Clojure(Script)'s missing piece
Find a file
2015-06-30 20:00:14 -04:00
src/com/rpl fixed cljs obj-extends? to have special case for nil because it cannot be reflected on like normal objects 2015-06-30 19:46:55 -04:00
test/com/rpl/specter add test runner file for cljs 2015-06-30 19:59:20 -04:00
.gitignore remove unwanted files 2015-06-24 16:21:24 -04:00
CHANGES.md 0.5.7 2015-06-29 23:44:18 -04:00
DEVELOPER.md fix formatting 2015-06-30 20:00:14 -04:00
LICENSE initial commit 2015-02-26 10:55:20 -05:00
project.clj add clojurescript dep and update clojure dep to 1.7.0, fix cljsbuild definition 2015-06-30 19:53:17 -04:00
README.md fix readme with update->transform change 2015-06-24 18:02:05 -04:00
repl.clj Testing 2015-06-24 14:28:33 -04:00
VERSION 0.5.7 2015-06-29 23:44:18 -04:00

Specter

Most of Clojure programming involves creating, manipulating, and transforming immutable values. However, as soon as your values become more complicated than a simple map or list like a list of maps of maps transforming these data structures becomes extremely cumbersome.

Specter is a library for querying and updating nested data structures. One way to think of it is "get-in" and "assoc-in" on steroids, though Specter works on any data structure, not just maps. It is similar to the concept of a "lens" in functional programming, though it has some important extensions.

Specter is fully extensible. At its core, its just a protocol for how to navigate within a data structure. By extending this protocol, you can use Specter to navigate any data structure or object you have.

Specter is a very high performance library. For example: the Specter equivalent to get-in runs 30% faster than get-in, and the Specter equivalent to update-in runs 5x faster than update-in. In each case the Specter code is equally as convenient.

Latest Version

The latest release version of Specter is hosted on Clojars:

Current Version

How to use

The usage of Specter will be explained via example. Suppose you have a sequence of maps, and you want to extract all the even values for :a keys. Here's how you do it:

user> (use 'com.rpl.specter)
nil
user> (select [ALL :a even?]
              [{:a 1} {:a 2} {:a 4} {:a 3}])
[2 4]

select extracts a sequence of results from a data structure. It takes in a "selector", which is a sequence of steps on how to navigate into that data structure. In this case, ALL looks at every element in the sequence, :a looks at the :a key for each element currently navigated to, and even? filters out any elements that aren't an even value.

If you had a map with a sequence as the value for the :a key, here's how to get all odd numbers in that sequence:

user> (use 'com.rpl.specter)
nil
user> (select [:a ALL odd?]
              {:a [1 2 3 5] :b :c})
[1 3 5]

Another function called transform is used to perform a transformation on a data structure. In addition to a selector, it takes in an "transform function" which specifies what to do with each element navigated to. For example, here's how to increment all the even values for :a keys in a sequence of maps:

user> (transform [ALL :a even?]
              inc
              [{:a 1} {:a 2} {:a 4} {:a 3}])
[{:a 1} {:a 3} {:a 5} {:a 3}]

Here's another example of transform:

user> (use 'com.rpl.specter)
nil
user> (transform [:a ALL odd?]
              dec
              {:a [1 2 3 5] :b :c})
{:b :c, :a [0 2 2 4]}

Specter comes with all sorts of built-in ways of navigating data structures. For example, here's how to increment the last odd number in a sequence:

user> (transform [(filterer odd?) LAST]
              inc
              [2 1 3 6 9 4 8])
[2 1 3 6 10 4 8]

filterer navigates you to a view of the sequence currently being looked at. LAST navigates you to the last element of whatever sequence you're looking at. But of course during transforms, the transforms are performed on the original data structure.

srange is a selector for looking at or replacing a subsequence of a sequence. For example, here's how to increment all the odd numbers between indexes 1 (inclusive) and 4 (exclusive):

user> (transform [(srange 1 4) ALL odd?] inc [0 1 2 3 4 5 6 7])
[0 2 2 4 4 5 6 7]

srange can also be used to replace that subsequence entirely with a new sequence. For example, here's how to replace the subsequence from index 2 to 4 with [-1 -1 -1]:

user> (transform (srange 2 4) (fn [_] [-1 -1 -1]) [0 1 2 3 4 5 6 7 8 9])
[0 1 -1 -1 -1 4 5 6 7 8 9]

The above can be written more concisely using the setval function, which is a wrapper around transform:

user> (setval (srange 2 4) [-1 -1 -1] [0 1 2 3 4 5 6 7 8 9])
[0 1 -1 -1 -1 4 5 6 7 8 9]

Here's how to concatenate the sequence [:a :b] to every nested sequence of a sequence:

user> (setval [ALL END] [:a :b] [[1] '(1 2) [:c]])
[[1 :a :b] (1 2 :a :b) [:c :a :b]]

END is a wrapper around srange-dynamic, which takes in functions that return the start index and end index given the structure.

walker is another useful selector that walks the data structure until a predicate is matched. Here's how to get all the numbers out of a map:

user> (select (walker number?)
              {2 [1 2 [6 7]] :a 4 :c {:a 1 :d [2 nil]}})
[2 1 2 1 2 6 7 4]

When doing more involved transformations, you often find you lose context when navigating deep within a data structure and need information "up" the data structure to perform the transformation. Specter solves this problem by allowing you to collect values during navigation to use in the transform function. Here's an example which transforms a sequence of maps by adding the value of the :b key to the value of the :a key, but only if the :a key is even:

user> (transform [ALL (collect-one :b) :a even?]
              +
              [{:a 1 :b 3} {:a 2 :b -10} {:a 4 :b 10} {:a 3}])
[{:b 3, :a 1} {:b -10, :a -8} {:b 10, :a 14} {:a 3}]

The transform function receives as arguments all the collected values followed by the navigated to value. So in this case + receives the value of the :b key followed by the value of the :a key, and the transform is performed to :a's value.

The four built-in ways for collecting values are VAL, collect, collect-one, and putval. VAL just adds whatever element it's currently on to the value list, while collect and collect-one take in a selector to navigate to the desired value. collect works just like select by finding a sequence of values, while collect-one expects to only navigate to a single value. Finally, putval adds an external value into the collected values list.

To make your own selector, implement the StructurePath protocol which looks like:

(defprotocol StructurePath
  (select* [this structure next-fn])
  (transform* [this structure next-fn])
  )

As an example, here is how Clojure keywords implement this protocol:

(extend-type clojure.lang.Keyword
  StructurePath
  (select* [kw structure next-fn]
    (next-fn (get structure kw)))
  (transform* [kw structure next-fn]
    (assoc structure kw (next-fn (get structure kw)))
    ))

next-fn represents the rest of the select or transform, respectively. As you can see, this implementation perfectly captures what it means to navigate via a keyword within a data structure. In the select case, it completes the select by calling next-fn on the value of the keyword. In the transform case, it transforms the nested data structure using next-fn and then replaces the current value of the keyword with that transformed data structure.

Finally, you can make select and transform work much faster by precompiling your selectors using the comp-paths function. There's about a 3x speed difference between the following two invocations of transform:

(def precompiled (comp-paths ALL :a even?))

(transform [ALL :a even?] inc structure)
(compiled-transform precompiled inc structure)

Depending on the details of the selector and the data being transformed, precompiling can sometimes provide more than a 10x speedup.

Some more examples:

Decrement every value in a map:

user> (transform [ALL LAST]
              dec
              {:a 1 :b 3})
{:b 2 :a 0}

Increment the value for :a key by 10:

user> (transform [:a (putval 10)]
              +
              {:a 1 :b 3})
{:b 3 :a 11}

Get every number divisible by 3 out of a sequence of sequences:

user> (select [ALL ALL #(= 0 (mod % 3))]
              [[1 2 3 4] [] [5 3 2 18] [2 4 6] [12]])
[3 3 18 6 12]

Append [:c :d] to every subsequence that has at least two even numbers:

user> (setval [ALL
               (selected? (filterer even?) (view count) #(>= % 2))
               END]
              [:c :d]
              [[1 2 3 4 5 6] [7 0 -1] [8 8] []])
[[1 2 3 4 5 6 :c :d] [7 0 -1] [8 8 :c :d] []]

For every map in a sequence, increment every number in :c's value if :a is even or increment :d if :a is odd:

user> (transform [ALL (if-path [:a even?] [:c ALL] :d)]
              inc
              [{:a 2 :c [1 2] :d 4} {:a 4 :c [0 10 -1]} {:a -1 :c [1 1 1] :d 1}])
[{:c [2 3], :d 4, :a 2} {:c [1 11 0], :a 4} {:c [1 1 1], :d 2, :a -1}]

Future work

  • Make it possible to parallelize selects/transforms
  • Any connection to transducers?
  • Add Clojurescript compatibility

License

Copyright 2015 Red Planet Labs, Inc. Specter is licensed under Apache License v2.0.