joinery

A library to enable traversal of in-memory graph-like data structures using Clojure(Script) map protocols.

Quickstart

(require '[cjsauer.joinery :refer [joined-map]])

;; Assume you have a local normalized data source.
;; joinery includes support for "table-like" sources out of the box:
(def db
  {:person/id {1 {:person/name "Calvin"
                  :person/friends [[:person/id 2]]
                  :person/pet [:pet/id 1]}
               2 {:person/name "Derek"
                  :person/friends [[:person/id 1]]}}
   :pet/id    {1 {:pet/name "Malcolm"
                  :pet/species :dog
                  :pet/owner [:person/id 1]}}})

;; Create a joined-map interface to the db
(def jm (joined-map db))

;; Use it like a normal map, and observe "joins" resolved on-demand
(get-in jm [:person/id 1 :person/pet])
;;=> #:pet{:name "Malcolm", :owner [:person/id 1], :species :dog}

;; Cardinality-many joins are resolved recursively
(get-in jm [:person/id 1 :person/friends])
;; => [#:person{:name "Derek", :friends [[:person/id 1]]}]

;; Cycles are possible
(get-in jm [:person/id 1 :person/pet :pet/owner])
;; => #:person{:name "Calvin", :friends [[:person/id 2]], :pet [:pet/id 1]}

;; Most of the expected map functions are implemented
;;   assoc, dissoc, find, seq, reduce, etc...
(def jm' (assoc-in jm [:person/id 1 :person/best-friend] [:pet/id 1]))
(get-in jm' [:person/id 1])
;; => #:person{:name "Calvin", :best-friend [:pet/id 1], ...}

;; Let's follow the newly added edge
(get-in jm' [:person/id 1 :person/best-friend])
;; => #:pet{:name "Malcolm", :owner [:person/id 1], :species :dog}

Advanced usage

The joined-map constructor accepts a couple more helpful options that provide additional means of customization. Firstly, by default, joined-map will use the provided db as the starting point of traversal. One can optionally provide a "starting entity" that will become the "current" value of the joined map:

(def db
  {:person/id {1 {:person/name "Calvin"
                  :person/friends [[:person/id 2]]
                  :person/pet [:pet/id 1]}
               2 {:person/name "Derek"
                  :person/friends [[:person/id 1]]}}
   :pet/id    {1 {:pet/name "Malcolm"
                  :pet/species :dog
                  :pet/owner [:person/id 1]}}})

;; Use the db as the backing source, but start at a different entity
(def jm (joined-map db {:ui.selected/user [:person/id 1]}))
jm
;; => #:ui.selected{:user [:person/id 1]}

;; Joins are resolved just as before
(:ui.selected/user jm)
;; => #:person{:friends [[:person/id 2]], :name "Calvin", :pet [:pet/id 1]}

In addition, one can customize the functionality of joins by implementing the Joinery protocol. Here is the default table-join implementation that ships with joinery:

(deftype TableIdentJoinery []
  Joinery
  (is-join? [_ v] (and (vector? v)
                    (= 2 (count v))
                    (keyword? (first v))))
  (join [_ table v] (get-in table v)))

;; Create a joined-map using a specific Joinery implementation:
(joined-map db entity (TableIdentJoinery.))

We can see above that the protocol is quite simple. We need to provide two things: how to identify what values should be treated as "links", and, given we've reached one of these links, how do we "resolve" it. With these two functions, we can obtain a map-like interface to a myriad of different normalized sources.

Prior Art

joinery came about while experimenting with the latest Clojure trend of in-memory databases, namely:

• *Pyramid's normalized structure
• *Fulcro's app state
• Asami's entity function
• DataScript's entity function
• Datomic's entity function

While not a database per se, Pathom3's Smart Map is another great source of inspiration. Pathom's scope is more broad in that smart map access can trigger arbitrary code to run (even network access), while joinery is mainly focused on local data structures only.

* joinery will work out of the box with these libraries

Development

Run the project's tests:

$ clojure -T:build test

Run the project's CI pipeline and build a JAR:

$ clojure -T:build ci

This will produce an updated pom.xml file with synchronized dependencies inside the META-INF directory inside target/classes and the JAR in target. You can update the version (and SCM tag) information in generated pom.xml by updating build.clj.

Install it locally (requires the ci task be run first):

$ clojure -T:build install

Deploy it to Clojars -- needs CLOJARS_USERNAME and CLOJARS_PASSWORD environment variables (requires the ci task be run first):

$ clojure -T:build deploy

License

Copyright © 2021 Calvin Sauer

EPLv1.0 is just the default for projects generated by clj-new: you are not required to open source this project, nor are you required to use EPLv1.0! Feel free to remove or change the LICENSE file and remove or update this section of the README.md file!

Distributed under the Eclipse Public License version 1.0.