Now we can eliminate buckets with multi-column (i.e. composite)
keys in equalities. Previous implementation was incorrect, so
we have to make an executor refactoring as well.

Planner may split some query into range of sub queries, where subquery
execution depends on results previous subquery. Then planner analyzes
query and if it finds similar situation he appends `motion` node to plan.
If executor finds `monion` node in execution plan, he executes
motion sub-query and saves result to virtual table, which linker will be used.

Previously,  unit tests and code lived in the same '*.rs' file.
We have written too much tests and these files became bloated
(up to 1000 lines) - it is difficult to navigate them. So, lets
make a refactoring.

Now we can build it with cargo doc.

At the moment IR (logical plan) is far from final release. Current
commit fixes some point on a way to the final API and full test
coverage.

To work with Calcite and Tarantool the planner needs to store values
in the plan leaf nodes. These values should be defined with exact types.
At the moment we need only several primitive types:
- boolean
- null
- number (decimal from decnumber library)
- string

As the planner analyzes SQL, it has to work with tree-valued logic
(3VL). At the moment we'll need only equivalence operation, so it
was implemented for the values in a 3VL manner.