The old approach with ZFSList would keep the two-dimensional array of
lines and their fields in memory (for a short time), which could easily
consume 100s of MiB with > 10000 snapshots / bookmarks (see #34)
fixes#61
config defines a single datastructure that can act both as a Map and as a Filter
(DatasetMapFilter)
Cleanup wildcard syntax along the way (also changes semantics).