sqlglot 27.19.0__py3-none-any.whl → 27.20.0__py3-none-any.whl

sqlglot/_version.py

@@ -28,7 +28,7 @@ version_tuple: VERSION_TUPLE
 commit_id: COMMIT_ID
 __commit_id__: COMMIT_ID
 
-__version__ = version = '27.19.0'
-__version_tuple__ = version_tuple = (27, 19, 0)
+__version__ = version = '27.20.0'
+__version_tuple__ = version_tuple = (27, 20, 0)
 
 __commit_id__ = commit_id = None

sqlglot/dialects/snowflake.py

@@ -543,6 +543,7 @@ class Snowflake(Dialect):
         exp.DataType.Type.VARCHAR: {
             *Dialect.TYPE_TO_EXPRESSIONS[exp.DataType.Type.VARCHAR],
             exp.Base64DecodeString,
+            exp.TryBase64DecodeString,
             exp.Base64Encode,
             exp.DecompressString,
             exp.MD5,
@@ -553,6 +554,7 @@ class Snowflake(Dialect):
             exp.Collate,
             exp.Collation,
             exp.HexDecodeString,
+            exp.TryHexDecodeString,
             exp.HexEncode,
             exp.Initcap,
             exp.RegexpExtract,
@@ -561,12 +563,16 @@ class Snowflake(Dialect):
             exp.Replace,
             exp.SHA,
             exp.SHA2,
+            exp.Soundex,
             exp.Space,
+            exp.SplitPart,
             exp.Uuid,
         },
         exp.DataType.Type.BINARY: {
             *Dialect.TYPE_TO_EXPRESSIONS[exp.DataType.Type.BINARY],
             exp.Base64DecodeBinary,
+            exp.TryBase64DecodeBinary,
+            exp.TryHexDecodeBinary,
             exp.Compress,
             exp.DecompressBinary,
             exp.MD5Digest,
@@ -586,6 +592,9 @@ class Snowflake(Dialect):
             exp.ParseUrl,
             exp.ParseIp,
         },
+        exp.DataType.Type.DECIMAL: {
+            exp.RegexpCount,
+        },
     }
 
     ANNOTATORS = {
@@ -607,6 +616,9 @@ class Snowflake(Dialect):
         },
         exp.ConcatWs: lambda self, e: self._annotate_by_args(e, "expressions"),
         exp.Reverse: _annotate_reverse,
+        exp.RegexpCount: lambda self, e: self._annotate_with_type(
+            e, exp.DataType.build("NUMBER", dialect="snowflake")
+        ),
     }
 
     TIME_MAPPING = {
@@ -793,6 +805,7 @@ class Snowflake(Dialect):
         FUNCTION_PARSERS = {
             **parser.Parser.FUNCTION_PARSERS,
             "DATE_PART": lambda self: self._parse_date_part(),
+            "DIRECTORY": lambda self: self._parse_directory(),
             "OBJECT_CONSTRUCT_KEEP_NULL": lambda self: self._parse_json_object(),
             "LISTAGG": lambda self: self._parse_string_agg(),
             "SEMANTIC_VIEW": lambda self: self._parse_semantic_view(),
@@ -902,6 +915,14 @@ class Snowflake(Dialect):
             ),
         }
 
+        def _parse_directory(self) -> exp.DirectoryStage:
+            table = self._parse_table_parts()
+
+            if isinstance(table, exp.Table):
+                table = table.this
+
+            return self.expression(exp.DirectoryStage, this=table)
+
         def _parse_use(self) -> exp.Use:
             if self._match_text_seq("SECONDARY", "ROLES"):
                 this = self._match_texts(("ALL", "NONE")) and exp.var(self._prev.text.upper())

sqlglot/dialects/sqlite.py

@@ -342,3 +342,12 @@ class SQLite(Dialect):
 
         def respectnulls_sql(self, expression: exp.RespectNulls) -> str:
             return self.sql(expression.this)
+
+        def windowspec_sql(self, expression: exp.WindowSpec) -> str:
+            if (
+                expression.text("kind").upper() == "RANGE"
+                and expression.text("start").upper() == "CURRENT ROW"
+            ):
+                return "RANGE CURRENT ROW"
+
+            return super().windowspec_sql(expression)

sqlglot/dialects/starrocks.py

@@ -32,6 +32,7 @@ def st_distance_sphere(self, expression: exp.StDistance) -> str:
 
 class StarRocks(MySQL):
     STRICT_JSON_PATH_SYNTAX = False
+    INDEX_OFFSET = 1
 
     class Tokenizer(MySQL.Tokenizer):
         KEYWORDS = {
@@ -49,6 +50,7 @@ class StarRocks(MySQL):
             "DATE_DIFF": lambda args: exp.DateDiff(
                 this=seq_get(args, 1), expression=seq_get(args, 2), unit=seq_get(args, 0)
             ),
+            "ARRAY_FLATTEN": exp.Flatten.from_arg_list,
             "REGEXP": exp.RegexpLike.from_arg_list,
         }
 
@@ -152,6 +154,7 @@ class StarRocks(MySQL):
             exp.DateDiff: lambda self, e: self.func(
                 "DATE_DIFF", unit_to_str(e), e.this, e.expression
             ),
+            exp.Flatten: rename_func("ARRAY_FLATTEN"),
             exp.JSONExtractScalar: arrow_json_extract_sql,
             exp.JSONExtract: arrow_json_extract_sql,
             exp.Property: property_sql,

sqlglot/expressions.py

@@ -120,19 +120,43 @@ class Expression(metaclass=_Expression):
     def __eq__(self, other) -> bool:
         return type(self) is type(other) and hash(self) == hash(other)
 
-    @property
-    def hashable_args(self) -> t.Any:
-        return frozenset(
-            (k, tuple(_norm_arg(a) for a in v) if type(v) is list else _norm_arg(v))
-            for k, v in self.args.items()
-            if not (v is None or v is False or (type(v) is list and not v))
-        )
-
     def __hash__(self) -> int:
-        if self._hash is not None:
-            return self._hash
-
-        return hash((self.__class__, self.hashable_args))
+        if self._hash is None:
+            nodes = []
+            queue = deque([self])
+
+            while queue:
+                node = queue.popleft()
+                nodes.append(node)
+
+                for v in node.iter_expressions():
+                    if v._hash is None:
+                        queue.append(v)
+
+            for node in reversed(nodes):
+                hash_ = hash(node.key)
+                t = type(node)
+
+                if t is Literal or t is Identifier:
+                    for k, v in sorted(node.args.items()):
+                        if v:
+                            hash_ = hash((hash_, k, v))
+                else:
+                    for k, v in sorted(node.args.items()):
+                        t = type(v)
+
+                        if t is list:
+                            for x in v:
+                                if x is not None and x is not False:
+                                    hash_ = hash((hash_, k, x.lower() if type(x) is str else x))
+                                else:
+                                    hash_ = hash((hash_, k))
+                        elif v is not None and v is not False:
+                            hash_ = hash((hash_, k, v.lower() if t is str else v))
+
+                node._hash = hash_
+        assert self._hash
+        return self._hash
 
     def __reduce__(self) -> t.Tuple[t.Callable, t.Tuple[t.List[t.Dict[str, t.Any]]]]:
         from sqlglot.serde import dump, load
@@ -369,6 +393,12 @@ class Expression(metaclass=_Expression):
             overwrite: assuming an index is given, this determines whether to overwrite the
                 list entry instead of only inserting a new value (i.e., like list.insert).
         """
+        expression: t.Optional[Expression] = self
+
+        while expression and expression._hash is not None:
+            expression._hash = None
+            expression = expression.parent
+
         if index is not None:
             expressions = self.args.get(arg_key) or []
 
@@ -2235,10 +2265,14 @@ class Prior(Expression):
 
 
 class Directory(Expression):
-    # https://spark.apache.org/docs/3.0.0-preview/sql-ref-syntax-dml-insert-overwrite-directory-hive.html
     arg_types = {"this": True, "local": False, "row_format": False}
 
 
+# https://docs.snowflake.com/en/user-guide/data-load-dirtables-query
+class DirectoryStage(Expression):
+    pass
+
+
 class ForeignKey(Expression):
     arg_types = {
         "expressions": False,
@@ -2298,10 +2332,6 @@ class Identifier(Expression):
     def quoted(self) -> bool:
         return bool(self.args.get("quoted"))
 
-    @property
-    def hashable_args(self) -> t.Any:
-        return (self.this, self.quoted)
-
     @property
     def output_name(self) -> str:
         return self.name
@@ -2536,10 +2566,6 @@ class LimitOptions(Expression):
 class Literal(Condition):
     arg_types = {"this": True, "is_string": True}
 
-    @property
-    def hashable_args(self) -> t.Any:
-        return (self.this, self.args.get("is_string"))
-
     @classmethod
     def number(cls, number) -> Literal:
         return cls(this=str(number), is_string=False)
@@ -6427,14 +6453,36 @@ class Base64DecodeBinary(Func):
     arg_types = {"this": True, "alphabet": False}
 
 
+# https://docs.snowflake.com/en/sql-reference/functions/base64_decode_string
 class Base64DecodeString(Func):
     arg_types = {"this": True, "alphabet": False}
 
 
+# https://docs.snowflake.com/en/sql-reference/functions/base64_encode
 class Base64Encode(Func):
     arg_types = {"this": True, "max_line_length": False, "alphabet": False}
 
 
+# https://docs.snowflake.com/en/sql-reference/functions/try_base64_decode_binary
+class TryBase64DecodeBinary(Func):
+    arg_types = {"this": True, "alphabet": False}
+
+
+# https://docs.snowflake.com/en/sql-reference/functions/try_base64_decode_string
+class TryBase64DecodeString(Func):
+    arg_types = {"this": True, "alphabet": False}
+
+
+# https://docs.snowflake.com/en/sql-reference/functions/try_hex_decode_binary
+class TryHexDecodeBinary(Func):
+    pass
+
+
+# https://docs.snowflake.com/en/sql-reference/functions/try_hex_decode_string
+class TryHexDecodeString(Func):
+    pass
+
+
 # https://trino.io/docs/current/functions/datetime.html#from_iso8601_timestamp
 class FromISO8601Timestamp(Func):
     _sql_names = ["FROM_ISO8601_TIMESTAMP"]
@@ -7249,6 +7297,15 @@ class RegexpSplit(Func):
     arg_types = {"this": True, "expression": True, "limit": False}
 
 
+class RegexpCount(Func):
+    arg_types = {
+        "this": True,
+        "expression": True,
+        "position": False,
+        "parameters": False,
+    }
+
+
 class Repeat(Func):
     arg_types = {"this": True, "times": True}
 
@@ -7712,10 +7769,6 @@ class TableColumn(Expression):
     pass
 
 
-def _norm_arg(arg):
-    return arg.lower() if type(arg) is str else arg
-
-
 ALL_FUNCTIONS = subclasses(__name__, Func, (AggFunc, Anonymous, Func))
 FUNCTION_BY_NAME = {name: func for func in ALL_FUNCTIONS for name in func.sql_names()}
 

sqlglot/generator.py

@@ -5334,3 +5334,6 @@ class Generator(metaclass=_Generator):
     def modelattribute_sql(self, expression: exp.ModelAttribute) -> str:
         self.unsupported("The model!attribute syntax is not supported")
         return ""
+
+    def directorystage_sql(self, expression: exp.DirectoryStage) -> str:
+        return self.func("DIRECTORY", expression.this)

sqlglot/helper.py

@@ -226,31 +226,13 @@ def while_changing(expression: Expression, func: t.Callable[[Expression], E]) ->
     Returns:
         The transformed expression.
     """
-    end_hash: t.Optional[int] = None
 
     while True:
-        # No need to walk the AST– we've already cached the hashes in the previous iteration
-        if end_hash is None:
-            for n in reversed(tuple(expression.walk())):
-                n._hash = hash(n)
-
         start_hash = hash(expression)
         expression = func(expression)
-
-        expression_nodes = tuple(expression.walk())
-
-        # Uncache previous caches so we can recompute them
-        for n in reversed(expression_nodes):
-            n._hash = None
-            n._hash = hash(n)
-
         end_hash = hash(expression)
 
         if start_hash == end_hash:
-            # ... and reset the hash so we don't risk it becoming out of date if a mutation happens
-            for n in expression_nodes:
-                n._hash = None
-
             break
 
     return expression

sqlglot/parser.py

@@ -1411,7 +1411,7 @@ class Parser(metaclass=_Parser):
 
     VIEW_ATTRIBUTES = {"ENCRYPTION", "SCHEMABINDING", "VIEW_METADATA"}
 
-    WINDOW_ALIAS_TOKENS = ID_VAR_TOKENS - {TokenType.ROWS}
+    WINDOW_ALIAS_TOKENS = ID_VAR_TOKENS - {TokenType.RANGE, TokenType.ROWS}
     WINDOW_BEFORE_PAREN_TOKENS = {TokenType.OVER}
     WINDOW_SIDES = {"FOLLOWING", "PRECEDING"}
 

{sqlglot-27.19.0.dist-info → sqlglot-27.20.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sqlglot
-Version: 27.19.0
+Version: 27.20.0
 Summary: An easily customizable SQL parser and transpiler
 Author-email: Toby Mao <toby.mao@gmail.com>
 License-Expression: MIT
@@ -89,12 +89,14 @@ pip3 install "sqlglot[rs]"
 Or with a local checkout:
 
 ```
+# Optionally prefix with UV=1 to use uv for the installation
 make install
 ```
 
 Requirements for development (optional):
 
 ```
+# Optionally prefix with UV=1 to use uv for the installation
 make install-dev
 ```
 

{sqlglot-27.19.0.dist-info → sqlglot-27.20.0.dist-info}/RECORD

@@ -1,15 +1,15 @@
 sqlglot/__init__.py,sha256=za08rtdPh2v7dOpGdNomttlIVGgTrKja7rPd6sQwaTg,5391
 sqlglot/__main__.py,sha256=022c173KqxsiABWTEpUIq_tJUxuNiW7a7ABsxBXqvu8,2069
 sqlglot/_typing.py,sha256=-1HPyr3w5COlSJWqlgt8jhFk2dyMvBuvVBqIX1wyVCM,642
-sqlglot/_version.py,sha256=wVbD9GMknBXDREyMG-Yx-Nw3OZVxyzYGofDpSiCozV4,708
+sqlglot/_version.py,sha256=W-gB5oTd5rWsYZXOhmfTMtDc_wY3Dl-VUjAJOtLX5qo,708
 sqlglot/diff.py,sha256=PtOllQMQa1Sw1-V2Y8eypmDqGujXYPaTOp_WLsWkAWk,17314
 sqlglot/errors.py,sha256=QNKMr-pzLUDR-tuMmn_GK6iMHUIVdb_YSJ_BhGEvuso,2126
-sqlglot/expressions.py,sha256=gR0XWI066giI1oStYhmswO7FY_pA0twxmc1P2XTo1tU,258468
-sqlglot/generator.py,sha256=ZHFzi6_EOJl5V-dnz71QJOIwtxj66oJdF6tUQoxBpck,226436
-sqlglot/helper.py,sha256=9nZjFVRBtMKFC3EdzpDQ6jkazFO19po6BF8xHiNGZIo,15111
+sqlglot/expressions.py,sha256=jQs1lR1Nh5nmXYEjaYVakXgMijNgaWmh1fsOOZ10imE,260181
+sqlglot/generator.py,sha256=KyFuqWQpawTj3rWV7ONKO4euqVTzV8aFU3desDu8fso,226565
+sqlglot/helper.py,sha256=OOt5_Mbmnl4Uy6WO6v7DR1iLPcb3v6ITybpq6usf3jw,14471
 sqlglot/jsonpath.py,sha256=SQgaxzaEYBN7At9dkTK4N1Spk6xHxvHL6QtCIP6iM30,7905
 sqlglot/lineage.py,sha256=Qj5ykuDNcATppb9vOjoIKBqRVLbu3OMPiZk9f3iyv40,15312
-sqlglot/parser.py,sha256=4UXHIDwR9mdbN5C7PVMffheR-cj3ipVJqIIK8W-o5ac,337436
+sqlglot/parser.py,sha256=bVpU72Ace-vGXkiqoojKXFGRpXxProri31Q3WP1TQuE,337453
 sqlglot/planner.py,sha256=ql7Li-bWJRcyXzNaZy_n6bQ6B2ZfunEIB8Ztv2xaxq4,14634
 sqlglot/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 sqlglot/schema.py,sha256=13H2qKQs27EKdTpDLOvcNnSTDAUbYNKjWtJs4aQCSOA,20509
@@ -42,12 +42,12 @@ sqlglot/dialects/prql.py,sha256=fwN-SPEGx-drwf1K0U2MByN-PkW3C_rOgQ3xeJeychg,7908
 sqlglot/dialects/redshift.py,sha256=FIwtP3yEg-way9pa32kxCJc6IaFkHVIvgYKZA-Ilmi0,15919
 sqlglot/dialects/risingwave.py,sha256=BqWwW1iT_OIVMwfRamaww79snnBwIgCfr22Go-ggO68,3289
 sqlglot/dialects/singlestore.py,sha256=0QqNYOucNklPQuyeGcsisLI97qPGx_RfWKOFarJz2qw,61711
-sqlglot/dialects/snowflake.py,sha256=D9GlZtIDLU_aqHWvGQUC8AS-U94WPBS_qEzErIJIoWY,78746
+sqlglot/dialects/snowflake.py,sha256=_SXsdDPNopi78MT7R_YHZHoY6WDB_tZ6pVvk5VcJaYQ,79485
 sqlglot/dialects/solr.py,sha256=pydnl4ml-3M1Fc4ALm6cMVO9h-5EtqZxPZH_91Nz1Ss,617
 sqlglot/dialects/spark.py,sha256=PzyhkelDzbCMgJ3RVHD6yyzLIFp9NdZfwVas5IymowM,10147
 sqlglot/dialects/spark2.py,sha256=qz36FT9k4iuiqboRpyG4VpKGkPR0P2fifmqgZ9gNUEU,14851
-sqlglot/dialects/sqlite.py,sha256=zzXEbnaLjJeg6hPLHricjpfSkuf8tpXECnjcHtoqIbw,13263
-sqlglot/dialects/starrocks.py,sha256=2gav0PSNgRdAGXzawdznZliBpglJoQ0wBxPI7ZIMsRw,11314
+sqlglot/dialects/sqlite.py,sha256=FuEDDyKZeeWVblknhFSMX7dNoS-ci5ktXpSXZeBK5xA,13592
+sqlglot/dialects/starrocks.py,sha256=-NWQa2gJbiMMfLauX-Jy9ciJ5DUzUOk2QkPbhglz5W4,11446
 sqlglot/dialects/tableau.py,sha256=oIawDzUITxGCWaEMB8OaNMPWhbC3U-2y09pYPm4eazc,2190
 sqlglot/dialects/teradata.py,sha256=7LxCcRwP0Idd_OnCzA57NCdheVjHcKC2aFAKG5N49IU,18202
 sqlglot/dialects/trino.py,sha256=Z7prRhCxIBh0KCxIQpWmVOIGHCJM9Xl5oRlqySxln4Y,4350
@@ -77,8 +77,8 @@ sqlglot/optimizer/qualify_tables.py,sha256=dA4ZazL7ShQh2JgBwpHuG-4c5lBw1TNzCnuN7
 sqlglot/optimizer/scope.py,sha256=UOTrbwqcTc5iRQf0WStgYWXpE24w6riZy-tJYA18yTw,31229
 sqlglot/optimizer/simplify.py,sha256=27IYsqbz1kyMlURSfRkm_ADSQJg-4805AOMFOjKKytU,51049
 sqlglot/optimizer/unnest_subqueries.py,sha256=kzWUVDlxs8z9nmRx-8U-pHXPtVZhEIwkKqmKhr2QLvc,10908
-sqlglot-27.19.0.dist-info/licenses/LICENSE,sha256=p1Yk0B4oa0l8Rh-_dYyy75d8spjPd_vTloXfz4FWxys,1065
-sqlglot-27.19.0.dist-info/METADATA,sha256=yaLkwFBKI3tGRmU2ayVsF3uEDoRgBLryVjdzeePUW4M,20703
-sqlglot-27.19.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-sqlglot-27.19.0.dist-info/top_level.txt,sha256=5kRskCGA_gVADF9rSfSzPdLHXqvfMusDYeHePfNY2nQ,8
-sqlglot-27.19.0.dist-info/RECORD,,
+sqlglot-27.20.0.dist-info/licenses/LICENSE,sha256=p1Yk0B4oa0l8Rh-_dYyy75d8spjPd_vTloXfz4FWxys,1065
+sqlglot-27.20.0.dist-info/METADATA,sha256=nO_Q85ybayInmXwdQV_wFMHF4kQSA0ujLlWo-H5r0I8,20825
+sqlglot-27.20.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+sqlglot-27.20.0.dist-info/top_level.txt,sha256=5kRskCGA_gVADF9rSfSzPdLHXqvfMusDYeHePfNY2nQ,8
+sqlglot-27.20.0.dist-info/RECORD,,