A toy implementation of naive-evaluated Datalog in Elixir, based on the blog
post The Essence of
Datalog. The
Datalog program given in the blog post is included in the Datalog.Example
module.
DO NOT USE THIS IN PRODUCTION, THIS IS FOR EDUCATIONAL PURPOSES ONLY.
-
No abstract types. I use structs for
Rule(%Datalog.Rule{}) andAtom(%Datalog.Atom{}), and two-tuples for the two variants ofTerm:VarandSym. -
No monadic do-notation. In a few places, I had to unpack the implicit iteration and flat-mapping that do-notation induces and make it explicit. The code looks a little strange as a result, with some
++andEnum.concat/1to get the desired level of list-nesting. Seeeval_atom/3for an example of this. -
No inline named-functions. When the blog author uses Haskell
whereclauses to define local functions inside a function scope, I've substitutedforcomprehensions or functions of the same name but different arities. Example:unify/2takes two atoms, andunify/1(local functiongoin the blog post) works over the zipped list of terms to unify from the inputs ofunify/2. -
No free-to-use fixpoint combinator. I have to tail-recurse, comparing the inputs to the outputs for equality. See
solve/1. -
Pretty-printed data-structures. I wanted the
inspectoutput to look like a Datalog program, so I implemented theInspectprotocol for the included structs.
$ iex -S mix
Erlang/OTP 22 [erts-10.7.2.1] [source] [64-bit] [smp:12:12] [ds:12:12:10] [async-threads:1] [hipe]
Interactive Elixir (1.10.3) - press Ctrl+C to exit (type h() ENTER for help)
iex(1)> Datalog.Example.ancestor
[
adviser("Andrew Rice", "Mistral Contrastin").,
adviser("Dominic Orchard", "Mistral Contrastin").,
adviser("Andy Hopper", "Andrew Rice").,
adviser("Alan Mycroft", "Dominic Orchard").,
adviser("David Wheeler", "Andy Hopper").,
adviser("Rod Burstall", "Alan Mycroft").,
adviser("Robin Milner", "Alan Mycroft").,
academicAncestor(X, Y) :- adviser(X, Y).,
academicAncestor(X, Z) :- adviser(X, Y), academicAncestor(Y, Z).,
query1(Intermediate) :-
academicAncestor("Robin Milner", Intermediate),
academicAncestor(Intermediate, "Mistral Contrastin")
.,
query2() :- academicAncestor("Alan Turing", "Mistral Contrastin").,
query3() :- academicAncestor("David Wheeler", "Mistral Contrastin").
]
iex(2)> Datalog.solve(v(1))
[
adviser("Andrew Rice", "Mistral Contrastin"),
adviser("Dominic Orchard", "Mistral Contrastin"),
adviser("Andy Hopper", "Andrew Rice"),
adviser("Alan Mycroft", "Dominic Orchard"),
adviser("David Wheeler", "Andy Hopper"),
adviser("Rod Burstall", "Alan Mycroft"),
adviser("Robin Milner", "Alan Mycroft"),
academicAncestor("Andrew Rice", "Mistral Contrastin"),
academicAncestor("Dominic Orchard", "Mistral Contrastin"),
academicAncestor("Andy Hopper", "Andrew Rice"),
academicAncestor("Alan Mycroft", "Dominic Orchard"),
academicAncestor("David Wheeler", "Andy Hopper"),
academicAncestor("Rod Burstall", "Alan Mycroft"),
academicAncestor("Robin Milner", "Alan Mycroft"),
academicAncestor("Andy Hopper", "Mistral Contrastin"),
academicAncestor("Alan Mycroft", "Mistral Contrastin"),
academicAncestor("David Wheeler", "Andrew Rice"),
academicAncestor("Rod Burstall", "Dominic Orchard"),
academicAncestor("Robin Milner", "Dominic Orchard"),
academicAncestor("David Wheeler", "Mistral Contrastin"),
academicAncestor("Rod Burstall", "Mistral Contrastin"),
academicAncestor("Robin Milner", "Mistral Contrastin"),
query1("Dominic Orchard"),
query1("Alan Mycroft"),
query3()
]
iex(3)> Datalog.query("query1", v(1))
[
[{{:var, "Intermediate"}, {:sym, "Dominic Orchard"}}],
[{{:var, "Intermediate"}, {:sym, "Alan Mycroft"}}]
]
iex(4)> Datalog.query("query2", v(1))
[]
iex(5)> Datalog.query("query3", v(1))
[[]]
iex(6)>