Panic when parsing byte unit with very long fractional part

[5225225]

assert_parses! {
    "9.00000000000000000000MB" => 9.megabytes(),
}

This test should pass, but instead you get

thread 'parse::parse_tests::accept' panicked at 'attempt to multiply with overflow', /rustc/8b09ba6a5d5c644fe0f1c27c7f9c80b334241707/library/core/src/num/mod.rs:843:5
stack backtrace:
   0: rust_begin_unwind
             at /rustc/8b09ba6a5d5c644fe0f1c27c7f9c80b334241707/library/std/src/panicking.rs:498:5
   1: core::panicking::panic_fmt
             at /rustc/8b09ba6a5d5c644fe0f1c27c7f9c80b334241707/library/core/src/panicking.rs:106:14
   2: core::panicking::panic
             at /rustc/8b09ba6a5d5c644fe0f1c27c7f9c80b334241707/library/core/src/panicking.rs:47:5
   3: core::num::<impl u64>::pow
             at /rustc/8b09ba6a5d5c644fe0f1c27c7f9c80b334241707/library/core/src/num/uint_macros.rs:1955:13
   4: ubyte::parse::<impl core::str::traits::FromStr for ubyte::byte_unit::ByteUnit>::from_str
             at ./src/parse.rs:77:47
   5: ubyte::parse::parse_tests::accept
             at ./src/parse.rs:183:9
   6: ubyte::parse::parse_tests::accept::{{closure}}
             at ./src/parse.rs:134:5
   7: core::ops::function::FnOnce::call_once
             at /rustc/8b09ba6a5d5c644fe0f1c27c7f9c80b334241707/library/core/src/ops/function.rs:227:5
   8: core::ops::function::FnOnce::call_once
             at /rustc/8b09ba6a5d5c644fe0f1c27c7f9c80b334241707/library/core/src/ops/function.rs:227:5
note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.

The problem is the .pow in

ubyte/src/parse.rs

Line 77 in c3cfe51

let frac_part = frac as f64 / 10u64.pow(frac_str.len() as u32) as f64;

I think this probably should be parsed successfully (as opposed to giving an error).

I'll have a look at trying to fix the arithmetic, but this kind of code is kinda hard to reason about for me.

[SergioBenitez]

This is the pow call panicking. The rustdocs make no mention of the call panicking under any conditions: https://doc.rust-lang.org/nightly/std/primitive.u64.html#method.pow. Still, there's clearly the potential for overflow here; I just wish it had been documented.