sqlglot.dialects.snowflake

   1from __future__ import annotations
   2
   3import typing as t
   4
   5from sqlglot import exp, generator, jsonpath, parser, tokens, transforms
   6from sqlglot.dialects.dialect import (
   7    Dialect,
   8    NormalizationStrategy,
   9    build_timetostr_or_tochar,
  10    binary_from_function,
  11    build_default_decimal_type,
  12    build_replace_with_optional_replacement,
  13    build_timestamp_from_parts,
  14    date_delta_sql,
  15    date_trunc_to_time,
  16    datestrtodate_sql,
  17    build_formatted_time,
  18    if_sql,
  19    inline_array_sql,
  20    max_or_greatest,
  21    min_or_least,
  22    rename_func,
  23    timestamptrunc_sql,
  24    timestrtotime_sql,
  25    var_map_sql,
  26    map_date_part,
  27    no_timestamp_sql,
  28    strposition_sql,
  29    timestampdiff_sql,
  30    no_make_interval_sql,
  31    groupconcat_sql,
  32)
  33from sqlglot.generator import unsupported_args
  34from sqlglot.helper import flatten, is_float, is_int, seq_get
  35from sqlglot.optimizer.scope import find_all_in_scope
  36from sqlglot.tokens import TokenType
  37
  38if t.TYPE_CHECKING:
  39    from sqlglot._typing import E, B
  40
  41
  42# from https://docs.snowflake.com/en/sql-reference/functions/to_timestamp.html
  43def _build_datetime(
  44    name: str, kind: exp.DataType.Type, safe: bool = False
  45) -> t.Callable[[t.List], exp.Func]:
  46    def _builder(args: t.List) -> exp.Func:
  47        value = seq_get(args, 0)
  48        scale_or_fmt = seq_get(args, 1)
  49
  50        int_value = value is not None and is_int(value.name)
  51        int_scale_or_fmt = scale_or_fmt is not None and scale_or_fmt.is_int
  52
  53        if isinstance(value, exp.Literal) or (value and scale_or_fmt):
  54            # Converts calls like `TO_TIME('01:02:03')` into casts
  55            if len(args) == 1 and value.is_string and not int_value:
  56                return (
  57                    exp.TryCast(this=value, to=exp.DataType.build(kind))
  58                    if safe
  59                    else exp.cast(value, kind)
  60                )
  61
  62            # Handles `TO_TIMESTAMP(str, fmt)` and `TO_TIMESTAMP(num, scale)` as special
  63            # cases so we can transpile them, since they're relatively common
  64            if kind == exp.DataType.Type.TIMESTAMP:
  65                if not safe and (int_value or int_scale_or_fmt):
  66                    # TRY_TO_TIMESTAMP('integer') is not parsed into exp.UnixToTime as
  67                    # it's not easily transpilable
  68                    return exp.UnixToTime(this=value, scale=scale_or_fmt)
  69                if not int_scale_or_fmt and not is_float(value.name):
  70                    expr = build_formatted_time(exp.StrToTime, "snowflake")(args)
  71                    expr.set("safe", safe)
  72                    return expr
  73
  74        if kind in (exp.DataType.Type.DATE, exp.DataType.Type.TIME) and not int_value:
  75            klass = exp.TsOrDsToDate if kind == exp.DataType.Type.DATE else exp.TsOrDsToTime
  76            formatted_exp = build_formatted_time(klass, "snowflake")(args)
  77            formatted_exp.set("safe", safe)
  78            return formatted_exp
  79
  80        return exp.Anonymous(this=name, expressions=args)
  81
  82    return _builder
  83
  84
  85def _build_object_construct(args: t.List) -> t.Union[exp.StarMap, exp.Struct]:
  86    expression = parser.build_var_map(args)
  87
  88    if isinstance(expression, exp.StarMap):
  89        return expression
  90
  91    return exp.Struct(
  92        expressions=[
  93            exp.PropertyEQ(this=k, expression=v) for k, v in zip(expression.keys, expression.values)
  94        ]
  95    )
  96
  97
  98def _build_datediff(args: t.List) -> exp.DateDiff:
  99    return exp.DateDiff(
 100        this=seq_get(args, 2), expression=seq_get(args, 1), unit=map_date_part(seq_get(args, 0))
 101    )
 102
 103
 104def _build_date_time_add(expr_type: t.Type[E]) -> t.Callable[[t.List], E]:
 105    def _builder(args: t.List) -> E:
 106        return expr_type(
 107            this=seq_get(args, 2),
 108            expression=seq_get(args, 1),
 109            unit=map_date_part(seq_get(args, 0)),
 110        )
 111
 112    return _builder
 113
 114
 115def _build_bitwise(expr_type: t.Type[B], name: str) -> t.Callable[[t.List], B | exp.Anonymous]:
 116    def _builder(args: t.List) -> B | exp.Anonymous:
 117        if len(args) == 3:
 118            return exp.Anonymous(this=name, expressions=args)
 119
 120        return binary_from_function(expr_type)(args)
 121
 122    return _builder
 123
 124
 125# https://docs.snowflake.com/en/sql-reference/functions/div0
 126def _build_if_from_div0(args: t.List) -> exp.If:
 127    lhs = exp._wrap(seq_get(args, 0), exp.Binary)
 128    rhs = exp._wrap(seq_get(args, 1), exp.Binary)
 129
 130    cond = exp.EQ(this=rhs, expression=exp.Literal.number(0)).and_(
 131        exp.Is(this=lhs, expression=exp.null()).not_()
 132    )
 133    true = exp.Literal.number(0)
 134    false = exp.Div(this=lhs, expression=rhs)
 135    return exp.If(this=cond, true=true, false=false)
 136
 137
 138# https://docs.snowflake.com/en/sql-reference/functions/zeroifnull
 139def _build_if_from_zeroifnull(args: t.List) -> exp.If:
 140    cond = exp.Is(this=seq_get(args, 0), expression=exp.Null())
 141    return exp.If(this=cond, true=exp.Literal.number(0), false=seq_get(args, 0))
 142
 143
 144# https://docs.snowflake.com/en/sql-reference/functions/zeroifnull
 145def _build_if_from_nullifzero(args: t.List) -> exp.If:
 146    cond = exp.EQ(this=seq_get(args, 0), expression=exp.Literal.number(0))
 147    return exp.If(this=cond, true=exp.Null(), false=seq_get(args, 0))
 148
 149
 150def _regexpilike_sql(self: Snowflake.Generator, expression: exp.RegexpILike) -> str:
 151    flag = expression.text("flag")
 152
 153    if "i" not in flag:
 154        flag += "i"
 155
 156    return self.func(
 157        "REGEXP_LIKE", expression.this, expression.expression, exp.Literal.string(flag)
 158    )
 159
 160
 161def _build_regexp_replace(args: t.List) -> exp.RegexpReplace:
 162    regexp_replace = exp.RegexpReplace.from_arg_list(args)
 163
 164    if not regexp_replace.args.get("replacement"):
 165        regexp_replace.set("replacement", exp.Literal.string(""))
 166
 167    return regexp_replace
 168
 169
 170def _show_parser(*args: t.Any, **kwargs: t.Any) -> t.Callable[[Snowflake.Parser], exp.Show]:
 171    def _parse(self: Snowflake.Parser) -> exp.Show:
 172        return self._parse_show_snowflake(*args, **kwargs)
 173
 174    return _parse
 175
 176
 177def _date_trunc_to_time(args: t.List) -> exp.DateTrunc | exp.TimestampTrunc:
 178    trunc = date_trunc_to_time(args)
 179    trunc.set("unit", map_date_part(trunc.args["unit"]))
 180    return trunc
 181
 182
 183def _unqualify_pivot_columns(expression: exp.Expression) -> exp.Expression:
 184    """
 185    Snowflake doesn't allow columns referenced in UNPIVOT to be qualified,
 186    so we need to unqualify them. Same goes for ANY ORDER BY <column>.
 187
 188    Example:
 189        >>> from sqlglot import parse_one
 190        >>> expr = parse_one("SELECT * FROM m_sales UNPIVOT(sales FOR month IN (m_sales.jan, feb, mar, april))")
 191        >>> print(_unqualify_pivot_columns(expr).sql(dialect="snowflake"))
 192        SELECT * FROM m_sales UNPIVOT(sales FOR month IN (jan, feb, mar, april))
 193    """
 194    if isinstance(expression, exp.Pivot):
 195        if expression.unpivot:
 196            expression = transforms.unqualify_columns(expression)
 197        else:
 198            for field in expression.fields:
 199                field_expr = seq_get(field.expressions if field else [], 0)
 200
 201                if isinstance(field_expr, exp.PivotAny):
 202                    unqualified_field_expr = transforms.unqualify_columns(field_expr)
 203                    t.cast(exp.Expression, field).set("expressions", unqualified_field_expr, 0)
 204
 205    return expression
 206
 207
 208def _flatten_structured_types_unless_iceberg(expression: exp.Expression) -> exp.Expression:
 209    assert isinstance(expression, exp.Create)
 210
 211    def _flatten_structured_type(expression: exp.DataType) -> exp.DataType:
 212        if expression.this in exp.DataType.NESTED_TYPES:
 213            expression.set("expressions", None)
 214        return expression
 215
 216    props = expression.args.get("properties")
 217    if isinstance(expression.this, exp.Schema) and not (props and props.find(exp.IcebergProperty)):
 218        for schema_expression in expression.this.expressions:
 219            if isinstance(schema_expression, exp.ColumnDef):
 220                column_type = schema_expression.kind
 221                if isinstance(column_type, exp.DataType):
 222                    column_type.transform(_flatten_structured_type, copy=False)
 223
 224    return expression
 225
 226
 227def _unnest_generate_date_array(unnest: exp.Unnest) -> None:
 228    generate_date_array = unnest.expressions[0]
 229    start = generate_date_array.args.get("start")
 230    end = generate_date_array.args.get("end")
 231    step = generate_date_array.args.get("step")
 232
 233    if not start or not end or not isinstance(step, exp.Interval) or step.name != "1":
 234        return
 235
 236    unit = step.args.get("unit")
 237
 238    unnest_alias = unnest.args.get("alias")
 239    if unnest_alias:
 240        unnest_alias = unnest_alias.copy()
 241        sequence_value_name = seq_get(unnest_alias.columns, 0) or "value"
 242    else:
 243        sequence_value_name = "value"
 244
 245    # We'll add the next sequence value to the starting date and project the result
 246    date_add = _build_date_time_add(exp.DateAdd)(
 247        [unit, exp.cast(sequence_value_name, "int"), exp.cast(start, "date")]
 248    ).as_(sequence_value_name)
 249
 250    # We use DATEDIFF to compute the number of sequence values needed
 251    number_sequence = Snowflake.Parser.FUNCTIONS["ARRAY_GENERATE_RANGE"](
 252        [exp.Literal.number(0), _build_datediff([unit, start, end]) + 1]
 253    )
 254
 255    unnest.set("expressions", [number_sequence])
 256    unnest.replace(exp.select(date_add).from_(unnest.copy()).subquery(unnest_alias))
 257
 258
 259def _transform_generate_date_array(expression: exp.Expression) -> exp.Expression:
 260    if isinstance(expression, exp.Select):
 261        for generate_date_array in expression.find_all(exp.GenerateDateArray):
 262            parent = generate_date_array.parent
 263
 264            # If GENERATE_DATE_ARRAY is used directly as an array (e.g passed into ARRAY_LENGTH), the transformed Snowflake
 265            # query is the following (it'll be unnested properly on the next iteration due to copy):
 266            # SELECT ref(GENERATE_DATE_ARRAY(...)) -> SELECT ref((SELECT ARRAY_AGG(*) FROM UNNEST(GENERATE_DATE_ARRAY(...))))
 267            if not isinstance(parent, exp.Unnest):
 268                unnest = exp.Unnest(expressions=[generate_date_array.copy()])
 269                generate_date_array.replace(
 270                    exp.select(exp.ArrayAgg(this=exp.Star())).from_(unnest).subquery()
 271                )
 272
 273            if (
 274                isinstance(parent, exp.Unnest)
 275                and isinstance(parent.parent, (exp.From, exp.Join))
 276                and len(parent.expressions) == 1
 277            ):
 278                _unnest_generate_date_array(parent)
 279
 280    return expression
 281
 282
 283def _build_regexp_extract(expr_type: t.Type[E]) -> t.Callable[[t.List], E]:
 284    def _builder(args: t.List) -> E:
 285        return expr_type(
 286            this=seq_get(args, 0),
 287            expression=seq_get(args, 1),
 288            position=seq_get(args, 2),
 289            occurrence=seq_get(args, 3),
 290            parameters=seq_get(args, 4),
 291            group=seq_get(args, 5) or exp.Literal.number(0),
 292        )
 293
 294    return _builder
 295
 296
 297def _regexpextract_sql(self, expression: exp.RegexpExtract | exp.RegexpExtractAll) -> str:
 298    # Other dialects don't support all of the following parameters, so we need to
 299    # generate default values as necessary to ensure the transpilation is correct
 300    group = expression.args.get("group")
 301
 302    # To avoid generating all these default values, we set group to None if
 303    # it's 0 (also default value) which doesn't trigger the following chain
 304    if group and group.name == "0":
 305        group = None
 306
 307    parameters = expression.args.get("parameters") or (group and exp.Literal.string("c"))
 308    occurrence = expression.args.get("occurrence") or (parameters and exp.Literal.number(1))
 309    position = expression.args.get("position") or (occurrence and exp.Literal.number(1))
 310
 311    return self.func(
 312        "REGEXP_SUBSTR" if isinstance(expression, exp.RegexpExtract) else "REGEXP_EXTRACT_ALL",
 313        expression.this,
 314        expression.expression,
 315        position,
 316        occurrence,
 317        parameters,
 318        group,
 319    )
 320
 321
 322def _json_extract_value_array_sql(
 323    self: Snowflake.Generator, expression: exp.JSONValueArray | exp.JSONExtractArray
 324) -> str:
 325    json_extract = exp.JSONExtract(this=expression.this, expression=expression.expression)
 326    ident = exp.to_identifier("x")
 327
 328    if isinstance(expression, exp.JSONValueArray):
 329        this: exp.Expression = exp.cast(ident, to=exp.DataType.Type.VARCHAR)
 330    else:
 331        this = exp.ParseJSON(this=f"TO_JSON({ident})")
 332
 333    transform_lambda = exp.Lambda(expressions=[ident], this=this)
 334
 335    return self.func("TRANSFORM", json_extract, transform_lambda)
 336
 337
 338def _eliminate_dot_variant_lookup(expression: exp.Expression) -> exp.Expression:
 339    if isinstance(expression, exp.Select):
 340        # This transformation is used to facilitate transpilation of BigQuery `UNNEST` operations
 341        # to Snowflake. It should not affect roundtrip because `Unnest` nodes cannot be produced
 342        # by Snowflake's parser.
 343        #
 344        # Additionally, at the time of writing this, BigQuery is the only dialect that produces a
 345        # `TableAlias` node that only fills `columns` and not `this`, due to `UNNEST_COLUMN_ONLY`.
 346        unnest_aliases = set()
 347        for unnest in find_all_in_scope(expression, exp.Unnest):
 348            unnest_alias = unnest.args.get("alias")
 349            if (
 350                isinstance(unnest_alias, exp.TableAlias)
 351                and not unnest_alias.this
 352                and len(unnest_alias.columns) == 1
 353            ):
 354                unnest_aliases.add(unnest_alias.columns[0].name)
 355
 356        if unnest_aliases:
 357            for c in find_all_in_scope(expression, exp.Column):
 358                if c.table in unnest_aliases:
 359                    bracket_lhs = c.args["table"]
 360                    bracket_rhs = exp.Literal.string(c.name)
 361                    c.replace(exp.Bracket(this=bracket_lhs, expressions=[bracket_rhs]))
 362
 363    return expression
 364
 365
 366class Snowflake(Dialect):
 367    # https://docs.snowflake.com/en/sql-reference/identifiers-syntax
 368    NORMALIZATION_STRATEGY = NormalizationStrategy.UPPERCASE
 369    NULL_ORDERING = "nulls_are_large"
 370    TIME_FORMAT = "'YYYY-MM-DD HH24:MI:SS'"
 371    SUPPORTS_USER_DEFINED_TYPES = False
 372    SUPPORTS_SEMI_ANTI_JOIN = False
 373    PREFER_CTE_ALIAS_COLUMN = True
 374    TABLESAMPLE_SIZE_IS_PERCENT = True
 375    COPY_PARAMS_ARE_CSV = False
 376    ARRAY_AGG_INCLUDES_NULLS = None
 377
 378    TIME_MAPPING = {
 379        "YYYY": "%Y",
 380        "yyyy": "%Y",
 381        "YY": "%y",
 382        "yy": "%y",
 383        "MMMM": "%B",
 384        "mmmm": "%B",
 385        "MON": "%b",
 386        "mon": "%b",
 387        "MM": "%m",
 388        "mm": "%m",
 389        "DD": "%d",
 390        "dd": "%-d",
 391        "DY": "%a",
 392        "dy": "%w",
 393        "HH24": "%H",
 394        "hh24": "%H",
 395        "HH12": "%I",
 396        "hh12": "%I",
 397        "MI": "%M",
 398        "mi": "%M",
 399        "SS": "%S",
 400        "ss": "%S",
 401        "FF6": "%f",
 402        "ff6": "%f",
 403    }
 404
 405    DATE_PART_MAPPING = {
 406        **Dialect.DATE_PART_MAPPING,
 407        "ISOWEEK": "WEEKISO",
 408    }
 409
 410    def quote_identifier(self, expression: E, identify: bool = True) -> E:
 411        # This disables quoting DUAL in SELECT ... FROM DUAL, because Snowflake treats an
 412        # unquoted DUAL keyword in a special way and does not map it to a user-defined table
 413        if (
 414            isinstance(expression, exp.Identifier)
 415            and isinstance(expression.parent, exp.Table)
 416            and expression.name.lower() == "dual"
 417        ):
 418            return expression  # type: ignore
 419
 420        return super().quote_identifier(expression, identify=identify)
 421
 422    class JSONPathTokenizer(jsonpath.JSONPathTokenizer):
 423        SINGLE_TOKENS = jsonpath.JSONPathTokenizer.SINGLE_TOKENS.copy()
 424        SINGLE_TOKENS.pop("$")
 425
 426    class Parser(parser.Parser):
 427        IDENTIFY_PIVOT_STRINGS = True
 428        DEFAULT_SAMPLING_METHOD = "BERNOULLI"
 429        COLON_IS_VARIANT_EXTRACT = True
 430
 431        ID_VAR_TOKENS = {
 432            *parser.Parser.ID_VAR_TOKENS,
 433            TokenType.MATCH_CONDITION,
 434        }
 435
 436        TABLE_ALIAS_TOKENS = parser.Parser.TABLE_ALIAS_TOKENS | {TokenType.WINDOW}
 437        TABLE_ALIAS_TOKENS.discard(TokenType.MATCH_CONDITION)
 438
 439        COLON_PLACEHOLDER_TOKENS = ID_VAR_TOKENS | {TokenType.NUMBER}
 440
 441        FUNCTIONS = {
 442            **parser.Parser.FUNCTIONS,
 443            "APPROX_PERCENTILE": exp.ApproxQuantile.from_arg_list,
 444            "ARRAY_CONSTRUCT": lambda args: exp.Array(expressions=args),
 445            "ARRAY_CONTAINS": lambda args: exp.ArrayContains(
 446                this=seq_get(args, 1), expression=seq_get(args, 0)
 447            ),
 448            "ARRAY_GENERATE_RANGE": lambda args: exp.GenerateSeries(
 449                # ARRAY_GENERATE_RANGE has an exlusive end; we normalize it to be inclusive
 450                start=seq_get(args, 0),
 451                end=exp.Sub(this=seq_get(args, 1), expression=exp.Literal.number(1)),
 452                step=seq_get(args, 2),
 453            ),
 454            "BITXOR": _build_bitwise(exp.BitwiseXor, "BITXOR"),
 455            "BIT_XOR": _build_bitwise(exp.BitwiseXor, "BITXOR"),
 456            "BITOR": _build_bitwise(exp.BitwiseOr, "BITOR"),
 457            "BIT_OR": _build_bitwise(exp.BitwiseOr, "BITOR"),
 458            "BITSHIFTLEFT": _build_bitwise(exp.BitwiseLeftShift, "BITSHIFTLEFT"),
 459            "BIT_SHIFTLEFT": _build_bitwise(exp.BitwiseLeftShift, "BIT_SHIFTLEFT"),
 460            "BITSHIFTRIGHT": _build_bitwise(exp.BitwiseRightShift, "BITSHIFTRIGHT"),
 461            "BIT_SHIFTRIGHT": _build_bitwise(exp.BitwiseRightShift, "BIT_SHIFTRIGHT"),
 462            "BOOLXOR": _build_bitwise(exp.Xor, "BOOLXOR"),
 463            "DATE": _build_datetime("DATE", exp.DataType.Type.DATE),
 464            "DATE_TRUNC": _date_trunc_to_time,
 465            "DATEADD": _build_date_time_add(exp.DateAdd),
 466            "DATEDIFF": _build_datediff,
 467            "DIV0": _build_if_from_div0,
 468            "EDITDISTANCE": lambda args: exp.Levenshtein(
 469                this=seq_get(args, 0), expression=seq_get(args, 1), max_dist=seq_get(args, 2)
 470            ),
 471            "FLATTEN": exp.Explode.from_arg_list,
 472            "GET_PATH": lambda args, dialect: exp.JSONExtract(
 473                this=seq_get(args, 0), expression=dialect.to_json_path(seq_get(args, 1))
 474            ),
 475            "HEX_DECODE_BINARY": exp.Unhex.from_arg_list,
 476            "IFF": exp.If.from_arg_list,
 477            "LAST_DAY": lambda args: exp.LastDay(
 478                this=seq_get(args, 0), unit=map_date_part(seq_get(args, 1))
 479            ),
 480            "LEN": lambda args: exp.Length(this=seq_get(args, 0), binary=True),
 481            "LENGTH": lambda args: exp.Length(this=seq_get(args, 0), binary=True),
 482            "NULLIFZERO": _build_if_from_nullifzero,
 483            "OBJECT_CONSTRUCT": _build_object_construct,
 484            "REGEXP_EXTRACT_ALL": _build_regexp_extract(exp.RegexpExtractAll),
 485            "REGEXP_REPLACE": _build_regexp_replace,
 486            "REGEXP_SUBSTR": _build_regexp_extract(exp.RegexpExtract),
 487            "REGEXP_SUBSTR_ALL": _build_regexp_extract(exp.RegexpExtractAll),
 488            "REPLACE": build_replace_with_optional_replacement,
 489            "RLIKE": exp.RegexpLike.from_arg_list,
 490            "SQUARE": lambda args: exp.Pow(this=seq_get(args, 0), expression=exp.Literal.number(2)),
 491            "TABLE": lambda args: exp.TableFromRows(this=seq_get(args, 0)),
 492            "TIMEADD": _build_date_time_add(exp.TimeAdd),
 493            "TIMEDIFF": _build_datediff,
 494            "TIMESTAMPADD": _build_date_time_add(exp.DateAdd),
 495            "TIMESTAMPDIFF": _build_datediff,
 496            "TIMESTAMPFROMPARTS": build_timestamp_from_parts,
 497            "TIMESTAMP_FROM_PARTS": build_timestamp_from_parts,
 498            "TIMESTAMPNTZFROMPARTS": build_timestamp_from_parts,
 499            "TIMESTAMP_NTZ_FROM_PARTS": build_timestamp_from_parts,
 500            "TRY_PARSE_JSON": lambda args: exp.ParseJSON(this=seq_get(args, 0), safe=True),
 501            "TRY_TO_DATE": _build_datetime("TRY_TO_DATE", exp.DataType.Type.DATE, safe=True),
 502            "TRY_TO_TIME": _build_datetime("TRY_TO_TIME", exp.DataType.Type.TIME, safe=True),
 503            "TRY_TO_TIMESTAMP": _build_datetime(
 504                "TRY_TO_TIMESTAMP", exp.DataType.Type.TIMESTAMP, safe=True
 505            ),
 506            "TO_CHAR": build_timetostr_or_tochar,
 507            "TO_DATE": _build_datetime("TO_DATE", exp.DataType.Type.DATE),
 508            "TO_NUMBER": lambda args: exp.ToNumber(
 509                this=seq_get(args, 0),
 510                format=seq_get(args, 1),
 511                precision=seq_get(args, 2),
 512                scale=seq_get(args, 3),
 513            ),
 514            "TO_TIME": _build_datetime("TO_TIME", exp.DataType.Type.TIME),
 515            "TO_TIMESTAMP": _build_datetime("TO_TIMESTAMP", exp.DataType.Type.TIMESTAMP),
 516            "TO_TIMESTAMP_LTZ": _build_datetime("TO_TIMESTAMP_LTZ", exp.DataType.Type.TIMESTAMPLTZ),
 517            "TO_TIMESTAMP_NTZ": _build_datetime("TO_TIMESTAMP_NTZ", exp.DataType.Type.TIMESTAMP),
 518            "TO_TIMESTAMP_TZ": _build_datetime("TO_TIMESTAMP_TZ", exp.DataType.Type.TIMESTAMPTZ),
 519            "TO_VARCHAR": exp.ToChar.from_arg_list,
 520            "ZEROIFNULL": _build_if_from_zeroifnull,
 521        }
 522
 523        FUNCTION_PARSERS = {
 524            **parser.Parser.FUNCTION_PARSERS,
 525            "DATE_PART": lambda self: self._parse_date_part(),
 526            "OBJECT_CONSTRUCT_KEEP_NULL": lambda self: self._parse_json_object(),
 527            "LISTAGG": lambda self: self._parse_string_agg(),
 528        }
 529        FUNCTION_PARSERS.pop("TRIM")
 530
 531        TIMESTAMPS = parser.Parser.TIMESTAMPS - {TokenType.TIME}
 532
 533        RANGE_PARSERS = {
 534            **parser.Parser.RANGE_PARSERS,
 535            TokenType.LIKE_ANY: parser.binary_range_parser(exp.LikeAny),
 536            TokenType.ILIKE_ANY: parser.binary_range_parser(exp.ILikeAny),
 537        }
 538
 539        ALTER_PARSERS = {
 540            **parser.Parser.ALTER_PARSERS,
 541            "UNSET": lambda self: self.expression(
 542                exp.Set,
 543                tag=self._match_text_seq("TAG"),
 544                expressions=self._parse_csv(self._parse_id_var),
 545                unset=True,
 546            ),
 547        }
 548
 549        STATEMENT_PARSERS = {
 550            **parser.Parser.STATEMENT_PARSERS,
 551            TokenType.GET: lambda self: self._parse_get(),
 552            TokenType.PUT: lambda self: self._parse_put(),
 553            TokenType.SHOW: lambda self: self._parse_show(),
 554        }
 555
 556        PROPERTY_PARSERS = {
 557            **parser.Parser.PROPERTY_PARSERS,
 558            "CREDENTIALS": lambda self: self._parse_credentials_property(),
 559            "FILE_FORMAT": lambda self: self._parse_file_format_property(),
 560            "LOCATION": lambda self: self._parse_location_property(),
 561            "TAG": lambda self: self._parse_tag(),
 562            "USING": lambda self: self._match_text_seq("TEMPLATE")
 563            and self.expression(exp.UsingTemplateProperty, this=self._parse_statement()),
 564        }
 565
 566        TYPE_CONVERTERS = {
 567            # https://docs.snowflake.com/en/sql-reference/data-types-numeric#number
 568            exp.DataType.Type.DECIMAL: build_default_decimal_type(precision=38, scale=0),
 569        }
 570
 571        SHOW_PARSERS = {
 572            "DATABASES": _show_parser("DATABASES"),
 573            "TERSE DATABASES": _show_parser("DATABASES"),
 574            "SCHEMAS": _show_parser("SCHEMAS"),
 575            "TERSE SCHEMAS": _show_parser("SCHEMAS"),
 576            "OBJECTS": _show_parser("OBJECTS"),
 577            "TERSE OBJECTS": _show_parser("OBJECTS"),
 578            "TABLES": _show_parser("TABLES"),
 579            "TERSE TABLES": _show_parser("TABLES"),
 580            "VIEWS": _show_parser("VIEWS"),
 581            "TERSE VIEWS": _show_parser("VIEWS"),
 582            "PRIMARY KEYS": _show_parser("PRIMARY KEYS"),
 583            "TERSE PRIMARY KEYS": _show_parser("PRIMARY KEYS"),
 584            "IMPORTED KEYS": _show_parser("IMPORTED KEYS"),
 585            "TERSE IMPORTED KEYS": _show_parser("IMPORTED KEYS"),
 586            "UNIQUE KEYS": _show_parser("UNIQUE KEYS"),
 587            "TERSE UNIQUE KEYS": _show_parser("UNIQUE KEYS"),
 588            "SEQUENCES": _show_parser("SEQUENCES"),
 589            "TERSE SEQUENCES": _show_parser("SEQUENCES"),
 590            "STAGES": _show_parser("STAGES"),
 591            "COLUMNS": _show_parser("COLUMNS"),
 592            "USERS": _show_parser("USERS"),
 593            "TERSE USERS": _show_parser("USERS"),
 594            "FILE FORMATS": _show_parser("FILE FORMATS"),
 595            "FUNCTIONS": _show_parser("FUNCTIONS"),
 596            "PROCEDURES": _show_parser("PROCEDURES"),
 597            "WAREHOUSES": _show_parser("WAREHOUSES"),
 598        }
 599
 600        CONSTRAINT_PARSERS = {
 601            **parser.Parser.CONSTRAINT_PARSERS,
 602            "WITH": lambda self: self._parse_with_constraint(),
 603            "MASKING": lambda self: self._parse_with_constraint(),
 604            "PROJECTION": lambda self: self._parse_with_constraint(),
 605            "TAG": lambda self: self._parse_with_constraint(),
 606        }
 607
 608        STAGED_FILE_SINGLE_TOKENS = {
 609            TokenType.DOT,
 610            TokenType.MOD,
 611            TokenType.SLASH,
 612        }
 613
 614        FLATTEN_COLUMNS = ["SEQ", "KEY", "PATH", "INDEX", "VALUE", "THIS"]
 615
 616        SCHEMA_KINDS = {"OBJECTS", "TABLES", "VIEWS", "SEQUENCES", "UNIQUE KEYS", "IMPORTED KEYS"}
 617
 618        NON_TABLE_CREATABLES = {"STORAGE INTEGRATION", "TAG", "WAREHOUSE", "STREAMLIT"}
 619
 620        LAMBDAS = {
 621            **parser.Parser.LAMBDAS,
 622            TokenType.ARROW: lambda self, expressions: self.expression(
 623                exp.Lambda,
 624                this=self._replace_lambda(
 625                    self._parse_assignment(),
 626                    expressions,
 627                ),
 628                expressions=[e.this if isinstance(e, exp.Cast) else e for e in expressions],
 629            ),
 630        }
 631
 632        def _parse_use(self) -> exp.Use:
 633            if self._match_text_seq("SECONDARY", "ROLES"):
 634                this = self._match_texts(("ALL", "NONE")) and exp.var(self._prev.text.upper())
 635                roles = None if this else self._parse_csv(lambda: self._parse_table(schema=False))
 636                return self.expression(
 637                    exp.Use, kind="SECONDARY ROLES", this=this, expressions=roles
 638                )
 639
 640            return super()._parse_use()
 641
 642        def _negate_range(
 643            self, this: t.Optional[exp.Expression] = None
 644        ) -> t.Optional[exp.Expression]:
 645            if not this:
 646                return this
 647
 648            query = this.args.get("query")
 649            if isinstance(this, exp.In) and isinstance(query, exp.Query):
 650                # Snowflake treats `value NOT IN (subquery)` as `VALUE <> ALL (subquery)`, so
 651                # we do this conversion here to avoid parsing it into `NOT value IN (subquery)`
 652                # which can produce different results (most likely a SnowFlake bug).
 653                #
 654                # https://docs.snowflake.com/en/sql-reference/functions/in
 655                # Context: https://github.com/tobymao/sqlglot/issues/3890
 656                return self.expression(
 657                    exp.NEQ, this=this.this, expression=exp.All(this=query.unnest())
 658                )
 659
 660            return self.expression(exp.Not, this=this)
 661
 662        def _parse_tag(self) -> exp.Tags:
 663            return self.expression(
 664                exp.Tags,
 665                expressions=self._parse_wrapped_csv(self._parse_property),
 666            )
 667
 668        def _parse_with_constraint(self) -> t.Optional[exp.Expression]:
 669            if self._prev.token_type != TokenType.WITH:
 670                self._retreat(self._index - 1)
 671
 672            if self._match_text_seq("MASKING", "POLICY"):
 673                policy = self._parse_column()
 674                return self.expression(
 675                    exp.MaskingPolicyColumnConstraint,
 676                    this=policy.to_dot() if isinstance(policy, exp.Column) else policy,
 677                    expressions=self._match(TokenType.USING)
 678                    and self._parse_wrapped_csv(self._parse_id_var),
 679                )
 680            if self._match_text_seq("PROJECTION", "POLICY"):
 681                policy = self._parse_column()
 682                return self.expression(
 683                    exp.ProjectionPolicyColumnConstraint,
 684                    this=policy.to_dot() if isinstance(policy, exp.Column) else policy,
 685                )
 686            if self._match(TokenType.TAG):
 687                return self._parse_tag()
 688
 689            return None
 690
 691        def _parse_with_property(self) -> t.Optional[exp.Expression] | t.List[exp.Expression]:
 692            if self._match(TokenType.TAG):
 693                return self._parse_tag()
 694
 695            return super()._parse_with_property()
 696
 697        def _parse_create(self) -> exp.Create | exp.Command:
 698            expression = super()._parse_create()
 699            if isinstance(expression, exp.Create) and expression.kind in self.NON_TABLE_CREATABLES:
 700                # Replace the Table node with the enclosed Identifier
 701                expression.this.replace(expression.this.this)
 702
 703            return expression
 704
 705        # https://docs.snowflake.com/en/sql-reference/functions/date_part.html
 706        # https://docs.snowflake.com/en/sql-reference/functions-date-time.html#label-supported-date-time-parts
 707        def _parse_date_part(self: Snowflake.Parser) -> t.Optional[exp.Expression]:
 708            this = self._parse_var() or self._parse_type()
 709
 710            if not this:
 711                return None
 712
 713            self._match(TokenType.COMMA)
 714            expression = self._parse_bitwise()
 715            this = map_date_part(this)
 716            name = this.name.upper()
 717
 718            if name.startswith("EPOCH"):
 719                if name == "EPOCH_MILLISECOND":
 720                    scale = 10**3
 721                elif name == "EPOCH_MICROSECOND":
 722                    scale = 10**6
 723                elif name == "EPOCH_NANOSECOND":
 724                    scale = 10**9
 725                else:
 726                    scale = None
 727
 728                ts = self.expression(exp.Cast, this=expression, to=exp.DataType.build("TIMESTAMP"))
 729                to_unix: exp.Expression = self.expression(exp.TimeToUnix, this=ts)
 730
 731                if scale:
 732                    to_unix = exp.Mul(this=to_unix, expression=exp.Literal.number(scale))
 733
 734                return to_unix
 735
 736            return self.expression(exp.Extract, this=this, expression=expression)
 737
 738        def _parse_bracket_key_value(self, is_map: bool = False) -> t.Optional[exp.Expression]:
 739            if is_map:
 740                # Keys are strings in Snowflake's objects, see also:
 741                # - https://docs.snowflake.com/en/sql-reference/data-types-semistructured
 742                # - https://docs.snowflake.com/en/sql-reference/functions/object_construct
 743                return self._parse_slice(self._parse_string())
 744
 745            return self._parse_slice(self._parse_alias(self._parse_assignment(), explicit=True))
 746
 747        def _parse_lateral(self) -> t.Optional[exp.Lateral]:
 748            lateral = super()._parse_lateral()
 749            if not lateral:
 750                return lateral
 751
 752            if isinstance(lateral.this, exp.Explode):
 753                table_alias = lateral.args.get("alias")
 754                columns = [exp.to_identifier(col) for col in self.FLATTEN_COLUMNS]
 755                if table_alias and not table_alias.args.get("columns"):
 756                    table_alias.set("columns", columns)
 757                elif not table_alias:
 758                    exp.alias_(lateral, "_flattened", table=columns, copy=False)
 759
 760            return lateral
 761
 762        def _parse_table_parts(
 763            self, schema: bool = False, is_db_reference: bool = False, wildcard: bool = False
 764        ) -> exp.Table:
 765            # https://docs.snowflake.com/en/user-guide/querying-stage
 766            if self._match(TokenType.STRING, advance=False):
 767                table = self._parse_string()
 768            elif self._match_text_seq("@", advance=False):
 769                table = self._parse_location_path()
 770            else:
 771                table = None
 772
 773            if table:
 774                file_format = None
 775                pattern = None
 776
 777                wrapped = self._match(TokenType.L_PAREN)
 778                while self._curr and wrapped and not self._match(TokenType.R_PAREN):
 779                    if self._match_text_seq("FILE_FORMAT", "=>"):
 780                        file_format = self._parse_string() or super()._parse_table_parts(
 781                            is_db_reference=is_db_reference
 782                        )
 783                    elif self._match_text_seq("PATTERN", "=>"):
 784                        pattern = self._parse_string()
 785                    else:
 786                        break
 787
 788                    self._match(TokenType.COMMA)
 789
 790                table = self.expression(exp.Table, this=table, format=file_format, pattern=pattern)
 791            else:
 792                table = super()._parse_table_parts(schema=schema, is_db_reference=is_db_reference)
 793
 794            return table
 795
 796        def _parse_table(
 797            self,
 798            schema: bool = False,
 799            joins: bool = False,
 800            alias_tokens: t.Optional[t.Collection[TokenType]] = None,
 801            parse_bracket: bool = False,
 802            is_db_reference: bool = False,
 803            parse_partition: bool = False,
 804            consume_pipe: bool = False,
 805        ) -> t.Optional[exp.Expression]:
 806            table = super()._parse_table(
 807                schema=schema,
 808                joins=joins,
 809                alias_tokens=alias_tokens,
 810                parse_bracket=parse_bracket,
 811                is_db_reference=is_db_reference,
 812                parse_partition=parse_partition,
 813            )
 814            if isinstance(table, exp.Table) and isinstance(table.this, exp.TableFromRows):
 815                table_from_rows = table.this
 816                for arg in exp.TableFromRows.arg_types:
 817                    if arg != "this":
 818                        table_from_rows.set(arg, table.args.get(arg))
 819
 820                table = table_from_rows
 821
 822            return table
 823
 824        def _parse_id_var(
 825            self,
 826            any_token: bool = True,
 827            tokens: t.Optional[t.Collection[TokenType]] = None,
 828        ) -> t.Optional[exp.Expression]:
 829            if self._match_text_seq("IDENTIFIER", "("):
 830                identifier = (
 831                    super()._parse_id_var(any_token=any_token, tokens=tokens)
 832                    or self._parse_string()
 833                )
 834                self._match_r_paren()
 835                return self.expression(exp.Anonymous, this="IDENTIFIER", expressions=[identifier])
 836
 837            return super()._parse_id_var(any_token=any_token, tokens=tokens)
 838
 839        def _parse_show_snowflake(self, this: str) -> exp.Show:
 840            scope = None
 841            scope_kind = None
 842
 843            # will identity SHOW TERSE SCHEMAS but not SHOW TERSE PRIMARY KEYS
 844            # which is syntactically valid but has no effect on the output
 845            terse = self._tokens[self._index - 2].text.upper() == "TERSE"
 846
 847            history = self._match_text_seq("HISTORY")
 848
 849            like = self._parse_string() if self._match(TokenType.LIKE) else None
 850
 851            if self._match(TokenType.IN):
 852                if self._match_text_seq("ACCOUNT"):
 853                    scope_kind = "ACCOUNT"
 854                elif self._match_text_seq("CLASS"):
 855                    scope_kind = "CLASS"
 856                    scope = self._parse_table_parts()
 857                elif self._match_text_seq("APPLICATION"):
 858                    scope_kind = "APPLICATION"
 859                    if self._match_text_seq("PACKAGE"):
 860                        scope_kind += " PACKAGE"
 861                    scope = self._parse_table_parts()
 862                elif self._match_set(self.DB_CREATABLES):
 863                    scope_kind = self._prev.text.upper()
 864                    if self._curr:
 865                        scope = self._parse_table_parts()
 866                elif self._curr:
 867                    scope_kind = "SCHEMA" if this in self.SCHEMA_KINDS else "TABLE"
 868                    scope = self._parse_table_parts()
 869
 870            return self.expression(
 871                exp.Show,
 872                **{
 873                    "terse": terse,
 874                    "this": this,
 875                    "history": history,
 876                    "like": like,
 877                    "scope": scope,
 878                    "scope_kind": scope_kind,
 879                    "starts_with": self._match_text_seq("STARTS", "WITH") and self._parse_string(),
 880                    "limit": self._parse_limit(),
 881                    "from": self._parse_string() if self._match(TokenType.FROM) else None,
 882                    "privileges": self._match_text_seq("WITH", "PRIVILEGES")
 883                    and self._parse_csv(lambda: self._parse_var(any_token=True, upper=True)),
 884                },
 885            )
 886
 887        def _parse_put(self) -> exp.Put | exp.Command:
 888            if self._curr.token_type != TokenType.STRING:
 889                return self._parse_as_command(self._prev)
 890
 891            return self.expression(
 892                exp.Put,
 893                this=self._parse_string(),
 894                target=self._parse_location_path(),
 895                properties=self._parse_properties(),
 896            )
 897
 898        def _parse_get(self) -> t.Optional[exp.Expression]:
 899            start = self._prev
 900
 901            # If we detect GET( then we need to parse a function, not a statement
 902            if self._match(TokenType.L_PAREN):
 903                self._retreat(self._index - 2)
 904                return self._parse_expression()
 905
 906            target = self._parse_location_path()
 907
 908            # Parse as command if unquoted file path
 909            if self._curr.token_type == TokenType.URI_START:
 910                return self._parse_as_command(start)
 911
 912            return self.expression(
 913                exp.Get,
 914                this=self._parse_string(),
 915                target=target,
 916                properties=self._parse_properties(),
 917            )
 918
 919        def _parse_location_property(self) -> exp.LocationProperty:
 920            self._match(TokenType.EQ)
 921            return self.expression(exp.LocationProperty, this=self._parse_location_path())
 922
 923        def _parse_file_location(self) -> t.Optional[exp.Expression]:
 924            # Parse either a subquery or a staged file
 925            return (
 926                self._parse_select(table=True, parse_subquery_alias=False)
 927                if self._match(TokenType.L_PAREN, advance=False)
 928                else self._parse_table_parts()
 929            )
 930
 931        def _parse_location_path(self) -> exp.Var:
 932            start = self._curr
 933            self._advance_any(ignore_reserved=True)
 934
 935            # We avoid consuming a comma token because external tables like @foo and @bar
 936            # can be joined in a query with a comma separator, as well as closing paren
 937            # in case of subqueries
 938            while self._is_connected() and not self._match_set(
 939                (TokenType.COMMA, TokenType.L_PAREN, TokenType.R_PAREN), advance=False
 940            ):
 941                self._advance_any(ignore_reserved=True)
 942
 943            return exp.var(self._find_sql(start, self._prev))
 944
 945        def _parse_lambda_arg(self) -> t.Optional[exp.Expression]:
 946            this = super()._parse_lambda_arg()
 947
 948            if not this:
 949                return this
 950
 951            typ = self._parse_types()
 952
 953            if typ:
 954                return self.expression(exp.Cast, this=this, to=typ)
 955
 956            return this
 957
 958        def _parse_foreign_key(self) -> exp.ForeignKey:
 959            # inlineFK, the REFERENCES columns are implied
 960            if self._match(TokenType.REFERENCES, advance=False):
 961                return self.expression(exp.ForeignKey)
 962
 963            # outoflineFK, explicitly names the columns
 964            return super()._parse_foreign_key()
 965
 966        def _parse_file_format_property(self) -> exp.FileFormatProperty:
 967            self._match(TokenType.EQ)
 968            if self._match(TokenType.L_PAREN, advance=False):
 969                expressions = self._parse_wrapped_options()
 970            else:
 971                expressions = [self._parse_format_name()]
 972
 973            return self.expression(
 974                exp.FileFormatProperty,
 975                expressions=expressions,
 976            )
 977
 978        def _parse_credentials_property(self) -> exp.CredentialsProperty:
 979            return self.expression(
 980                exp.CredentialsProperty,
 981                expressions=self._parse_wrapped_options(),
 982            )
 983
 984    class Tokenizer(tokens.Tokenizer):
 985        STRING_ESCAPES = ["\\", "'"]
 986        HEX_STRINGS = [("x'", "'"), ("X'", "'")]
 987        RAW_STRINGS = ["$$"]
 988        COMMENTS = ["--", "//", ("/*", "*/")]
 989        NESTED_COMMENTS = False
 990
 991        KEYWORDS = {
 992            **tokens.Tokenizer.KEYWORDS,
 993            "FILE://": TokenType.URI_START,
 994            "BYTEINT": TokenType.INT,
 995            "EXCLUDE": TokenType.EXCEPT,
 996            "FILE FORMAT": TokenType.FILE_FORMAT,
 997            "GET": TokenType.GET,
 998            "ILIKE ANY": TokenType.ILIKE_ANY,
 999            "LIKE ANY": TokenType.LIKE_ANY,
1000            "MATCH_CONDITION": TokenType.MATCH_CONDITION,
1001            "MATCH_RECOGNIZE": TokenType.MATCH_RECOGNIZE,
1002            "MINUS": TokenType.EXCEPT,
1003            "NCHAR VARYING": TokenType.VARCHAR,
1004            "PUT": TokenType.PUT,
1005            "REMOVE": TokenType.COMMAND,
1006            "RM": TokenType.COMMAND,
1007            "SAMPLE": TokenType.TABLE_SAMPLE,
1008            "SQL_DOUBLE": TokenType.DOUBLE,
1009            "SQL_VARCHAR": TokenType.VARCHAR,
1010            "STORAGE INTEGRATION": TokenType.STORAGE_INTEGRATION,
1011            "TAG": TokenType.TAG,
1012            "TIMESTAMP_TZ": TokenType.TIMESTAMPTZ,
1013            "TOP": TokenType.TOP,
1014            "WAREHOUSE": TokenType.WAREHOUSE,
1015            "STAGE": TokenType.STAGE,
1016            "STREAMLIT": TokenType.STREAMLIT,
1017        }
1018        KEYWORDS.pop("/*+")
1019
1020        SINGLE_TOKENS = {
1021            **tokens.Tokenizer.SINGLE_TOKENS,
1022            "$": TokenType.PARAMETER,
1023        }
1024
1025        VAR_SINGLE_TOKENS = {"$"}
1026
1027        COMMANDS = tokens.Tokenizer.COMMANDS - {TokenType.SHOW}
1028
1029    class Generator(generator.Generator):
1030        PARAMETER_TOKEN = "$"
1031        MATCHED_BY_SOURCE = False
1032        SINGLE_STRING_INTERVAL = True
1033        JOIN_HINTS = False
1034        TABLE_HINTS = False
1035        QUERY_HINTS = False
1036        AGGREGATE_FILTER_SUPPORTED = False
1037        SUPPORTS_TABLE_COPY = False
1038        COLLATE_IS_FUNC = True
1039        LIMIT_ONLY_LITERALS = True
1040        JSON_KEY_VALUE_PAIR_SEP = ","
1041        INSERT_OVERWRITE = " OVERWRITE INTO"
1042        STRUCT_DELIMITER = ("(", ")")
1043        COPY_PARAMS_ARE_WRAPPED = False
1044        COPY_PARAMS_EQ_REQUIRED = True
1045        STAR_EXCEPT = "EXCLUDE"
1046        SUPPORTS_EXPLODING_PROJECTIONS = False
1047        ARRAY_CONCAT_IS_VAR_LEN = False
1048        SUPPORTS_CONVERT_TIMEZONE = True
1049        EXCEPT_INTERSECT_SUPPORT_ALL_CLAUSE = False
1050        SUPPORTS_MEDIAN = True
1051        ARRAY_SIZE_NAME = "ARRAY_SIZE"
1052
1053        TRANSFORMS = {
1054            **generator.Generator.TRANSFORMS,
1055            exp.ApproxDistinct: rename_func("APPROX_COUNT_DISTINCT"),
1056            exp.ArgMax: rename_func("MAX_BY"),
1057            exp.ArgMin: rename_func("MIN_BY"),
1058            exp.ArrayConcat: lambda self, e: self.arrayconcat_sql(e, name="ARRAY_CAT"),
1059            exp.ArrayContains: lambda self, e: self.func("ARRAY_CONTAINS", e.expression, e.this),
1060            exp.ArrayIntersect: rename_func("ARRAY_INTERSECTION"),
1061            exp.AtTimeZone: lambda self, e: self.func(
1062                "CONVERT_TIMEZONE", e.args.get("zone"), e.this
1063            ),
1064            exp.BitwiseOr: rename_func("BITOR"),
1065            exp.BitwiseXor: rename_func("BITXOR"),
1066            exp.BitwiseLeftShift: rename_func("BITSHIFTLEFT"),
1067            exp.BitwiseRightShift: rename_func("BITSHIFTRIGHT"),
1068            exp.Create: transforms.preprocess([_flatten_structured_types_unless_iceberg]),
1069            exp.DateAdd: date_delta_sql("DATEADD"),
1070            exp.DateDiff: date_delta_sql("DATEDIFF"),
1071            exp.DatetimeAdd: date_delta_sql("TIMESTAMPADD"),
1072            exp.DatetimeDiff: timestampdiff_sql,
1073            exp.DateStrToDate: datestrtodate_sql,
1074            exp.DayOfMonth: rename_func("DAYOFMONTH"),
1075            exp.DayOfWeek: rename_func("DAYOFWEEK"),
1076            exp.DayOfWeekIso: rename_func("DAYOFWEEKISO"),
1077            exp.DayOfYear: rename_func("DAYOFYEAR"),
1078            exp.Explode: rename_func("FLATTEN"),
1079            exp.Extract: lambda self, e: self.func(
1080                "DATE_PART", map_date_part(e.this, self.dialect), e.expression
1081            ),
1082            exp.FileFormatProperty: lambda self,
1083            e: f"FILE_FORMAT=({self.expressions(e, 'expressions', sep=' ')})",
1084            exp.FromTimeZone: lambda self, e: self.func(
1085                "CONVERT_TIMEZONE", e.args.get("zone"), "'UTC'", e.this
1086            ),
1087            exp.GenerateSeries: lambda self, e: self.func(
1088                "ARRAY_GENERATE_RANGE", e.args["start"], e.args["end"] + 1, e.args.get("step")
1089            ),
1090            exp.GroupConcat: lambda self, e: groupconcat_sql(self, e, sep=""),
1091            exp.If: if_sql(name="IFF", false_value="NULL"),
1092            exp.JSONExtractArray: _json_extract_value_array_sql,
1093            exp.JSONExtractScalar: lambda self, e: self.func(
1094                "JSON_EXTRACT_PATH_TEXT", e.this, e.expression
1095            ),
1096            exp.JSONObject: lambda self, e: self.func("OBJECT_CONSTRUCT_KEEP_NULL", *e.expressions),
1097            exp.JSONPathRoot: lambda *_: "",
1098            exp.JSONValueArray: _json_extract_value_array_sql,
1099            exp.Levenshtein: unsupported_args("ins_cost", "del_cost", "sub_cost")(
1100                rename_func("EDITDISTANCE")
1101            ),
1102            exp.LocationProperty: lambda self, e: f"LOCATION={self.sql(e, 'this')}",
1103            exp.LogicalAnd: rename_func("BOOLAND_AGG"),
1104            exp.LogicalOr: rename_func("BOOLOR_AGG"),
1105            exp.Map: lambda self, e: var_map_sql(self, e, "OBJECT_CONSTRUCT"),
1106            exp.MakeInterval: no_make_interval_sql,
1107            exp.Max: max_or_greatest,
1108            exp.Min: min_or_least,
1109            exp.ParseJSON: lambda self, e: self.func(
1110                "TRY_PARSE_JSON" if e.args.get("safe") else "PARSE_JSON", e.this
1111            ),
1112            exp.PartitionedByProperty: lambda self, e: f"PARTITION BY {self.sql(e, 'this')}",
1113            exp.PercentileCont: transforms.preprocess(
1114                [transforms.add_within_group_for_percentiles]
1115            ),
1116            exp.PercentileDisc: transforms.preprocess(
1117                [transforms.add_within_group_for_percentiles]
1118            ),
1119            exp.Pivot: transforms.preprocess([_unqualify_pivot_columns]),
1120            exp.RegexpExtract: _regexpextract_sql,
1121            exp.RegexpExtractAll: _regexpextract_sql,
1122            exp.RegexpILike: _regexpilike_sql,
1123            exp.Rand: rename_func("RANDOM"),
1124            exp.Select: transforms.preprocess(
1125                [
1126                    transforms.eliminate_window_clause,
1127                    transforms.eliminate_distinct_on,
1128                    transforms.explode_projection_to_unnest(),
1129                    transforms.eliminate_semi_and_anti_joins,
1130                    _transform_generate_date_array,
1131                    _eliminate_dot_variant_lookup,
1132                ]
1133            ),
1134            exp.SHA: rename_func("SHA1"),
1135            exp.StarMap: rename_func("OBJECT_CONSTRUCT"),
1136            exp.StartsWith: rename_func("STARTSWITH"),
1137            exp.EndsWith: rename_func("ENDSWITH"),
1138            exp.StrPosition: lambda self, e: strposition_sql(
1139                self, e, func_name="CHARINDEX", supports_position=True
1140            ),
1141            exp.StrToDate: lambda self, e: self.func("DATE", e.this, self.format_time(e)),
1142            exp.StringToArray: rename_func("STRTOK_TO_ARRAY"),
1143            exp.Stuff: rename_func("INSERT"),
1144            exp.StPoint: rename_func("ST_MAKEPOINT"),
1145            exp.TimeAdd: date_delta_sql("TIMEADD"),
1146            exp.Timestamp: no_timestamp_sql,
1147            exp.TimestampAdd: date_delta_sql("TIMESTAMPADD"),
1148            exp.TimestampDiff: lambda self, e: self.func(
1149                "TIMESTAMPDIFF", e.unit, e.expression, e.this
1150            ),
1151            exp.TimestampTrunc: timestamptrunc_sql(),
1152            exp.TimeStrToTime: timestrtotime_sql,
1153            exp.TimeToUnix: lambda self, e: f"EXTRACT(epoch_second FROM {self.sql(e, 'this')})",
1154            exp.ToArray: rename_func("TO_ARRAY"),
1155            exp.ToChar: lambda self, e: self.function_fallback_sql(e),
1156            exp.ToDouble: rename_func("TO_DOUBLE"),
1157            exp.TsOrDsAdd: date_delta_sql("DATEADD", cast=True),
1158            exp.TsOrDsDiff: date_delta_sql("DATEDIFF"),
1159            exp.TsOrDsToDate: lambda self, e: self.func(
1160                "TRY_TO_DATE" if e.args.get("safe") else "TO_DATE", e.this, self.format_time(e)
1161            ),
1162            exp.TsOrDsToTime: lambda self, e: self.func(
1163                "TRY_TO_TIME" if e.args.get("safe") else "TO_TIME", e.this, self.format_time(e)
1164            ),
1165            exp.Unhex: rename_func("HEX_DECODE_BINARY"),
1166            exp.UnixToTime: rename_func("TO_TIMESTAMP"),
1167            exp.Uuid: rename_func("UUID_STRING"),
1168            exp.VarMap: lambda self, e: var_map_sql(self, e, "OBJECT_CONSTRUCT"),
1169            exp.WeekOfYear: rename_func("WEEKOFYEAR"),
1170            exp.Xor: rename_func("BOOLXOR"),
1171        }
1172
1173        SUPPORTED_JSON_PATH_PARTS = {
1174            exp.JSONPathKey,
1175            exp.JSONPathRoot,
1176            exp.JSONPathSubscript,
1177        }
1178
1179        TYPE_MAPPING = {
1180            **generator.Generator.TYPE_MAPPING,
1181            exp.DataType.Type.NESTED: "OBJECT",
1182            exp.DataType.Type.STRUCT: "OBJECT",
1183            exp.DataType.Type.BIGDECIMAL: "DOUBLE",
1184        }
1185
1186        TOKEN_MAPPING = {
1187            TokenType.AUTO_INCREMENT: "AUTOINCREMENT",
1188        }
1189
1190        PROPERTIES_LOCATION = {
1191            **generator.Generator.PROPERTIES_LOCATION,
1192            exp.CredentialsProperty: exp.Properties.Location.POST_WITH,
1193            exp.LocationProperty: exp.Properties.Location.POST_WITH,
1194            exp.PartitionedByProperty: exp.Properties.Location.POST_SCHEMA,
1195            exp.SetProperty: exp.Properties.Location.UNSUPPORTED,
1196            exp.VolatileProperty: exp.Properties.Location.UNSUPPORTED,
1197        }
1198
1199        UNSUPPORTED_VALUES_EXPRESSIONS = {
1200            exp.Map,
1201            exp.StarMap,
1202            exp.Struct,
1203            exp.VarMap,
1204        }
1205
1206        RESPECT_IGNORE_NULLS_UNSUPPORTED_EXPRESSIONS = (exp.ArrayAgg,)
1207
1208        def with_properties(self, properties: exp.Properties) -> str:
1209            return self.properties(properties, wrapped=False, prefix=self.sep(""), sep=" ")
1210
1211        def values_sql(self, expression: exp.Values, values_as_table: bool = True) -> str:
1212            if expression.find(*self.UNSUPPORTED_VALUES_EXPRESSIONS):
1213                values_as_table = False
1214
1215            return super().values_sql(expression, values_as_table=values_as_table)
1216
1217        def datatype_sql(self, expression: exp.DataType) -> str:
1218            expressions = expression.expressions
1219            if (
1220                expressions
1221                and expression.is_type(*exp.DataType.STRUCT_TYPES)
1222                and any(isinstance(field_type, exp.DataType) for field_type in expressions)
1223            ):
1224                # The correct syntax is OBJECT [ (<key> <value_type [NOT NULL] [, ...]) ]
1225                return "OBJECT"
1226
1227            return super().datatype_sql(expression)
1228
1229        def tonumber_sql(self, expression: exp.ToNumber) -> str:
1230            return self.func(
1231                "TO_NUMBER",
1232                expression.this,
1233                expression.args.get("format"),
1234                expression.args.get("precision"),
1235                expression.args.get("scale"),
1236            )
1237
1238        def timestampfromparts_sql(self, expression: exp.TimestampFromParts) -> str:
1239            milli = expression.args.get("milli")
1240            if milli is not None:
1241                milli_to_nano = milli.pop() * exp.Literal.number(1000000)
1242                expression.set("nano", milli_to_nano)
1243
1244            return rename_func("TIMESTAMP_FROM_PARTS")(self, expression)
1245
1246        def cast_sql(self, expression: exp.Cast, safe_prefix: t.Optional[str] = None) -> str:
1247            if expression.is_type(exp.DataType.Type.GEOGRAPHY):
1248                return self.func("TO_GEOGRAPHY", expression.this)
1249            if expression.is_type(exp.DataType.Type.GEOMETRY):
1250                return self.func("TO_GEOMETRY", expression.this)
1251
1252            return super().cast_sql(expression, safe_prefix=safe_prefix)
1253
1254        def trycast_sql(self, expression: exp.TryCast) -> str:
1255            value = expression.this
1256
1257            if value.type is None:
1258                from sqlglot.optimizer.annotate_types import annotate_types
1259
1260                value = annotate_types(value, dialect=self.dialect)
1261
1262            if value.is_type(*exp.DataType.TEXT_TYPES, exp.DataType.Type.UNKNOWN):
1263                return super().trycast_sql(expression)
1264
1265            # TRY_CAST only works for string values in Snowflake
1266            return self.cast_sql(expression)
1267
1268        def log_sql(self, expression: exp.Log) -> str:
1269            if not expression.expression:
1270                return self.func("LN", expression.this)
1271
1272            return super().log_sql(expression)
1273
1274        def unnest_sql(self, expression: exp.Unnest) -> str:
1275            unnest_alias = expression.args.get("alias")
1276            offset = expression.args.get("offset")
1277
1278            unnest_alias_columns = unnest_alias.columns if unnest_alias else []
1279            value = seq_get(unnest_alias_columns, 0) or exp.to_identifier("value")
1280
1281            columns = [
1282                exp.to_identifier("seq"),
1283                exp.to_identifier("key"),
1284                exp.to_identifier("path"),
1285                offset.pop() if isinstance(offset, exp.Expression) else exp.to_identifier("index"),
1286                value,
1287                exp.to_identifier("this"),
1288            ]
1289
1290            if unnest_alias:
1291                unnest_alias.set("columns", columns)
1292            else:
1293                unnest_alias = exp.TableAlias(this="_u", columns=columns)
1294
1295            table_input = self.sql(expression.expressions[0])
1296            if not table_input.startswith("INPUT =>"):
1297                table_input = f"INPUT => {table_input}"
1298
1299            explode = f"TABLE(FLATTEN({table_input}))"
1300            alias = self.sql(unnest_alias)
1301            alias = f" AS {alias}" if alias else ""
1302            value = "" if isinstance(expression.parent, (exp.From, exp.Join)) else f"{value} FROM "
1303
1304            return f"{value}{explode}{alias}"
1305
1306        def show_sql(self, expression: exp.Show) -> str:
1307            terse = "TERSE " if expression.args.get("terse") else ""
1308            history = " HISTORY" if expression.args.get("history") else ""
1309            like = self.sql(expression, "like")
1310            like = f" LIKE {like}" if like else ""
1311
1312            scope = self.sql(expression, "scope")
1313            scope = f" {scope}" if scope else ""
1314
1315            scope_kind = self.sql(expression, "scope_kind")
1316            if scope_kind:
1317                scope_kind = f" IN {scope_kind}"
1318
1319            starts_with = self.sql(expression, "starts_with")
1320            if starts_with:
1321                starts_with = f" STARTS WITH {starts_with}"
1322
1323            limit = self.sql(expression, "limit")
1324
1325            from_ = self.sql(expression, "from")
1326            if from_:
1327                from_ = f" FROM {from_}"
1328
1329            privileges = self.expressions(expression, key="privileges", flat=True)
1330            privileges = f" WITH PRIVILEGES {privileges}" if privileges else ""
1331
1332            return f"SHOW {terse}{expression.name}{history}{like}{scope_kind}{scope}{starts_with}{limit}{from_}{privileges}"
1333
1334        def describe_sql(self, expression: exp.Describe) -> str:
1335            # Default to table if kind is unknown
1336            kind_value = expression.args.get("kind") or "TABLE"
1337            kind = f" {kind_value}" if kind_value else ""
1338            this = f" {self.sql(expression, 'this')}"
1339            expressions = self.expressions(expression, flat=True)
1340            expressions = f" {expressions}" if expressions else ""
1341            return f"DESCRIBE{kind}{this}{expressions}"
1342
1343        def generatedasidentitycolumnconstraint_sql(
1344            self, expression: exp.GeneratedAsIdentityColumnConstraint
1345        ) -> str:
1346            start = expression.args.get("start")
1347            start = f" START {start}" if start else ""
1348            increment = expression.args.get("increment")
1349            increment = f" INCREMENT {increment}" if increment else ""
1350
1351            order = expression.args.get("order")
1352            if order is not None:
1353                order_clause = " ORDER" if order else " NOORDER"
1354            else:
1355                order_clause = ""
1356
1357            return f"AUTOINCREMENT{start}{increment}{order_clause}"
1358
1359        def cluster_sql(self, expression: exp.Cluster) -> str:
1360            return f"CLUSTER BY ({self.expressions(expression, flat=True)})"
1361
1362        def struct_sql(self, expression: exp.Struct) -> str:
1363            keys = []
1364            values = []
1365
1366            for i, e in enumerate(expression.expressions):
1367                if isinstance(e, exp.PropertyEQ):
1368                    keys.append(
1369                        exp.Literal.string(e.name) if isinstance(e.this, exp.Identifier) else e.this
1370                    )
1371                    values.append(e.expression)
1372                else:
1373                    keys.append(exp.Literal.string(f"_{i}"))
1374                    values.append(e)
1375
1376            return self.func("OBJECT_CONSTRUCT", *flatten(zip(keys, values)))
1377
1378        @unsupported_args("weight", "accuracy")
1379        def approxquantile_sql(self, expression: exp.ApproxQuantile) -> str:
1380            return self.func("APPROX_PERCENTILE", expression.this, expression.args.get("quantile"))
1381
1382        def alterset_sql(self, expression: exp.AlterSet) -> str:
1383            exprs = self.expressions(expression, flat=True)
1384            exprs = f" {exprs}" if exprs else ""
1385            file_format = self.expressions(expression, key="file_format", flat=True, sep=" ")
1386            file_format = f" STAGE_FILE_FORMAT = ({file_format})" if file_format else ""
1387            copy_options = self.expressions(expression, key="copy_options", flat=True, sep=" ")
1388            copy_options = f" STAGE_COPY_OPTIONS = ({copy_options})" if copy_options else ""
1389            tag = self.expressions(expression, key="tag", flat=True)
1390            tag = f" TAG {tag}" if tag else ""
1391
1392            return f"SET{exprs}{file_format}{copy_options}{tag}"
1393
1394        def strtotime_sql(self, expression: exp.StrToTime):
1395            safe_prefix = "TRY_" if expression.args.get("safe") else ""
1396            return self.func(
1397                f"{safe_prefix}TO_TIMESTAMP", expression.this, self.format_time(expression)
1398            )
1399
1400        def timestampsub_sql(self, expression: exp.TimestampSub):
1401            return self.sql(
1402                exp.TimestampAdd(
1403                    this=expression.this,
1404                    expression=expression.expression * -1,
1405                    unit=expression.unit,
1406                )
1407            )
1408
1409        def jsonextract_sql(self, expression: exp.JSONExtract):
1410            this = expression.this
1411
1412            # JSON strings are valid coming from other dialects such as BQ
1413            return self.func(
1414                "GET_PATH",
1415                exp.ParseJSON(this=this) if this.is_string else this,
1416                expression.expression,
1417            )
1418
1419        def timetostr_sql(self, expression: exp.TimeToStr) -> str:
1420            this = expression.this
1421            if this.is_string:
1422                this = exp.cast(this, exp.DataType.Type.TIMESTAMP)
1423
1424            return self.func("TO_CHAR", this, self.format_time(expression))
1425
1426        def datesub_sql(self, expression: exp.DateSub) -> str:
1427            value = expression.expression
1428            if value:
1429                value.replace(value * (-1))
1430            else:
1431                self.unsupported("DateSub cannot be transpiled if the subtracted count is unknown")
1432
1433            return date_delta_sql("DATEADD")(self, expression)
1434
1435        def select_sql(self, expression: exp.Select) -> str:
1436            limit = expression.args.get("limit")
1437            offset = expression.args.get("offset")
1438            if offset and not limit:
1439                expression.limit(exp.Null(), copy=False)
1440            return super().select_sql(expression)
1441
1442        def createable_sql(self, expression: exp.Create, locations: t.DefaultDict) -> str:
1443            is_materialized = expression.find(exp.MaterializedProperty)
1444            copy_grants_property = expression.find(exp.CopyGrantsProperty)
1445
1446            if expression.kind == "VIEW" and is_materialized and copy_grants_property:
1447                # For materialized views, COPY GRANTS is located *before* the columns list
1448                # This is in contrast to normal views where COPY GRANTS is located *after* the columns list
1449                # We default CopyGrantsProperty to POST_SCHEMA which means we need to output it POST_NAME if a materialized view is detected
1450                # ref: https://docs.snowflake.com/en/sql-reference/sql/create-materialized-view#syntax
1451                # ref: https://docs.snowflake.com/en/sql-reference/sql/create-view#syntax
1452                post_schema_properties = locations[exp.Properties.Location.POST_SCHEMA]
1453                post_schema_properties.pop(post_schema_properties.index(copy_grants_property))
1454
1455                this_name = self.sql(expression.this, "this")
1456                copy_grants = self.sql(copy_grants_property)
1457                this_schema = self.schema_columns_sql(expression.this)
1458                this_schema = f"{self.sep()}{this_schema}" if this_schema else ""
1459
1460                return f"{this_name}{self.sep()}{copy_grants}{this_schema}"
1461
1462            return super().createable_sql(expression, locations)
1463
1464        def arrayagg_sql(self, expression: exp.ArrayAgg) -> str:
1465            this = expression.this
1466
1467            # If an ORDER BY clause is present, we need to remove it from ARRAY_AGG
1468            # and add it later as part of the WITHIN GROUP clause
1469            order = this if isinstance(this, exp.Order) else None
1470            if order:
1471                expression.set("this", order.this.pop())
1472
1473            expr_sql = super().arrayagg_sql(expression)
1474
1475            if order:
1476                expr_sql = self.sql(exp.WithinGroup(this=expr_sql, expression=order))
1477
1478            return expr_sql
1479
1480        def array_sql(self, expression: exp.Array) -> str:
1481            expressions = expression.expressions
1482
1483            first_expr = seq_get(expressions, 0)
1484            if isinstance(first_expr, exp.Select):
1485                # SELECT AS STRUCT foo AS alias_foo -> ARRAY_AGG(OBJECT_CONSTRUCT('alias_foo', foo))
1486                if first_expr.text("kind").upper() == "STRUCT":
1487                    object_construct_args = []
1488                    for expr in first_expr.expressions:
1489                        # Alias case: SELECT AS STRUCT foo AS alias_foo -> OBJECT_CONSTRUCT('alias_foo', foo)
1490                        # Column case: SELECT AS STRUCT foo -> OBJECT_CONSTRUCT('foo', foo)
1491                        name = expr.this if isinstance(expr, exp.Alias) else expr
1492
1493                        object_construct_args.extend([exp.Literal.string(expr.alias_or_name), name])
1494
1495                    array_agg = exp.ArrayAgg(
1496                        this=_build_object_construct(args=object_construct_args)
1497                    )
1498
1499                    first_expr.set("kind", None)
1500                    first_expr.set("expressions", [array_agg])
1501
1502                    return self.sql(first_expr.subquery())
1503
1504            return inline_array_sql(self, expression)