In the optparse library, the wrong string is used to remove the i, m, and x options from the encoding options when casting an argument to a Regexp. It removes all characters that are NOT i, m, or x.…
Fixed issue in optparse that was passing the incorrect options when c…
…asting an argument to a Regexp.
Better yet, just remove the caret, so that the negation of the string…
… is not used.
Showing you all comments on commits in this comparison.
Should not have used regex syntax here, delete takes a string. This was deleting all characters in the intersection of o and the characters [,^,i,m,x, and ].