arohner.spectrum

https://github.com/arohner/spectrum.git

git clone 'https://github.com/arohner/spectrum.git'

(ql:quickload :arohner.spectrum)

---

spectrum

A library for doing static analysis of Clojure code, catching clojure.spec conform errors at compile time.

Wait what?

It's like core.typed, but it relies on clojure.spec annotations.

So it's an optional static type system?

Kind-of. It finds errors at compile time, and predicates kind of look like types. So sure.

Current Status

Developer Preview, not yet ready for any kind of use.

Current development is working towards making spectrum self-check.

Goals

• usable
• pragmatic
• readable implementation
• fast
• configurable strictness levels
• incremental checking

Non-goals

• perfection
• correctness
• requiring 100% spec coverage.

In particular, spectrum aims to be fast and usable, and catching bugs. It aims to have low false positives, at the expense of potentially not catching 100% of type errors.

A tool that catches 80% of bugs that you use every day is better than a 100% tool that you don't use. Spectrum will converge on 100% correct, but won't guarantee correctness for a while.

Usage

Still under active development, so this is incomplete.

Use clojure.spec as normal. Then, at the repl:

(require '[spectrum.flow :as f])
(require '[spectrum.types :as t])

(f/infer-var #'foo)

There is also

(f/infer-form '(foo 3))

which is useful when you want to debug the signature of a form.

infer-form can optionally take a map of keywordized variables to types

(f/infer-form '(string? x) {:x #'string?})
#Value[true]

(f/infer-form '(string? x) {:x (t/value-t 3)})
#Value[false]

Limitations

• It's still very early.

Justification

• because I could
• working example of types = proofs = predicates
• Some properties of a program are easier to prove at runtime. Some at compile-time
• this is extra checking on top of spec, not a replacement for it
• to get HN to shut up about how adding a static type system to a dynamic language is impossible
• why not just use spec by itself?
• instrument doesn't check return values
• check works best on pure functions with good generators.
• Not all clojure code is pure
• Not always easy to write a good generator for all functions (hello, any fn that takes a DB connection)
• Not always easy to write a generator w/ 100% coverage
• you can't spec non-defn functions
• you can't spec higher order functions
• generative testing with side-effects doesn't work well
• generative testing gives false negatives on rare codepaths
• static typing guarantees all codepaths are checked
• spec doesn't deal with non-fn vars at all (binding, alter-var-root, set!, etc)
• generative testing can be slow

Plan

Spectrum is still very early, and not ready for production use. Current development is focused on making spectrum self-check.

License

Copyright © 2019 Allen Rohner

Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported