This alters how strings inside ""s are tokenized: instead of ignoring the "s, it considers all the content inside them as a single token.
It also adds ; and = to the character ignore lists.
The consequences for already existing parsers should be null.
