CURRENT_USER()",
+ write={"spark": "CURRENT_USER()"},
+ )
diff --git a/tests/dialects/test_sqlite.py b/tests/dialects/test_sqlite.py
index fd9e52b..583d5be 100644
--- a/tests/dialects/test_sqlite.py
+++ b/tests/dialects/test_sqlite.py
@@ -12,7 +12,7 @@ class TestSQLite(Validator):
self.validate_identity("INSERT OR ROLLBACK INTO foo (x, y) VALUES (1, 2)")
self.validate_all(
- "CREATE TABLE foo (id INTEGER PRIMARY KEY ASC)",
+ "CREATE TABLE foo (id INTEGER PRIMARY KEY ASC)",
write={"sqlite": "CREATE TABLE foo (id INTEGER PRIMARY KEY ASC)"},
)
self.validate_all(
diff --git a/tests/dialects/test_tsql.py b/tests/dialects/test_tsql.py
index 60867be..d9ee4ae 100644
--- a/tests/dialects/test_tsql.py
+++ b/tests/dialects/test_tsql.py
@@ -547,11 +547,11 @@ WHERE
def test_string(self):
self.validate_all(
"SELECT N'test'",
- write={"spark": "SELECT N'test'"},
+ write={"spark": "SELECT 'test'"},
)
self.validate_all(
"SELECT n'test'",
- write={"spark": "SELECT N'test'"},
+ write={"spark": "SELECT 'test'"},
)
self.validate_all(
"SELECT '''test'''",
@@ -621,3 +621,13 @@ WHERE
"tsql": """SELECT "x" FROM "a"."b" FOR SYSTEM_TIME ALL AS alias""",
},
)
+
+ def test_current_user(self):
+ self.validate_all(
+ "SUSER_NAME()",
+ write={"spark": "CURRENT_USER()"},
+ )
+ self.validate_all(
+ "SUSER_SNAME()",
+ write={"spark": "CURRENT_USER()"},
+ )
diff --git a/tests/fixtures/identity.sql b/tests/fixtures/identity.sql
index 5fab65b..54e5583 100644
--- a/tests/fixtures/identity.sql
+++ b/tests/fixtures/identity.sql
@@ -666,6 +666,7 @@ INSERT INTO x VALUES (1, 'a', 2.0)
INSERT INTO x VALUES (1, 'a', 2.0), (1, 'a', 3.0), (X(), y[1], z.x)
INSERT INTO y (a, b, c) SELECT a, b, c FROM x
INSERT INTO y (SELECT 1) UNION (SELECT 2)
+INSERT INTO result_table (WITH test AS (SELECT * FROM source_table) SELECT * FROM test)
INSERT OVERWRITE TABLE x IF EXISTS SELECT * FROM y
INSERT OVERWRITE TABLE a.b IF EXISTS SELECT * FROM y
INSERT OVERWRITE DIRECTORY 'x' SELECT 1
@@ -728,6 +729,8 @@ SELECT * INTO newevent FROM event
SELECT * INTO TEMPORARY newevent FROM event
SELECT * INTO UNLOGGED newevent FROM event
ALTER TABLE integers ADD COLUMN k INT
+ALTER TABLE integers ADD COLUMN k INT FIRST
+ALTER TABLE integers ADD COLUMN k INT AFTER m
ALTER TABLE integers ADD COLUMN IF NOT EXISTS k INT
ALTER TABLE IF EXISTS integers ADD COLUMN k INT
ALTER TABLE integers ADD COLUMN l INT DEFAULT 10
diff --git a/tests/fixtures/optimizer/optimizer.sql b/tests/fixtures/optimizer/optimizer.sql
index c5112b2..9e7880c 100644
--- a/tests/fixtures/optimizer/optimizer.sql
+++ b/tests/fixtures/optimizer/optimizer.sql
@@ -462,6 +462,7 @@ SELECT
FROM "db1"."tbl" AS "tbl"
CROSS JOIN "db2"."tbl" AS "tbl_2";
+# execute: false
SELECT
*,
IFF(
diff --git a/tests/fixtures/optimizer/qualify_columns.sql b/tests/fixtures/optimizer/qualify_columns.sql
index df65e65..74e2d0a 100644
--- a/tests/fixtures/optimizer/qualify_columns.sql
+++ b/tests/fixtures/optimizer/qualify_columns.sql
@@ -35,6 +35,9 @@ SELECT 1 AS "1", 2 + 3 AS _col_1 FROM x AS x;
SELECT a + b FROM x;
SELECT x.a + x.b AS _col_0 FROM x AS x;
+SELECT l.a FROM x l WHERE a IN (select a FROM x ORDER by a);
+SELECT l.a AS a FROM x AS l WHERE l.a IN (SELECT x.a AS a FROM x AS x ORDER BY a);
+
# execute: false
SELECT a, SUM(b) FROM x WHERE a > 1 AND b > 1 GROUP BY a;
SELECT x.a AS a, SUM(x.b) AS _col_1 FROM x AS x WHERE x.a > 1 AND x.b > 1 GROUP BY x.a;
@@ -46,15 +49,14 @@ SELECT SUM(a) AS a FROM x HAVING SUM(a) > 3;
SELECT SUM(x.a) AS a FROM x AS x HAVING SUM(x.a) > 3;
SELECT SUM(a) AS c FROM x HAVING c > 3;
-SELECT SUM(x.a) AS c FROM x AS x HAVING c > 3;
+SELECT SUM(x.a) AS c FROM x AS x HAVING SUM(x.a) > 3;
# execute: false
SELECT SUM(a) AS a FROM x HAVING a > 3;
-SELECT SUM(x.a) AS a FROM x AS x HAVING a > 3;
+SELECT SUM(x.a) AS a FROM x AS x HAVING SUM(x.a) > 3;
-# execute: false
-SELECT SUM(a) AS c FROM x HAVING SUM(c) > 3;
-SELECT SUM(x.a) AS c FROM x AS x HAVING SUM(c) > 3;
+SELECT SUM(a) AS c FROM x HAVING SUM(b) > 3;
+SELECT SUM(x.a) AS c FROM x AS x HAVING SUM(x.b) > 3;
SELECT a AS j, b FROM x ORDER BY j;
SELECT x.a AS j, x.b AS b FROM x AS x ORDER BY j;
@@ -95,6 +97,7 @@ SELECT COALESCE(x.a) AS d FROM x AS x JOIN y AS y ON x.b = y.b GROUP BY COALESCE
SELECT a + 1 AS d FROM x WHERE d > 1;
SELECT x.a + 1 AS d FROM x AS x WHERE x.a + 1 > 1;
+# execute: false
SELECT a + 1 AS d, d + 2 FROM x;
SELECT x.a + 1 AS d, x.a + 1 + 2 AS _col_1 FROM x AS x;
@@ -124,6 +127,10 @@ SELECT DATE_TRUNC('week', x.a) AS a FROM x AS x;
SELECT DATE_TRUNC(a, MONTH) AS a FROM x;
SELECT DATE_TRUNC(x.a, MONTH) AS a FROM x AS x;
+# execute: false
+SELECT x FROM READ_PARQUET('path.parquet', hive_partition=1);
+SELECT _q_0.x AS x FROM READ_PARQUET('path.parquet', hive_partition = 1) AS _q_0;
+
--------------------------------------
-- Derived tables
--------------------------------------
@@ -262,11 +269,9 @@ SELECT x.a AS d, x.b AS b FROM x AS x;
SELECT * EXCEPT(b) REPLACE(a AS d) FROM x;
SELECT x.a AS d FROM x AS x;
-# execute: false
SELECT x.* EXCEPT(a), y.* FROM x, y;
SELECT x.b AS b, y.b AS b, y.c AS c FROM x AS x, y AS y;
-# execute: false
SELECT * EXCEPT(a) FROM x;
SELECT x.b AS b FROM x AS x;
@@ -338,12 +343,11 @@ SELECT t.c AS c FROM x AS x LATERAL VIEW EXPLODE(x.a) t AS c;
SELECT aa FROM x, UNNEST(a) AS t(aa);
SELECT t.aa AS aa FROM x AS x, UNNEST(x.a) AS t(aa);
-# execute: false
# dialect: bigquery
+# execute: false
SELECT aa FROM x, UNNEST(a) AS aa;
SELECT aa AS aa FROM x AS x, UNNEST(x.a) AS aa;
-# execute: false
# dialect: presto
SELECT x.a, i.b FROM x CROSS JOIN UNNEST(SPLIT(b, ',')) AS i(b);
SELECT x.a AS a, i.b AS b FROM x AS x CROSS JOIN UNNEST(SPLIT(x.b, ',')) AS i(b);
diff --git a/tests/fixtures/optimizer/simplify.sql b/tests/fixtures/optimizer/simplify.sql
index 08e8700..54ec64b 100644
--- a/tests/fixtures/optimizer/simplify.sql
+++ b/tests/fixtures/optimizer/simplify.sql
@@ -282,6 +282,9 @@ x * (1 - y);
-1 + 3;
2;
+1 - 2 - 4;
+-5;
+
-(-1);
1;
@@ -307,19 +310,22 @@ x * (1 - y);
0.0219;
1 / 3;
-0;
+1 / 3;
+
+1 / 3.0;
+0.3333333333333333333333333333;
20.0 / 6;
3.333333333333333333333333333;
10 / 5;
-2;
+10 / 5;
(1.0 * 3) * 4 - 2 * (5 / 2);
-8.0;
+12.0 - 2 * (5 / 2);
6 - 2 + 4 * 2 + a;
-a + 12;
+12 + a;
a + 1 + 1 + 2;
a + 4;
@@ -376,7 +382,7 @@ interval '1' year + date '1998-01-01';
CAST('1999-01-01' AS DATE);
interval '1' year + date '1998-01-01' + 3 * 7 * 4;
-84 + CAST('1999-01-01' AS DATE);
+CAST('1999-01-01' AS DATE) + 84;
date '1998-12-01' - interval '90' foo;
CAST('1998-12-01' AS DATE) - INTERVAL '90' foo;
diff --git a/tests/fixtures/optimizer/tpc-ds/tpc-ds.sql b/tests/fixtures/optimizer/tpc-ds/tpc-ds.sql
index b92ad37..d9a06cc 100644
--- a/tests/fixtures/optimizer/tpc-ds/tpc-ds.sql
+++ b/tests/fixtures/optimizer/tpc-ds/tpc-ds.sql
@@ -6145,7 +6145,7 @@ WITH web_v1 AS
(
SELECT ws_item_sk item_sk,
d_date,
- sum(Sum(ws_sales_price)) OVER (partition BY ws_item_sk ORDER BY d_date rows BETWEEN UNBOUNDED PRECEDING AND CURRENT row) cume_sales
+ sum(Sum(ws_sales_price)) OVER (partition BY ws_item_sk ORDER BY d_date rows BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) cume_sales
FROM web_sales ,
date_dim
WHERE ws_sold_date_sk=d_date_sk
@@ -6156,7 +6156,7 @@ WITH web_v1 AS
(
SELECT ss_item_sk item_sk,
d_date,
- sum(sum(ss_sales_price)) OVER (partition BY ss_item_sk ORDER BY d_date rows BETWEEN UNBOUNDED PRECEDING AND CURRENT row) cume_sales
+ sum(sum(ss_sales_price)) OVER (partition BY ss_item_sk ORDER BY d_date rows BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) cume_sales
FROM store_sales ,
date_dim
WHERE ss_sold_date_sk=d_date_sk
@@ -6171,8 +6171,8 @@ FROM (
d_date ,
web_sales ,
store_sales ,
- max(web_sales) OVER (partition BY item_sk ORDER BY d_date rows BETWEEN UNBOUNDED PRECEDING AND CURRENT row) web_cumulative ,
- max(store_sales) OVER (partition BY item_sk ORDER BY d_date rows BETWEEN UNBOUNDED PRECEDING AND CURRENT row) store_cumulative
+ max(web_sales) OVER (partition BY item_sk ORDER BY d_date rows BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) web_cumulative ,
+ max(store_sales) OVER (partition BY item_sk ORDER BY d_date rows BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) store_cumulative
FROM (
SELECT
CASE
@@ -6206,7 +6206,7 @@ WITH "date_dim_2" AS (
SELECT
"web_sales"."ws_item_sk" AS "item_sk",
"date_dim"."d_date" AS "d_date",
- SUM(SUM("web_sales"."ws_sales_price")) OVER (PARTITION BY "web_sales"."ws_item_sk" ORDER BY "date_dim"."d_date" rows BETWEEN UNBOUNDED PRECEDING AND CURRENT row) AS "cume_sales"
+ SUM(SUM("web_sales"."ws_sales_price")) OVER (PARTITION BY "web_sales"."ws_item_sk" ORDER BY "date_dim"."d_date" rows BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) AS "cume_sales"
FROM "web_sales" AS "web_sales"
JOIN "date_dim_2" AS "date_dim"
ON "web_sales"."ws_sold_date_sk" = "date_dim"."d_date_sk"
@@ -6219,7 +6219,7 @@ WITH "date_dim_2" AS (
SELECT
"store_sales"."ss_item_sk" AS "item_sk",
"date_dim"."d_date" AS "d_date",
- SUM(SUM("store_sales"."ss_sales_price")) OVER (PARTITION BY "store_sales"."ss_item_sk" ORDER BY "date_dim"."d_date" rows BETWEEN UNBOUNDED PRECEDING AND CURRENT row) AS "cume_sales"
+ SUM(SUM("store_sales"."ss_sales_price")) OVER (PARTITION BY "store_sales"."ss_item_sk" ORDER BY "date_dim"."d_date" rows BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) AS "cume_sales"
FROM "store_sales" AS "store_sales"
JOIN "date_dim_2" AS "date_dim"
ON "store_sales"."ss_sold_date_sk" = "date_dim"."d_date_sk"
@@ -6242,12 +6242,12 @@ WITH "date_dim_2" AS (
WHEN NOT "web"."item_sk" IS NULL
THEN "web"."item_sk"
ELSE "store"."item_sk"
- END ORDER BY CASE WHEN NOT "web"."d_date" IS NULL THEN "web"."d_date" ELSE "store"."d_date" END rows BETWEEN UNBOUNDED PRECEDING AND CURRENT row) AS "web_cumulative",
+ END ORDER BY CASE WHEN NOT "web"."d_date" IS NULL THEN "web"."d_date" ELSE "store"."d_date" END rows BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) AS "web_cumulative",
MAX("store"."cume_sales") OVER (PARTITION BY CASE
WHEN NOT "web"."item_sk" IS NULL
THEN "web"."item_sk"
ELSE "store"."item_sk"
- END ORDER BY CASE WHEN NOT "web"."d_date" IS NULL THEN "web"."d_date" ELSE "store"."d_date" END rows BETWEEN UNBOUNDED PRECEDING AND CURRENT row) AS "store_cumulative"
+ END ORDER BY CASE WHEN NOT "web"."d_date" IS NULL THEN "web"."d_date" ELSE "store"."d_date" END rows BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW) AS "store_cumulative"
FROM "web_v1" AS "web"
FULL JOIN "store_v1" AS "store"
ON "web"."d_date" = "store"."d_date" AND "web"."item_sk" = "store"."item_sk"
diff --git a/tests/test_executor.py b/tests/test_executor.py
index c455f3a..6bf7d6a 100644
--- a/tests/test_executor.py
+++ b/tests/test_executor.py
@@ -1,3 +1,4 @@
+import datetime
import unittest
from datetime import date
@@ -513,6 +514,8 @@ class TestExecutor(unittest.TestCase):
self.assertEqual(result.rows, [(3,)])
def test_scalar_functions(self):
+ now = datetime.datetime.now()
+
for sql, expected in [
("CONCAT('a', 'b')", "ab"),
("CONCAT('a', NULL)", None),
@@ -569,6 +572,26 @@ class TestExecutor(unittest.TestCase):
("NULL IS NOT NULL", False),
("NULL = NULL", None),
("NULL <> NULL", None),
+ ("YEAR(CURRENT_TIMESTAMP)", now.year),
+ ("MONTH(CURRENT_TIME)", now.month),
+ ("DAY(CURRENT_DATETIME())", now.day),
+ ("YEAR(CURRENT_DATE())", now.year),
+ ("MONTH(CURRENT_DATE())", now.month),
+ ("DAY(CURRENT_DATE())", now.day),
+ ("YEAR(CURRENT_TIMESTAMP) + 1", now.year + 1),
+ (
+ "YEAR(CURRENT_TIMESTAMP) IN (YEAR(CURRENT_TIMESTAMP) + 1, YEAR(CURRENT_TIMESTAMP) * 10)",
+ False,
+ ),
+ ("YEAR(CURRENT_TIMESTAMP) = (YEAR(CURRENT_TIMESTAMP))", True),
+ ("YEAR(CURRENT_TIMESTAMP) <> (YEAR(CURRENT_TIMESTAMP))", False),
+ ("YEAR(CURRENT_DATE()) + 1", now.year + 1),
+ (
+ "YEAR(CURRENT_DATE()) IN (YEAR(CURRENT_DATE()) + 1, YEAR(CURRENT_DATE()) * 10)",
+ False,
+ ),
+ ("YEAR(CURRENT_DATE()) = (YEAR(CURRENT_DATE()))", True),
+ ("YEAR(CURRENT_DATE()) <> (YEAR(CURRENT_DATE()))", False),
]:
with self.subTest(sql):
result = execute(f"SELECT {sql}")
diff --git a/tests/test_expressions.py b/tests/test_expressions.py
index c22f13e..b09b2ab 100644
--- a/tests/test_expressions.py
+++ b/tests/test_expressions.py
@@ -194,30 +194,31 @@ class TestExpressions(unittest.TestCase):
def test_replace_placeholders(self):
self.assertEqual(
exp.replace_placeholders(
- parse_one("select * from :tbl1 JOIN :tbl2 ON :col1 = :col2 WHERE :col3 > 100"),
- tbl1="foo",
- tbl2="bar",
- col1="a",
- col2="b",
- col3="c",
+ parse_one("select * from :tbl1 JOIN :tbl2 ON :col1 = :str1 WHERE :col2 > :int1"),
+ tbl1=exp.to_identifier("foo"),
+ tbl2=exp.to_identifier("bar"),
+ col1=exp.to_identifier("a"),
+ col2=exp.to_identifier("c"),
+ str1="b",
+ int1=100,
).sql(),
- "SELECT * FROM foo JOIN bar ON a = b WHERE c > 100",
+ "SELECT * FROM foo JOIN bar ON a = 'b' WHERE c > 100",
)
self.assertEqual(
exp.replace_placeholders(
- parse_one("select * from ? JOIN ? ON ? = ? WHERE ? > 100"),
- "foo",
- "bar",
- "a",
+ parse_one("select * from ? JOIN ? ON ? = ? WHERE ? = 'bla'"),
+ exp.to_identifier("foo"),
+ exp.to_identifier("bar"),
+ exp.to_identifier("a"),
"b",
- "c",
+ "bla",
).sql(),
- "SELECT * FROM foo JOIN bar ON a = b WHERE c > 100",
+ "SELECT * FROM foo JOIN bar ON a = 'b' WHERE 'bla' = 'bla'",
)
self.assertEqual(
exp.replace_placeholders(
parse_one("select * from ? WHERE ? > 100"),
- "foo",
+ exp.to_identifier("foo"),
).sql(),
"SELECT * FROM foo WHERE ? > 100",
)
@@ -229,12 +230,12 @@ class TestExpressions(unittest.TestCase):
)
self.assertEqual(
exp.replace_placeholders(
- parse_one("select * from (SELECT :col1 FROM ?) WHERE :col2 > 100"),
- "tbl1",
- "tbl2",
+ parse_one("select * from (SELECT :col1 FROM ?) WHERE :col2 > ?"),
+ exp.to_identifier("tbl1"),
+ 100,
"tbl3",
- col1="a",
- col2="b",
+ col1=exp.to_identifier("a"),
+ col2=exp.to_identifier("b"),
col3="c",
).sql(),
"SELECT * FROM (SELECT a FROM tbl1) WHERE b > 100",
@@ -526,6 +527,7 @@ class TestExpressions(unittest.TestCase):
self.assertIsInstance(parse_one("VARIANCE(a)"), exp.Variance)
self.assertIsInstance(parse_one("VARIANCE_POP(a)"), exp.VariancePop)
self.assertIsInstance(parse_one("YEAR(a)"), exp.Year)
+ self.assertIsInstance(parse_one("HLL(a)"), exp.Hll)
def test_column(self):
column = parse_one("a.b.c.d")
diff --git a/tests/test_optimizer.py b/tests/test_optimizer.py
index 597fa6f..d077570 100644
--- a/tests/test_optimizer.py
+++ b/tests/test_optimizer.py
@@ -106,7 +106,6 @@ class TestOptimizer(unittest.TestCase):
):
title = meta.get("title") or f"{i}, {sql}"
dialect = meta.get("dialect")
- execute = execute if meta.get("execute") is None else False
leave_tables_isolated = meta.get("leave_tables_isolated")
func_kwargs = {**kwargs}
@@ -114,7 +113,13 @@ class TestOptimizer(unittest.TestCase):
func_kwargs["leave_tables_isolated"] = string_to_bool(leave_tables_isolated)
future = pool.submit(parse_and_optimize, func, sql, dialect, **func_kwargs)
- results[future] = (sql, title, expected, dialect, execute)
+ results[future] = (
+ sql,
+ title,
+ expected,
+ dialect,
+ execute if meta.get("execute") is None else False,
+ )
for future in as_completed(results):
optimized = future.result()
diff --git a/tests/test_tokens.py b/tests/test_tokens.py
index 909eb18..8481f4d 100644
--- a/tests/test_tokens.py
+++ b/tests/test_tokens.py
@@ -13,6 +13,7 @@ class TestTokens(unittest.TestCase):
("foo --comment", ["comment"]),
("foo", []),
("foo /*comment 1*/ /*comment 2*/", ["comment 1", "comment 2"]),
+ ("foo\n-- comment", [" comment"]),
]
for sql, comment in sql_comment: