p5-Math-BigInt: update to 1.999818.

* Fix CPAN RT #130736 regarding numify() on a Math::BigFloat with a large
  number of digits and an equally large, but negative, exponent.

* Fix a problem in to_ieee754() due to rounding in bpow().

1.999817 2019-10-11

* Avoid non-integers in intermediate computations in Math::BigInt::Calc. This
  causes errors due to rounding issues, e.g., 47265625999999999 / 1000000000
  is 47265626 with double precision.

* Remove api_version() and the corresponding test file. There is no need to
  check which methods are supported by a backend library now that each backend
  library is a subclass of Math::BigInt::Lib. Methods not provided in the
  backend library are provided by the parent class.

* Add to_ieee745() and from_ieee754().

* Add backermann() and ackermann() for the Ackermann function.

* Add buparrow() and uparrow() for Knuth's up-arrow notation.

* Add information about the github repository.

* Update links in SUPPORT section in POD.

* Fix bpow(). It returned NaN when it should truncate to zero.

* Make blsft() in Math::BigInt allow bases up until the largest unsigned
  integer.

* Make _lsft() in Math::BigInt::Calc handle an arbitrarily large base.

* Add new methods bdigitsum() and digitsum() to Math::BigInt. Add new method
  _digitsum() to Math::BigInt::Lib.

* Add new methods is_non_negative() and is_non_positive().

* Extend the default collation sequence used by to_base() and from_base() to
  all the 94 printable ASCII characters except blank/space.

* Make new() in Math::BigFloat handle octal floating point numbers.

* Slightly more robust t/01load.t.

* Remove unused variables.

* Miscellaneous code cleanup.

(wiz)