Make printf trim trailing zeroes with %[Gg] format

[MCJack123]

This PR modifies the behavior of pico_printf to automatically trim trailing zeroes when using the %G or %g format specifiers. This is the behavior that is specified by GCC:

The ‘%g’ and ‘%G’ conversions print the argument in the style of ‘%e’ or ‘%E’ (respectively) if the exponent would be less than -4 or greater than or equal to the precision; otherwise they use the ‘%f’ style. A precision of 0, is taken as 1. Trailing zeros are removed from the fractional portion of the result and a decimal-point character appears only if it is followed by a digit.

Previously, printf("%.14g", 2000) would output 2000.0000000000, but with this change it outputs just 2000. This is useful for programs that want the minimum-length conversion of a number with the most precision necessary. (My example is Lua, which uses sprintf(s, "%.14g", num) to convert numbers to strings; previously this would result in unnecessarily large conversions for small integers.)

This change does not affect the default printf - newlib exhibits the same behavior as pico_printf currently does, so that would have to be adjusted as well (and I can say after reading the code that I'm not the one to do that). Users will have to link pico_printf manually to benefit AFAICT.

Please let me know if this breaks anything - I didn't have a chance to test this fully (I'm embroiled in debugging some other stuff), but looking at the code I don't think there should be much to break.

[kilograham]

Note that people using the SDK CMake build get pico_printf by default.