Impedance

Transform Clojure maps at speed.

Rationale

In electronics, impedance mismatch is a case where an electrical load's or source's impedance doesn't match that of the driver/sink system (respectively). Impedance matching allows us to maximize the power transfer across that boundary.

A similar phenomenon happens in software, where in the boundary between processes there is a mismatch between the representation model in one and the other. This library, like a transformer in electronics, allows us to easily transform an input map to look how our system would expect and maximize the information flow.

Just maps

Maps are pervasive in Clojure for information modeling. While you can cheat and access vectors associatively, the library does not yet support mapping over sequences.

Why not Meander?

Meander is great. It's feature rich and beautiful. It is a whole arsenal. On the other hand, Impedance is a simple hatchet. Good for one thing, sharp enough, and simple. It's also faster.

Differences

• No scanning or Cartesian product. Only leaves can be matched.

Performance differences

As tested with criterium:

{:a {:b ?x :c ?y} "x" ?z} ;; from
{:x ?x :y ?y :z ?z} ;; to
;;; Bench transform default
;;;              Execution time mean : 107.636019 ns
;;; Bench transform checked
;;;              Execution time mean : 76.051257 ns
;;; Bench transform unchecked
;;;              Execution time mean : 73.565921 ns
;;; Bench meander
;;;              Execution time mean : 430.781144 ns

Usage

Dependency

[bsless/impedance "0.0.0-alpha1"]

A basic transformer

(require '[impedance.transform :as t])
(def f (t/transform {:a {:b ?x :c ?y} "x" ?z} {:x (inc ?x) :y ?y :z ?z}))
(f {:a {:b 1 :c 2} "x" 3}) ;; => {:x 2, :y 2, :z 3}

Creating a transformer programmatically

Warning: uses eval. Use at your own risk.

(def from {:a {:b ?x :c ?y} "x" ?z})
(def to {:x (inc ?x) :y ?y :z ?z})
(def f (t/eval-transform from to))
(f {:a {:b 1 :c 2} "x" 3}) ;; => {:x 2, :y 2, :z 3}

Inferred Context - Poor Man's jq

An alternate syntax for transformations is the inferred context one:

(require '[impedance.context :as c])
(c/with-inferred-context ctx
  {:a {:b (or ^:? [:foo :bar] ^:? [:fizz :buzz])
       :c ^:? [:fizz :bazz]}
   :x ^:? [:foo :quux]})

Here, specify only the shape of the desired output, and use the :? metadata on vectors to mark them as paths referring to the map ctx.

This syntax can be used to create and define functions:

(fntx :checked
      {:a {:b (or ^:? [:foo :bar] ^:? [:fizz :buzz])
           :c ^:? [:fizz :bazz]}
       :x ^:? [:foo :quux]})
(defntx f :unchecked
  {:a {:b (or ^:? [:foo :bar] ^:? [:fizz :buzz])
       :c ^:? [:fizz :bazz]}
   :x ^:? [:foo :quux]})

Context compiler modes

All context compilers and transformers receive an optional argument for mode, which can be one of the following:

• :poly - will use clojure.core/get to get values out of a collection. Highly polymorphic at the slight cost of performance.
• :unchecked - will use direct method invoke of .valAt on an object. Will throw if an object doesn't implement ILookup and NPE for nil.
• :checked - same as :unchecked but nil checks before every call to .valAt.

License

Copyright © 2020 Ben Sless

This program and the accompanying materials are made available under the terms of the Eclipse Public License 2.0 which is available at http://www.eclipse.org/legal/epl-2.0.

This Source Code may also be made available under the following Secondary Licenses when the conditions for such availability set forth in the Eclipse Public License, v. 2.0 are satisfied: GNU General Public License as published by the Free Software Foundation, either version 2 of the License, or (at your option) any later version, with the GNU Classpath Exception which is available at https://www.gnu.org/software/classpath/license.html.