Merging Arrays in Clojure

Question

I need to merge a collection of arrays based on id.

Example data:

EDIT: (changed to match Clojure data structures)

 [{:id 1, :region :NA, :name :Test1, :OS :W}
  {:id 1, :region :EU, :name :Test2, :OS :W}
  {:id 2, :region :AS, :name :test3, :OS :L}
  {:id 2, :region :AS, :name :test4, :OS :M}]

Becomes:

EDIT: (changed to match Clojure data structures)

[{:id 1, :region [:NA :EU], :name [:Test1 :Test2] ,:OS [:W]}
 {:id 2, :region [:AS] :name [:test3 :Test4], :OS [:L :M]}]

| is the delimiter (changeable) If possible, also would like alphabetical order as well.

Answer 1

You can use some combination of functions from clojure.set (if you change the outermost vector to set). Specifically clojure.set/index looks promising.

Answer 2

(def data
 [{:id 1, :region :NA, :name :Test1, :OS :W}
  {:id 1, :region :EU, :name :Test2, :OS :W}
  {:id 2, :region :AS, :name :test3, :OS :L}
  {:id 2, :region :AS, :name :test4, :OS :M}])

(defn key-join
  "join of map by key , value is distinct."
  [map-list]
  (let [keys (keys (first map-list))]
       (into {} (for [k keys] [k (vec (set (map #(% k) map-list)))]))))

(defn group-reduce [key map-list]
  (let [gdata (group-by key map-list)]
    (into [] (for [[k m] gdata] (let [m (key-join m)](assoc m key ((key m) 0)))))))



user=> (group-reduce :id data)
[{:name [:Test2 :Test1], :OS [:W], :region [:EU :NA], :id 1} {:name [:test3 :test4], :OS [:L :M], :region [:AS], :id 2}]

Answer 3

You can use the merge-with function as shown below in the example.

Firstly, we define some helper functions

(defn collect [& xs]
  (apply vector (-> xs distinct sort)))

The collect function makes sure that the items in xs are unique and sorted and finally returns them in a vector.

(defn merge-keys [k xs]
  (map #(apply merge-with collect %) (vals (group-by k xs))))

merge-keys first groups the hash-maps in xs by a primary key (in your case :id ), takes each list of grouped items and merges the values of the keys using the collect function from above.

(def xs [{:id 1, :region :NA, :name :Test1, :OS :W}
         {:id 1, :region :EU, :name :Test2, :OS :W}
         {:id 2, :region :AS, :name :test3, :OS :L}
         {:id 2, :region :AS, :name :test4, :OS :M}])

(merge-keys :id xs)
=> ({:id [1],
     :region [:EU :NA],
     :name [:Test1 :Test2],
     :OS [:W]}
    {:id [2],
     :region [:AS],
     :name [:test3 :test4],
     :OS [:L :M]})

Note however that even the :id key now has vector associated with it. You can easily un-vector it by either introducing an if statement in collect which associates a single value with the key instead of a vector...

(defn collect [& xs]
  (let [cs (apply vector (-> xs distinct sort))]
    (if (= 1 (count cs)) (first cs) cs)))

...or take the result from merge-keys and do

(map #(update-in % [:id] first) result)

which will only un-vector the :id map entry