Adding upstream version 26.22.1.

Signed-off-by: Daniel Baumann <daniel@debian.org>

tests/dialects/test_bigquery.py

@@ -177,6 +177,10 @@ class TestBigQuery(Validator):
         self.validate_identity(
             "CREATE OR REPLACE VIEW test (tenant_id OPTIONS (description='Test description on table creation')) AS SELECT 1 AS tenant_id, 1 AS customer_id",
         )
+        self.validate_identity(
+            "--c\nARRAY_AGG(v IGNORE NULLS)",
+            "ARRAY_AGG(v IGNORE NULLS) /* c */",
+        )
         self.validate_identity(
             'SELECT r"\\t"',
             "SELECT '\\\\t'",
@@ -1732,6 +1736,9 @@ WHERE
         )
 
     def test_errors(self):
+        with self.assertRaises(ParseError):
+            self.parse_one("SELECT * FROM a - b.c.d2")
+
         with self.assertRaises(TokenError):
             transpile("'\\'", read="bigquery")
 

tests/dialects/test_dialect.py

@@ -11,6 +11,7 @@ from sqlglot import (
     parse_one,
 )
 from sqlglot.dialects import BigQuery, Hive, Snowflake
+from sqlglot.dialects.dialect import Version
 from sqlglot.parser import logger as parser_logger
 
 
@@ -134,14 +135,25 @@ class TestDialect(Validator):
             "oracle, normalization_strategy = lowercase, version = 19.5"
         )
         self.assertEqual(oracle_with_settings.normalization_strategy.value, "LOWERCASE")
-        self.assertEqual(oracle_with_settings.settings, {"version": "19.5"})
+        self.assertEqual(oracle_with_settings.version, Version("19.5"))
 
-        bool_settings = Dialect.get_or_raise("oracle, s1=TruE, s2=1, s3=FaLse, s4=0, s5=nonbool")
+        class MyDialect(Dialect):
+            SUPPORTED_SETTINGS = {"s1", "s2", "s3", "s4", "s5"}
+
+        bool_settings = Dialect.get_or_raise("mydialect, s1=TruE, s2=1, s3=FaLse, s4=0, s5=nonbool")
         self.assertEqual(
             bool_settings.settings,
             {"s1": True, "s2": True, "s3": False, "s4": False, "s5": "nonbool"},
         )
 
+        with self.assertRaises(ValueError) as cm:
+            Dialect.get_or_raise("tsql,normalisation_strategy=case_sensitive")
+
+        self.assertEqual(
+            "Unknown setting 'normalisation_strategy'. Did you mean normalization_strategy?",
+            str(cm.exception),
+        )
+
     def test_compare_dialects(self):
         bigquery_class = Dialect["bigquery"]
         bigquery_object = BigQuery()
@@ -170,7 +182,9 @@ class TestDialect(Validator):
 
     def test_compare_dialect_versions(self):
         ddb_v1 = Dialect.get_or_raise("duckdb, version=1.0")
-        ddb_v1_2 = Dialect.get_or_raise("duckdb, foo=bar, version=1.0")
+        ddb_v1_2 = Dialect.get_or_raise(
+            "duckdb, normalization_strategy=case_sensitive, version=1.0"
+        )
         ddb_v2 = Dialect.get_or_raise("duckdb, version=2.2.4")
         ddb_latest = Dialect.get_or_raise("duckdb")
 

tests/dialects/test_duckdb.py

@@ -9,6 +9,7 @@ class TestDuckDB(Validator):
     dialect = "duckdb"
 
     def test_duckdb(self):
+        self.validate_identity("SELECT * FROM my_ducklake.demo AT (VERSION => 2)")
         self.validate_identity("SELECT UUIDV7()")
         self.validate_identity("SELECT TRY(LOG(0))")
         self.validate_identity("x::timestamp", "CAST(x AS TIMESTAMP)")

tests/dialects/test_snowflake.py

@@ -2667,3 +2667,25 @@ SINGLE = TRUE""",
         for node in (max_by, min_by):
             self.assertEqual(len(node.this.expressions), 1)
             self.assertIsInstance(node.expression, exp.Column)
+
+    def test_create_view_copy_grants(self):
+        # for normal views, 'COPY GRANTS' goes *after* the column list. ref: https://docs.snowflake.com/en/sql-reference/sql/create-view#syntax
+        self.validate_identity(
+            "CREATE OR REPLACE VIEW FOO (A, B) COPY GRANTS AS SELECT A, B FROM TBL"
+        )
+
+        # for materialized views, 'COPY GRANTS' must go *before* the column list or an error will be thrown. ref: https://docs.snowflake.com/en/sql-reference/sql/create-materialized-view#syntax
+        self.validate_identity(
+            "CREATE OR REPLACE MATERIALIZED VIEW FOO COPY GRANTS (A, B) AS SELECT A, B FROM TBL"
+        )
+
+        # check that only 'COPY GRANTS' goes before the column list and other properties still go after
+        self.validate_identity(
+            "CREATE OR REPLACE MATERIALIZED VIEW FOO COPY GRANTS (A, B) COMMENT='foo' TAG (a='b') AS SELECT A, B FROM TBL"
+        )
+
+        # no COPY GRANTS
+        self.validate_identity("CREATE OR REPLACE VIEW FOO (A, B) AS SELECT A, B FROM TBL")
+        self.validate_identity(
+            "CREATE OR REPLACE MATERIALIZED VIEW FOO (A, B) AS SELECT A, B FROM TBL"
+        )

tests/dialects/test_tsql.py

@@ -1318,6 +1318,7 @@ WHERE
         )
 
     def test_isnull(self):
+        self.validate_identity("ISNULL(x, y)")
         self.validate_all("ISNULL(x, y)", write={"spark": "COALESCE(x, y)"})
 
     def test_json(self):

tests/fixtures/optimizer/annotate_functions.sql

@@ -28,6 +28,21 @@ TIME;
 TIME_SUB(CAST('09:05:03' AS TIME), INTERVAL 2 HOUR);
 TIME;
 
+SORT_ARRAY(ARRAY(tbl.str_col));
+ARRAY<STRING>;
+
+SORT_ARRAY(ARRAY(tbl.double_col));
+ARRAY<DOUBLE>;
+
+SORT_ARRAY(ARRAY(tbl.bigint_col));
+ARRAY<BIGINT>;
+
+tbl.bigint || tbl.str_col;
+VARCHAR;
+
+tbl.str_col || tbl.bigint;
+VARCHAR;
+
 --------------------------------------
 -- Spark2 / Spark3 / Databricks
 --------------------------------------

tests/fixtures/optimizer/annotate_types.sql

@@ -86,3 +86,9 @@ TIME;
 # dialect: bigquery
 EXTRACT(day from x);
 INT;
+
+CASE WHEN x THEN CAST(y AS DECIMAL(18, 2)) ELSE NULL END;
+DECIMAL(18,2);
+
+CASE WHEN x THEN NULL ELSE CAST(y AS DECIMAL(18, 2)) END;
+DECIMAL(18,2);

tests/fixtures/optimizer/optimizer.sql

@@ -693,14 +693,14 @@ PIVOT(MAX("SOURCE"."VALUE") FOR "SOURCE"."KEY" IN ('a', 'b', 'c')) AS "FINAL"("I
 # dialect: snowflake
 SELECT * FROM m_sales AS m_sales(empid, dept, jan, feb) UNPIVOT(sales FOR month IN (jan, feb)) ORDER BY empid;
 SELECT
-  "_Q_0"."EMPID" AS "EMPID",
-  "_Q_0"."DEPT" AS "DEPT",
-  "_Q_0"."MONTH" AS "MONTH",
-  "_Q_0"."SALES" AS "SALES"
+  "M_SALES"."EMPID" AS "EMPID",
+  "M_SALES"."DEPT" AS "DEPT",
+  "M_SALES"."MONTH" AS "MONTH",
+  "M_SALES"."SALES" AS "SALES"
 FROM "M_SALES" AS "M_SALES"("EMPID", "DEPT", "JAN", "FEB")
-UNPIVOT("SALES" FOR "MONTH" IN ("JAN", "FEB")) AS "_Q_0"
+UNPIVOT("SALES" FOR "MONTH" IN ("JAN", "FEB")) AS "M_SALES"
 ORDER BY
-  "_Q_0"."EMPID";
+  "M_SALES"."EMPID";
 
 # title: unpivoted table source, unpivot has column aliases
 # execute: false
@@ -747,28 +747,28 @@ ORDER BY
 # note: the named columns aren not supported by BQ but we add them here to avoid defining a schema
 SELECT * FROM produce AS produce(product, q1, q2, q3, q4) UNPIVOT(sales FOR quarter IN (q1, q2, q3, q4));
 SELECT
-  `_q_0`.`product` AS `product`,
-  `_q_0`.`quarter` AS `quarter`,
-  `_q_0`.`sales` AS `sales`
+  `produce`.`product` AS `product`,
+  `produce`.`quarter` AS `quarter`,
+  `produce`.`sales` AS `sales`
 FROM `produce` AS `produce`
-UNPIVOT(`sales` FOR `quarter` IN (`produce`.`q1`, `produce`.`q2`, `produce`.`q3`, `produce`.`q4`)) AS `_q_0`;
+UNPIVOT(`sales` FOR `quarter` IN (`produce`.`q1`, `produce`.`q2`, `produce`.`q3`, `produce`.`q4`)) AS `produce`;
 
 # title: unpivoted table source with multiple value columns
 # execute: false
 # dialect: bigquery
 SELECT * FROM produce AS produce(product, q1, q2, q3, q4) UNPIVOT((first_half_sales, second_half_sales) FOR semesters IN ((Q1, Q2) AS 'semester_1', (Q3, Q4) AS 'semester_2'));
 SELECT
-  `_q_0`.`product` AS `product`,
-  `_q_0`.`semesters` AS `semesters`,
-  `_q_0`.`first_half_sales` AS `first_half_sales`,
-  `_q_0`.`second_half_sales` AS `second_half_sales`
+  `produce`.`product` AS `product`,
+  `produce`.`semesters` AS `semesters`,
+  `produce`.`first_half_sales` AS `first_half_sales`,
+  `produce`.`second_half_sales` AS `second_half_sales`
 FROM `produce` AS `produce`
 UNPIVOT((`first_half_sales`, `second_half_sales`) FOR 
   `semesters` IN (
     (`produce`.`q1`, `produce`.`q2`) AS 'semester_1',
     (`produce`.`q3`, `produce`.`q4`) AS 'semester_2'
   )
-) AS `_q_0`;
+) AS `produce`;
 
 # title: quoting is preserved
 # dialect: snowflake

tests/fixtures/optimizer/tpc-ds/tpc-ds.sql

@@ -11105,11 +11105,9 @@ LEFT JOIN "ws" AS "ws"
   AND "ss"."ss_item_sk" = "ws"."ws_item_sk"
   AND "ss"."ss_sold_year" = "ws"."ws_sold_year"
 WHERE
-  "cs"."cs_qty" > 0
-  AND "ss"."ss_sold_year" = 1999
-  AND "ws"."ws_qty" > 0
-  AND NOT "cs"."cs_qty" IS NULL
-  AND NOT "ws"."ws_qty" IS NULL
+  "ss"."ss_sold_year" = 1999
+  AND COALESCE("cs"."cs_qty", 0) > 0
+  AND COALESCE("ws"."ws_qty", 0) > 0
 ORDER BY
   "ss_item_sk",
   "ss"."ss_qty" DESC,

tests/test_optimizer.py

@@ -51,7 +51,9 @@ def normalize(expression, **kwargs):
 
 
 def simplify(expression, **kwargs):
-    return optimizer.simplify.simplify(expression, constant_propagation=True, **kwargs)
+    return optimizer.simplify.simplify(
+        expression, constant_propagation=True, coalesce_simplification=True, **kwargs
+    )
 
 
 def annotate_functions(expression, **kwargs):

tests/test_parser.py

@@ -985,6 +985,9 @@ class TestParser(unittest.TestCase):
             {"line": 1, "col": 81, "start": 69, "end": 80},
         )
 
+        ast = parse_one("SELECT FOO()")
+        self.assertEqual(ast.find(exp.Anonymous).meta, {"line": 1, "col": 10, "start": 7, "end": 9})
+
     def test_quoted_identifier_meta(self):
         sql = 'SELECT "a" FROM "test_schema"."test_table_a"'
         ast = parse_one(sql)