Doc Comments / Phobos /
Add your comments here...
I was looking into the std.conv module the other evening, and I had thought it had conversion functions for all of the D numerical types, but sadly I discovered that it only did the whole-number types. So the next evening I looked around and found std.string.atof() that does only float and double, and std.math2's "real atof( in char )" function, which does a pretty good job with float, double, and real right out of the box. With a bit of playing around with std.math2 yet the next evening I managed to make some useful wrapper functions that worked well with D's remaining floating-point types: ifloat, idouble, ireal, cfloat, cdouble, and creal.
I really think it would be great if Walter would add these functions, or functions like these to the std.conv before D v1.0. It would make std.conv more complete, and on top of that, it would save a person's time looking all around for functions to convert number-types that were converted with std.string.toString() functions back to their original D's numerical value(s).
Some templates (e.g. to, roundTo, parse) have been added to the D 2.x version. The std2 project contains a port for D 1.x.