A fast Haskell library for converting between double precision floating point numbers and text strings. It is implemented as a binding to the V8-derived C++ double-conversion library.
The default read is pretty slow on doubles; there is bytestring-lexing which is faster than default, but I'm guessing double-conversion could be even faster – or perhaps fast_float would be a better option?
IIUC, this only exposes functions from doubles to strings, not the other way around, but https://github.com/google/double-conversion/blob/master/double-conversion/string-to-double.h has functions from string-to-double – would it make sense to include both directions?
The default
read
is pretty slow on doubles; there is bytestring-lexing which is faster than default, but I'm guessing double-conversion could be even faster – or perhaps fast_float would be a better option?