Document the relation between AST/Frontend/FrontendMixin/Backend

[ltfish]

The following is quoted from @zardus in the Slack channel. We should put them in the documentation.

generally speaking:

• claripy abstracts expressions using claripy.AST objects
• Frontends provide different paradigms for handling those expressions. The FullFrontend (should really be named SolverFrontend or something) handles them using something like an SMT solver, LightFrontend (should really be named ApproximationFrontend or something) handles them by using an abstract (and approximating) data domain.
• each Frontend needs to, at some point, do actual operation and evaluations on the AST. ASTs don't support this on their own. Instead, Backends translate ASTs into BackendObjects (i.e., python primitives, Z3 expressions, strided intervals for VSA, etc) and also handle the appropriate state-tracking objects (such as Z3 solvers)
• Frontends take ASTs as inputs and use Backends to backend.convert() those ASTs into BackendObjects that can be evaluated and otherwise reasoned about
• FrontendMixins customize the operation of Frontends. For example, ModelCacheMixin caches solutions from an SMT solver

[zardus]

Done yesterday :-)

angr/angr-doc@043e50a

This might be a good use-case for some sort of diagram.

[ltfish]

Awesome!