daft.functions.hash#

def hash(
    *exprs: Expression,
    seed: Any | None = None,
    hash_function: Literal["xxhash", "xxhash32", "xxhash64", "xxhash3_64", "murmurhash3", "sha1"] | None = "xxhash",
) -> Expression:
    """Hashes the values in the Expression.

    Uses the specified hash function to hash the values in the expression. Default to [XXH3_64bits](https://xxhash.com/) non-cryptographic hash function.

    Args:
        exprs: One or more expressions (or column names/wildcards) to hash.
        seed (optional): Seed used for generating the hash. Defaults to 0.
        hash_function (optional): Hash function to use. One of "xxhash" (alias for "xxhash3_64"), "xxhash32", "xxhash64", "xxhash3_64", "murmurhash3", or "sha1". Defaults to "xxhash" (alias for "xxhash3_64").

    Returns:
        Expression (UInt64 Expression): The hashed expression.

    Note:
        Null values will produce a hash value instead of being propagated as null.

    """
    # Only pass hash_function if explicitly provided to maintain backward compatibility in string representation
    kwargs = {}
    if not exprs:
        raise ValueError("hash() requires at least one expression")
    normalized = []
    for expr in exprs:
        if isinstance(expr, str):
            resolved = col(expr)
            normalized.append(resolved)
        else:
            resolved = Expression._to_expression(expr)
            normalized.append(resolved)
    if seed is not None:
        kwargs["seed"] = seed
    if hash_function is not None:
        kwargs["hash_function"] = hash_function
    return Expression._call_builtin_scalar_fn("hash", *normalized, **kwargs)