Member Junction
    Preparing search index...

    Variable ALLOWED_SQL_FUNCTIONSConst

    ALLOWED_SQL_FUNCTIONS: {
        aggregates: readonly [
            "COUNT",
            "COUNT_BIG",
            "SUM",
            "AVG",
            "MIN",
            "MAX",
            "STDEV",
            "STDEVP",
            "VAR",
            "VARP",
            "STRING_AGG",
            "CHECKSUM_AGG",
            "ARRAY_AGG",
            "JSONB_AGG",
            "JSON_AGG",
            "BOOL_AND",
            "BOOL_OR",
            "EVERY",
            "PERCENTILE_CONT",
            "PERCENTILE_DISC",
            "MODE",
        ];
        conditional: readonly ["CASE", "WHEN", "THEN", "ELSE", "END"];
        conversion: readonly [
            "CAST",
            "CONVERT",
            "TRY_CAST",
            "TRY_CONVERT",
            "FORMAT",
            "TO_CHAR",
            "TO_DATE",
            "TO_NUMBER",
            "TO_TIMESTAMP",
        ];
        date: readonly [
            "DATEPART",
            "DATEDIFF",
            "DATEADD",
            "YEAR",
            "MONTH",
            "DAY",
            "HOUR",
            "MINUTE",
            "SECOND",
            "GETDATE",
            "GETUTCDATE",
            "SYSDATETIME",
            "EOMONTH",
            "NOW",
            "CURRENT_DATE",
            "CURRENT_TIME",
            "CURRENT_TIMESTAMP",
            "LOCALTIMESTAMP",
            "DATE_TRUNC",
            "DATE_PART",
            "EXTRACT",
            "AGE",
            "TO_DATE",
            "TO_TIMESTAMP",
            "MAKE_DATE",
            "MAKE_TIMESTAMP",
            "JUSTIFY_DAYS",
        ];
        logical: readonly [
            "AND",
            "OR",
            "NOT",
            "IS",
            "NULL",
            "LIKE",
            "BETWEEN",
            "IN",
            "ILIKE",
            "SIMILAR",
            "ANY",
            "ALL",
            "SOME",
            "EXISTS",
        ];
        math: readonly [
            "ABS",
            "CEILING",
            "FLOOR",
            "ROUND",
            "POWER",
            "SQRT",
            "LOG",
            "LOG10",
            "EXP",
            "SIGN",
            "RAND",
            "CEIL",
            "RANDOM",
            "TRUNC",
            "MOD",
            "DIV",
            "GREATEST",
            "LEAST",
            "WIDTH_BUCKET",
        ];
        nullHandling: readonly [
            "ISNULL",
            "COALESCE",
            "NULLIF",
            "IIF",
            "GREATEST",
            "LEAST",
        ];
        ordering: readonly [
            "ASC",
            "ASCENDING",
            "DESC",
            "DESCENDING",
            "OVER",
            "PARTITION",
            "BY",
            "ORDER",
            "ROWS",
            "RANGE",
            "UNBOUNDED",
            "PRECEDING",
            "FOLLOWING",
            "CURRENT",
            "ROW",
            "NULLS",
            "FIRST",
            "LAST",
            "FILTER",
            "WITHIN",
            "GROUPING",
        ];
        string: readonly [
            "LEN",
            "LENGTH",
            "UPPER",
            "LOWER",
            "LTRIM",
            "RTRIM",
            "TRIM",
            "LEFT",
            "RIGHT",
            "SUBSTRING",
            "CHARINDEX",
            "REPLACE",
            "CONCAT",
            "STUFF",
            "POSITION",
            "STRPOS",
            "SPLIT_PART",
            "INITCAP",
            "LPAD",
            "RPAD",
            "REPEAT",
            "REVERSE",
            "TRANSLATE",
            "REGEXP_REPLACE",
            "REGEXP_MATCHES",
            "REGEXP_SUBSTR",
            "TO_CHAR",
            "MD5",
            "ASCII",
            "CHR",
        ];
    } = ...

    Safe SQL functions allowed in expressions, organized by category.

    This is a deny-by-default SAFETY list, not a dialect grammar. Its job is to stop a user-supplied expression from smuggling in something that is not a read-only scalar/aggregate call — it is not, and must not become, a per-platform parser. So it carries the union of every supported dialect's read-only vocabulary, and checkFunctionNames consults the flattened union.

    Keying the buckets off ResolvePlatformKey is the alternative, and it costs more than it buys: a PostgreSQL name that reaches a SQL Server tenant is rejected by the SQL Server parser on its own — DATE_TRUNC(...) simply never parses there. The union therefore cannot widen the blast radius past what the target server will already accept, whereas a dialect switch here would add a second independent place for the dialect to be resolved, and so a second place for it to be wrong.

    Before this list held the PostgreSQL names, a perfectly ordinary PG filter such as DATE_TRUNC('month', CreatedAt) was rejected outright with "Function 'DATE_TRUNC' is not allowed" on every PostgreSQL tenant.

    Every name here is read-only by construction. Anything that writes, escalates, or reaches outside the statement stays on DANGEROUS_SQL_KEYWORDS, which is checked FIRST (step 2 of validate) — a name added here can never unblock one that is denied there.

    Type Declaration

    • Readonlyaggregates: readonly [
          "COUNT",
          "COUNT_BIG",
          "SUM",
          "AVG",
          "MIN",
          "MAX",
          "STDEV",
          "STDEVP",
          "VAR",
          "VARP",
          "STRING_AGG",
          "CHECKSUM_AGG",
          "ARRAY_AGG",
          "JSONB_AGG",
          "JSON_AGG",
          "BOOL_AND",
          "BOOL_OR",
          "EVERY",
          "PERCENTILE_CONT",
          "PERCENTILE_DISC",
          "MODE",
      ]
    • Readonlyconditional: readonly ["CASE", "WHEN", "THEN", "ELSE", "END"]
    • Readonlyconversion: readonly [
          "CAST",
          "CONVERT",
          "TRY_CAST",
          "TRY_CONVERT",
          "FORMAT",
          "TO_CHAR",
          "TO_DATE",
          "TO_NUMBER",
          "TO_TIMESTAMP",
      ]
    • Readonlydate: readonly [
          "DATEPART",
          "DATEDIFF",
          "DATEADD",
          "YEAR",
          "MONTH",
          "DAY",
          "HOUR",
          "MINUTE",
          "SECOND",
          "GETDATE",
          "GETUTCDATE",
          "SYSDATETIME",
          "EOMONTH",
          "NOW",
          "CURRENT_DATE",
          "CURRENT_TIME",
          "CURRENT_TIMESTAMP",
          "LOCALTIMESTAMP",
          "DATE_TRUNC",
          "DATE_PART",
          "EXTRACT",
          "AGE",
          "TO_DATE",
          "TO_TIMESTAMP",
          "MAKE_DATE",
          "MAKE_TIMESTAMP",
          "JUSTIFY_DAYS",
      ]
    • Readonlylogical: readonly [
          "AND",
          "OR",
          "NOT",
          "IS",
          "NULL",
          "LIKE",
          "BETWEEN",
          "IN",
          "ILIKE",
          "SIMILAR",
          "ANY",
          "ALL",
          "SOME",
          "EXISTS",
      ]
    • Readonlymath: readonly [
          "ABS",
          "CEILING",
          "FLOOR",
          "ROUND",
          "POWER",
          "SQRT",
          "LOG",
          "LOG10",
          "EXP",
          "SIGN",
          "RAND",
          "CEIL",
          "RANDOM",
          "TRUNC",
          "MOD",
          "DIV",
          "GREATEST",
          "LEAST",
          "WIDTH_BUCKET",
      ]
    • ReadonlynullHandling: readonly ["ISNULL", "COALESCE", "NULLIF", "IIF", "GREATEST", "LEAST"]
    • Readonlyordering: readonly [
          "ASC",
          "ASCENDING",
          "DESC",
          "DESCENDING",
          "OVER",
          "PARTITION",
          "BY",
          "ORDER",
          "ROWS",
          "RANGE",
          "UNBOUNDED",
          "PRECEDING",
          "FOLLOWING",
          "CURRENT",
          "ROW",
          "NULLS",
          "FIRST",
          "LAST",
          "FILTER",
          "WITHIN",
          "GROUPING",
      ]
    • Readonlystring: readonly [
          "LEN",
          "LENGTH",
          "UPPER",
          "LOWER",
          "LTRIM",
          "RTRIM",
          "TRIM",
          "LEFT",
          "RIGHT",
          "SUBSTRING",
          "CHARINDEX",
          "REPLACE",
          "CONCAT",
          "STUFF",
          "POSITION",
          "STRPOS",
          "SPLIT_PART",
          "INITCAP",
          "LPAD",
          "RPAD",
          "REPEAT",
          "REVERSE",
          "TRANSLATE",
          "REGEXP_REPLACE",
          "REGEXP_MATCHES",
          "REGEXP_SUBSTR",
          "TO_CHAR",
          "MD5",
          "ASCII",
          "CHR",
      ]

    Settings

    Member Visibility

    On This Page

    WHY THIS IS A UNION RATHER THAN A PER-DIALECT TABLE