Probe Implementation

[ReubenJ]

Implementation of the method from "Just-in-time learning for bottom-up enumerative synthesis".

TODO:

1. Init Probe top-down iterator
2. init update function for the probabilities of the grammar
3. write priority function that multiplies the priority function of the program and returns the most likely program
4. update derivation_heuristic accordingly
5. write tests

• check that log probabilities align with the intended enumeration order
• re-use parent's probabilities to calculate the current probability again.

Possible future work:

• Add alternating search procedures to Probe? #8 ?
• Add BottomUp iterator for Probe

[THinnerichs]

I added a prototypical implementation of Probe to Garden. I still have to write proper tests + docs.

@ReubenJ I am overwriting some core functions of Herb, like priority_function, rulenode_log_probability,... . Also, I am not happy with overwriting all top-down iterators and having to do the same for all possible iterator-type classes.
Maybe there is a smoother way?