module Decimal:sig..end
decimal type alias provides more readable serializations to s-expressions, at
    the cost of lower precision.  For example:
    # sexp_of_decimal 3.000000000001;;
    - : Sexp.t = 3
    # sexp_of_float 3.000000000001;;
    - : Sexp.t = 3.0000000000010000889
    
    Also, the decimal sexp-converter will fail when provided with nan or infinity.
    # float_of_sexp (Sexp.Atom "nan");;
    - : float = nan
    # decimal_of_sexp (Sexp.Atom "nan");;
    Exception:
    (Sexplib.Conv.Of_sexp_error (Failure common.ml.Decimal_nan_or_inf) nan).
    typet =float
val compare : t -> t -> intval t_of_sexp : Sexplib.Sexp.t -> tval sexp_of_t : t -> Sexplib.Sexp.tval bin_t : t Bin_prot.Type_class.tval bin_read_t : t Bin_prot.Read.readerval __bin_read_t__ : (int -> t) Bin_prot.Read.readerval bin_reader_t : t Bin_prot.Type_class.readerval bin_size_t : t Bin_prot.Size.sizerval bin_write_t : t Bin_prot.Write.writerval bin_writer_t : t Bin_prot.Type_class.writer