Add checked, wrapping and saturating arithmetic for rem/mod

[kimikage]

The default arithmetic for rem is still checked arithmetic.

saturating_{rem|mod} behaves somewhat strangely in the case of the overflow. (cf. #221 (comment))

julia> x = typemin(Q0f7); y = -eps(Q0f7);

julia> saturating_div(x, y) # 128 --> 127
127

julia> saturating_rem(x, y) # != 0
-0.008Q0f7

julia> x == saturating_div(x, y) * y + saturating_rem(x, y)
true

This also adds the support for 3-arg rem. Only RoundToZero(default), RoundDown(mod), and RoundUp are supported. In RoundUp mode, rem causes OverflowError with unsigned (i.e. Normed) values. Otherwise, rem/mod does not cause an overflow.

This also adds dummy checked, wrapping and saturating rem for type modulus to prevent MethodError due to lexical replacement of x % X.

This completes the implementation of the checked arithmetic. Performance tweaks without changing behavior and the rest of #185 can be carried out after the release of v0.9.0.

Closes #221

[codecov]

Codecov Report

Merging #230 into master will increase coverage by 0.13%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #230      +/-   ##
==========================================
+ Coverage   95.42%   95.55%   +0.13%     
==========================================
  Files           6        6              
  Lines         677      698      +21     
==========================================
+ Hits          646      667      +21     
  Misses         31       31              

Impacted Files	Coverage Δ
src/FixedPointNumbers.jl	96.49% <100.00%> (+0.33%)	⬆️
src/fixed.jl	97.91% <100.00%> (-0.03%)	⬇️
src/normed.jl	93.48% <100.00%> (-0.07%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update c6b8a9c...e4ebc3b. Read the comment docs.

[timholy]

I really like how much you've been able centralize here, nice work as always!

[kimikage]

There is a bit more discussion about what to do with these arithmetic methods in v0.9.0, but I'll merge this first.
Although my plan is JuliaMath/CheckedArithmetic.jl#9, @timholy may have a slightly different idea (cf. https://discourse.julialang.org/t/rfc-what-should-the-arithmetic-within-the-fixedpointnumbers-be/46697/36).