Fix(duckdb, clickhouse)!: QUALIFY comes after WINDOW clause in queries (#3745)

* Fix(duckdb): QUALIFY comes after WINDOW clause in queries

* Refactor, don't eliminate qualify for clickhouse

Author: georgesittas

sqlglot/dialects/bigquery.py

@@ -735,6 +735,13 @@ class Generator(generator.Generator):
             exp.VolatileProperty: exp.Properties.Location.UNSUPPORTED,
         }
 
+        # WINDOW comes after QUALIFY
+        # https://cloud.google.com/bigquery/docs/reference/standard-sql/query-syntax#window_clause
+        AFTER_HAVING_MODIFIER_TRANSFORMS = {
+            "qualify": generator.Generator.AFTER_HAVING_MODIFIER_TRANSFORMS["qualify"],
+            "windows": generator.Generator.AFTER_HAVING_MODIFIER_TRANSFORMS["windows"],
+        }
+
         # from: https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#reserved_keywords
         RESERVED_KEYWORDS = {
             "all",

sqlglot/dialects/clickhouse.py

@@ -2,7 +2,7 @@
 
 import typing as t
 
-from sqlglot import exp, generator, parser, tokens, transforms
+from sqlglot import exp, generator, parser, tokens
 from sqlglot.dialects.dialect import (
     Dialect,
     arg_max_or_min_no_count,
@@ -789,7 +789,6 @@ class Generator(generator.Generator):
             exp.Quantile: _quantile_sql,
             exp.RegexpLike: lambda self, e: self.func("match", e.this, e.expression),
             exp.Rand: rename_func("randCanonical"),
-            exp.Select: transforms.preprocess([transforms.eliminate_qualify]),
             exp.StartsWith: rename_func("startsWith"),
             exp.StrPosition: lambda self, e: self.func(
                 "position", e.this, e.args.get("substr"), e.args.get("position")

sqlglot/generator.py

@@ -406,13 +406,13 @@ class Generator(metaclass=_Generator):
     AFTER_HAVING_MODIFIER_TRANSFORMS = {
         "cluster": lambda self, e: self.sql(e, "cluster"),
         "distribute": lambda self, e: self.sql(e, "distribute"),
-        "qualify": lambda self, e: self.sql(e, "qualify"),
         "sort": lambda self, e: self.sql(e, "sort"),
         "windows": lambda self, e: (
             self.seg("WINDOW ") + self.expressions(e, key="windows", flat=True)
             if e.args.get("windows")
             else ""
         ),
+        "qualify": lambda self, e: self.sql(e, "qualify"),
     }
 
     TOKEN_MAPPING: t.Dict[TokenType, str] = {}

tests/dialects/test_clickhouse.py

@@ -83,6 +83,9 @@ def test_clickhouse(self):
         self.validate_identity("TRUNCATE TABLE t1 ON CLUSTER test_cluster")
         self.validate_identity("TRUNCATE DATABASE db")
         self.validate_identity("TRUNCATE DATABASE db ON CLUSTER test_cluster")
+        self.validate_identity(
+            "SELECT number, COUNT() OVER (PARTITION BY number % 3) AS partition_count FROM numbers(10) WINDOW window_name AS (PARTITION BY number) QUALIFY partition_count = 4 ORDER BY number"
+        )
         self.validate_identity(
             "SELECT id, quantileGK(100, 0.95)(reading) OVER (PARTITION BY id ORDER BY id RANGE BETWEEN 30000 PRECEDING AND CURRENT ROW) AS window FROM table"
         )

tests/dialects/test_dialect.py

@@ -2267,7 +2267,7 @@ def test_qualify(self):
             write={
                 "duckdb": "SELECT * FROM t QUALIFY COUNT(*) OVER () > 1",
                 "snowflake": "SELECT * FROM t QUALIFY COUNT(*) OVER () > 1",
-                "clickhouse": "SELECT * FROM (SELECT *, COUNT(*) OVER () AS _w FROM t) AS _t WHERE _w > 1",
+                "clickhouse": "SELECT * FROM t QUALIFY COUNT(*) OVER () > 1",
                 "mysql": "SELECT * FROM (SELECT *, COUNT(*) OVER () AS _w FROM t) AS _t WHERE _w > 1",
                 "oracle": "SELECT * FROM (SELECT *, COUNT(*) OVER () AS _w FROM t) _t WHERE _w > 1",
                 "postgres": "SELECT * FROM (SELECT *, COUNT(*) OVER () AS _w FROM t) AS _t WHERE _w > 1",
@@ -2279,7 +2279,7 @@ def test_qualify(self):
             write={
                 "duckdb": 'SELECT "user id", some_id, 1 AS other_id, 2 AS "2 nd id" FROM t QUALIFY COUNT(*) OVER () > 1',
                 "snowflake": 'SELECT "user id", some_id, 1 AS other_id, 2 AS "2 nd id" FROM t QUALIFY COUNT(*) OVER () > 1',
-                "clickhouse": 'SELECT "user id", some_id, other_id, "2 nd id" FROM (SELECT "user id", some_id, 1 AS other_id, 2 AS "2 nd id", COUNT(*) OVER () AS _w FROM t) AS _t WHERE _w > 1',
+                "clickhouse": 'SELECT "user id", some_id, 1 AS other_id, 2 AS "2 nd id" FROM t QUALIFY COUNT(*) OVER () > 1',
                 "mysql": "SELECT `user id`, some_id, other_id, `2 nd id` FROM (SELECT `user id`, some_id, 1 AS other_id, 2 AS `2 nd id`, COUNT(*) OVER () AS _w FROM t) AS _t WHERE _w > 1",
                 "oracle": 'SELECT "user id", some_id, other_id, "2 nd id" FROM (SELECT "user id", some_id, 1 AS other_id, 2 AS "2 nd id", COUNT(*) OVER () AS _w FROM t) _t WHERE _w > 1',
                 "postgres": 'SELECT "user id", some_id, other_id, "2 nd id" FROM (SELECT "user id", some_id, 1 AS other_id, 2 AS "2 nd id", COUNT(*) OVER () AS _w FROM t) AS _t WHERE _w > 1',

tests/dialects/test_duckdb.py

@@ -361,6 +361,10 @@ def test_duckdb(self):
         self.validate_identity(
             "SELECT * FROM (PIVOT Cities ON Year USING SUM(Population) GROUP BY Country) AS pivot_alias"
         )
+        self.validate_identity(
+            # QUALIFY comes after WINDOW
+            "SELECT schema_name, function_name, ROW_NUMBER() OVER my_window AS function_rank FROM DUCKDB_FUNCTIONS() WINDOW my_window AS (PARTITION BY schema_name ORDER BY function_name) QUALIFY ROW_NUMBER() OVER my_window < 3"
+        )
         self.validate_identity("DATE_SUB('YEAR', col, '2020-01-01')").assert_is(exp.Anonymous)
         self.validate_identity("DATESUB('YEAR', col, '2020-01-01')").assert_is(exp.Anonymous)