Cassandra's IntegerType is implemented as arbitrary precision Java BigIntegers and DecimalType is implemented as arbitrary precision Java BigDecimals, see the documentation and the marshallers from the cassandra source:
i was also thinking node-bigint is the way to go, however, that would require a compiled dependency and the target installation location have gmp installed on it. What do you think?
Or maybe https://github.com/eschnou/node-bignumber
I'm not sure about that one. Its a direct port from a browser version and it doesn't have any tests. Furthermore, the last time it had anything doen to it was over a year ago.
When I pull very large Ints (64bit), I loss precision. Because I have to work with large values, I'm already working with node-bigint. These are lookup UIDs, which causes havac for me. Being able to retrieve large Ints without lossing precision will make a lot of problems go away.
https://github.com/justmoon/node-bignum doesn't required compiled dependency
how do you figure? bignum.cc is in the root and the install script is "node-gyp configure build". More than likely I'll be working in this, or node-num which is a pure JS version.
"node-gyp configure build"