Support for Float from std package math/big

[wienczny]

Currently "big" types from c* are mapped using "gopkg.in/inf.v0". Please add support for the std library math/big

[Zariel]

Any particular reason? The decimal type in Cassandra are infinite precision decimals which are encoded with a scale and an unscaled part, as far as I am aware the math/big package does not support scaled/unscaled decimals.

[wienczny]

I need to convert to and from a swagger model. The model is generated using go-swagger which converts swagger "decimal numbers" to float64. The problem is that there is no easy conversion between those two.

[Zariel]

Are those floating point numbers or arbitrary precision decimals?

[wienczny]

Those values are regular javascript floating point numbers on the wire. I'm using java.lang.BigDecimal for calculations the client side though

[Zariel]

Then your losing precision over the wire, gopkg.in/inf.v0 provides types which are compatible with java.lang.BigDecimal

[wienczny]

You are right. That's why I don't want to loose more precision converting values. Maybe new types will be added with OAI/OpenAPI-Specification#607

[Zariel]

There is a proposal to a big decimal type to the standard library see golang/go#12127 but that wont solve the issue of the incompatible types you are encountering. I would recommend converting them to strings or such and then converting back to the decimal type. Unless you have any suggestions I don't see how to implement the decimal type in Cassandra with the types in math/big.