Feat: add distribute by alias for hive window functions closes #2043

sqlglot/dialects/hive.py

@@ -220,6 +220,7 @@ class Tokenizer(tokens.Tokenizer):
     class Parser(parser.Parser):
         LOG_DEFAULTS_TO_LN = True
         STRICT_CAST = False
+        PARTITION_BY_TOKENS = {*parser.Parser.PARTITION_BY_TOKENS, TokenType.DISTRIBUTE_BY}
 
         FUNCTIONS = {
             **parser.Parser.FUNCTIONS,

sqlglot/parser.py

@@ -812,6 +812,7 @@ class Parser(metaclass=_Parser):
     WINDOW_ALIAS_TOKENS = ID_VAR_TOKENS - {TokenType.ROWS}
     WINDOW_BEFORE_PAREN_TOKENS = {TokenType.OVER}
     WINDOW_SIDES = {"FOLLOWING", "PRECEDING"}
+    PARTITION_BY_TOKENS = {TokenType.PARTITION_BY}
 
     ADD_CONSTRAINT_TOKENS = {TokenType.CONSTRAINT, TokenType.PRIMARY_KEY, TokenType.FOREIGN_KEY}
 
@@ -1623,7 +1624,7 @@ def _parse_locking(self) -> exp.LockingProperty:
         )
 
     def _parse_partition_by(self) -> t.List[t.Optional[exp.Expression]]:
-        if self._match(TokenType.PARTITION_BY):
+        if self._match_set(self.PARTITION_BY_TOKENS):
             return self._parse_csv(self._parse_conjunction)
         return []
 

tests/dialects/test_hive.py

@@ -390,6 +390,9 @@ def test_order_by(self):
         )
 
     def test_hive(self):
+        self.validate_identity(
+            "SELECT ROW() OVER (DISTRIBUTE BY x)", "SELECT ROW() OVER (PARTITION BY x)"
+        )
         self.validate_identity("SELECT transform")
         self.validate_identity("SELECT * FROM test DISTRIBUTE BY y SORT BY x DESC ORDER BY l")
         self.validate_identity(