why `char` for trits?

[Rudxain]

Why is it used almost everywhere in the code? Its size is 4 times as big as u8, and it's less ergonomic to write collections of chars than binary strings and byte arrays:

['0','-','+',]; // char array: clunky
*b"0-+"; // octet array: terser and efficient

An even better approach is:

/// Balanced Ternary digit ("Trit")
#[derive(Debug, Copy, Clone, Eq, PartialEq, Ord, PartialOrd)]
#[repr(i8)]
enum BT {
    /// Negative
    N = -1,
    /// Zero
    Z = 0,
    /// Positive
    P = 1,
}

Then define a macro that takes a string of BT chars and maps them to an array or Vec of BT variants.

Then, in the future, instead of Vec<BT>, we could copy the code from bitvec (is the license compatible?) and adapt it to ternary, for better optimization