Yes, typically you would select a range of keys with min/max_key, and use the pipe: option for any other filtering, mapping, pattern matching, etc.
Example, if you want to match keys that are maps containing a key :foo with number value:
CubDB.select(db, pipe: [
filter: fn
{%{foo: n}, _value} when is_number(n) ->
true
_ ->
false
end
])
When something can be expressed both with min_key/max_key and with filter, you should prefer the first though, as that avoids even loading from disk unnecessary entries. Of course, you can combine the two (and use more than just filter: map, take, etc. are available).
