The bigint and biginteger functions throw on double values outside of long range

This works fine:

but passing any Float or Double values outside the range of a long throw an exception:

Cause: bigint and biginteger cover a series of possible input cases but did not have an explicit case for Float or Double, so was falling back to default.

Solution: Add check for Float/Double case that coerces the input to double, then uses BigDecimal.valueOf(), then converts to a BigInteger and so on as with other types.

Patch: clj-1197-make-bigint-work-on-all-doubles-v1.txt