Add the following features to the concordancer search interface, as proposed here:

1. Apply an unlimited number of sets of filters simultaneously.
2. To appear in the output, a line of the corpus must have tokens which fulfill all the sets of filters specified.
3. Filters can be added one by one and deleted