The MF recoverer had an unfortunate off-by-one error (> instead of >=), which the first commit fixes. The second commit makes the check involved faster. The final commit makes all recoverers use the same threshold for determining if there are enough shifts to constitute success. The commit messages have a little more detail, although none is too fussy, since this is a relatively simple PR.
The MF recoverer had an unfortunate off-by-one error (
>instead of>=), which the first commit fixes. The second commit makes the check involved faster. The final commit makes all recoverers use the same threshold for determining if there are enough shifts to constitute success. The commit messages have a little more detail, although none is too fussy, since this is a relatively simple PR.