Upgrade to Bitarray v3 and new backend algorithms

Fixes

Update the code to match the new syntaxis of bitarray package
(mostly renaming ba.itersearch(...) functions into ba.search(...) ).

Plus minor bugfixes.

New functionality (backend version)

Add support_surplus parameter into iter_keys_of_intent function.
Now one can find ∆-equivalent keys of an intent,
that are minimal descriptions that describe the same objects as the intent
plus at most support_surplus more additional objects.
The function in question is called csp.mine_equivalence_classes.iter_keys_of_intent(...).

Implement MRG-Exp algorithm (Carpathia-G-Rare) for Minimal Rare Itemset mining.
A minimal rare itemset (or a minimal rare description) is a minimal subset of attributes
that describes less than (or equal to) max_support objects.
The function is called csp.mine_equivalence_classes.iter_minimal_rare_itemsets_via_mrgexp(...).

Implement Clustering via MRG-Exp algorithm for Minimal Broad Clustering mining.
A minimal broad clustering is a minimal subset of attributes that, together,
cover more than (or equal to) min_coverage objects.
The function is called csp.mine_equivalence_classes.iter_minimal_broad_clusterings_via_mrgexp(...).

Add level-wise equivalence class iteration.
The level-wise optimisation is supposed to fasten up the iteration of the equivalence class.
However, currently, it only fastens up the iteration through the equivalence class of the maximal description
(that often describes no objects).
The function is called csp.mine_equivalence_classes.iter_equivalence_class_levelwise(...).
The older and more straightforward iteration procedure is implemented in
csp.mine_equivalence_classes.iter_equivalence_class(...).

New functionality (frontend version)

Now caspailleur also downloads the context's metadata together with the context itself.
The function name stays the same: csp.io.from_fca_repo(...).

Human API

This release introduces "Human API" for mining concepts, implications, and descriptions in a user-friendly way.
This new Human API is covered in ReadMe file.
While, one can still construct efficient mining pipelines using old function-based API.

Also, the release contains some bug fixes (e.g. in computing delta-stability).

The first release

The first release of caspailleur.
Yet, the package is stable for the last 4 months.