fix: handle Unicode characters correctly in mutator byte offset calculation

[stevennevins]

Fix source code corruption when slither-mutate processes Solidity files containing Unicode characters (e.g., arrow symbols, non-ASCII comments).

Example

// → Unicode arrow
contract Counter {
    function increment() public {
        number++;
    }
}

The → arrow is 3 bytes but 1 character. This 2-byte difference causes mutations to be applied at wrong positions.

Without fix - garbage output:

nurevert()
fufunction setNumber
nu++number

With fix - correct mutations:

revert()
function setNumber
++number