Library POCS.Lab1.StatDbImpl

Require Import POCS.
Require Import VariablesAPI.
Require Import StatDbAPI.

An implementation of StatDB

StatDB is built on top of the Variables API, which maintains two variables VarCount and VarSum. StatDB reads and writes these two variables using the API provided in the VarsAPI module. VarCount stores the number of integers that VarSum sums up. The mean is just VarSum/VarCount.

Module StatDB (v : VarsAPI) <: StatDbAPI.

  Import ListNotations.

  Definition add (v : nat) : proc unit :=
    sum <- VarSum;
    count <- VarCount;
    _ <- v.write VarSum (sum + v);
    _ <- v.write VarCount (count + 1);
    Ret tt.

Exercise : complete the implementation of mean

  Definition mean : proc (option nat) :=
    Ret None.

  Definition init' : proc InitResult :=
    _ <- v.write VarCount 0;
    _ <- v.write VarSum 0;
    Ret Initialized.

  Definition init := then_init v.init init'.

  Definition recover : proc unit :=

Exercise : complete the implementation of the abstraction function:

  Definition statdb_abstraction (vars_state : VariablesAPI.State) (statdb_state : StatDbAPI.State) : Prop :=

  Definition abstr : Abstraction StatDbAPI.State :=
      {| abstraction := statdb_abstraction |}.

  Example abstr_1_ok : statdb_abstraction (VariablesAPI.mkState 0 0) nil.

Exercise : complete the proof for the following admitted examples

Exercise : complete the proof of add

  Theorem add_ok : forall v, proc_spec (add_spec v) (add v) recover abstr.

Exercise : complete the proof of mean

  Theorem mean_ok : proc_spec mean_spec mean recover abstr.

  Theorem recover_noop : rec_noop recover abstr no_crash.

End StatDB.

Require Import VariablesImpl.
Module StatDBImpl := StatDB Vars.
Print Assumptions StatDBImpl.abstr_2_nok.
Print Assumptions StatDBImpl.abstr_3_ok.
Print Assumptions StatDBImpl.add_ok.
Print Assumptions StatDBImpl.mean_ok.