Merge 'simulator: add counterexample minimization' from Alperen Keleş

This PR introduces counterexample minimization(shrinking) in the
simulator. It will require changes to the current structure in various
places, so I've opened it as a draft PR for now, in order to not
overwhelm the reviewers all at once.
- [x] Turn interactions plans into sequences of properties instead of
sequences of interactions, adding a semantic layer.
- [x] Add assumptions to the properties, rendering a property invalid if
its assumptions are not valid.
- [x] Record the failure point in a failing assertion, as shrinking by
definition works when the same assertion fails for a smaller input.
- [ ] Add shrinking at three levels,
  - [x] top level(removing whole properties),
  - [x] property level(removing interactions within properties),
  - [ ] interaction level(shrinking values in interactions to smaller
ones).
- [ ] Add [marauders](https://github.com/alpaylan/marauders) as a dev
dependency, inject custom mutations to a testing branch for evaluating
the simulator performance.
- [ ] Integrate the simulator evaluation with the CI.

Closes #623

simulator/generation/mod.rs

@@ -4,24 +4,41 @@ use anarchist_readable_name_generator_lib::readable_name_custom;
 use rand::{distributions::uniform::SampleUniform, Rng};
 
 pub mod plan;
+pub mod property;
 pub mod query;
 pub mod table;
 
+/// Arbitrary trait for generating random values
+/// An implementation of arbitrary is assumed to be a uniform sampling of
+/// the possible values of the type, with a bias towards smaller values for
+/// practicality.
 pub trait Arbitrary {
     fn arbitrary<R: Rng>(rng: &mut R) -> Self;
 }
 
+/// ArbitraryFrom trait for generating random values from a given value
+/// ArbitraryFrom allows for constructing relations, where the generated
+/// value is dependent on the given value. These relations could be constraints
+/// such as generating an integer within an interval, or a value that fits in a table,
+/// or a predicate satisfying a given table row.
 pub trait ArbitraryFrom<T> {
-    fn arbitrary_from<R: Rng>(rng: &mut R, t: &T) -> Self;
+    fn arbitrary_from<R: Rng>(rng: &mut R, t: T) -> Self;
 }
 
+/// Frequency is a helper function for composing different generators with different frequency
+/// of occurences.
+/// The type signature for the `N` parameter is a bit complex, but it
+/// roughly corresponds to a type that can be summed, compared, subtracted and sampled, which are
+/// the operations we require for the implementation.
+// todo: switch to a simpler type signature that can accomodate all integer and float types, which
+//       should be enough for our purposes.
 pub(crate) fn frequency<
     'a,
     T,
     R: rand::Rng,
     N: Sum + PartialOrd + Copy + Default + SampleUniform + SubAssign,
 >(
-    choices: Vec<(N, Box<dyn FnOnce(&mut R) -> T + 'a>)>,
+    choices: Vec<(N, Box<dyn Fn(&mut R) -> T + 'a>)>,
     rng: &mut R,
 ) -> T {
     let total = choices.iter().map(|(weight, _)| *weight).sum::<N>();
@@ -37,6 +54,7 @@ pub(crate) fn frequency<
     unreachable!()
 }
 
+/// one_of is a helper function for composing different generators with equal probability of occurence.
 pub(crate) fn one_of<'a, T, R: rand::Rng>(
     choices: Vec<Box<dyn Fn(&mut R) -> T + 'a>>,
     rng: &mut R,
@@ -45,15 +63,20 @@ pub(crate) fn one_of<'a, T, R: rand::Rng>(
     choices[index](rng)
 }
 
+/// pick is a helper function for uniformly picking a random element from a slice
 pub(crate) fn pick<'a, T, R: rand::Rng>(choices: &'a [T], rng: &mut R) -> &'a T {
     let index = rng.gen_range(0..choices.len());
     &choices[index]
 }
 
+/// pick_index is typically used for picking an index from a slice to later refer to the element
+/// at that index.
 pub(crate) fn pick_index<R: rand::Rng>(choices: usize, rng: &mut R) -> usize {
     rng.gen_range(0..choices)
 }
 
+/// gen_random_text uses `anarchist_readable_name_generator_lib` to generate random
+/// readable names for tables, columns, text values etc.
 fn gen_random_text<T: Rng>(rng: &mut T) -> String {
     let big_text = rng.gen_ratio(1, 1000);
     if big_text {