Christian Ebert <blacktrash <at> gmx.net> writes: > See also: fe38b0a3a928, apparently put aside during refactor of > highlight. Since we've broken this before, maybe a test would be a good idea? Cheers, Dirkjan