用clojure中的定界符分割一个序列?

split a sequence by delimiter in clojure?

说我有一个Clojure序列,例如

1
'(1 2 3 6 7 8)

我想将其拆分,以便每当遇到一个可被3整除的元素时拆分列表,以便结果看起来像

1
'((1 2) (3) (6 7 8))

(编辑:我真正需要的是

1
[[1 2] [3] [6 7 8]]

,但我也将使用序列版本:)

在Clojure中执行此操作的最佳方法是什么?

partition-by没有帮助:

1
2
(partition-by #(= (rem % 3) 0) '(1 2 3 6 7 8))
; => ((1 2) (3 6) (7 8))

split-with关闭:

1
2
(split-with #(not (= (rem % 3) 0)) '(1 2 3 6 7 8))
; => [(1 2) (3 6 7 8)]


像这样吗

1
2
3
4
5
6
7
8
9
(defn partition-with
  [f coll]
  (lazy-seq
    (when-let [s (seq coll)]
      (let [run (cons (first s) (take-while (complement f) (next s)))]
        (cons run (partition-with f (seq (drop (count run) s))))))))

(partition-with #(= (rem % 3) 0) [1 2 3 6 7 8 9 12 13 15 16 17 18])
=> ((1 2) (3) (6 7 8) (9) (12 13) (15 16 17) (18))


以及另一种基于还原的老派解决方案:

1
2
3
4
5
6
7
8
9
10
11
user> (defn split-all [pred items]
        (when (seq items)
          (apply conj (reduce (fn [[acc curr] x]
                                (if (pred x)
                                  [(conj acc curr) [x]]
                                  [acc (conj curr x)]))
                              [[] []] items))))
#'user/split-all

user> (split-all #(zero? (rem % 3)) '(1 2 3 6 7 8 10 11 12))
;;=> [[1 2] [3] [6 7 8 10 11] [12]]

这是一个有趣的问题。 我最近在图珀洛(Tupelo)库中添加了函数split-using,在这里看起来很合适。 我在下面的代码中保留了spyx调试语句,以便您可以看到进展情况:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
(ns tst.clj.core
  (:use clojure.test tupelo.test)
  (:require
    [tupelo.core :as t]  ))
(t/refer-tupelo)

(defn start-segment? [vals]
  (zero? (rem (first vals) 3)))

(defn partition-using [pred vals-in]
  (loop [vals   vals-in
         result []]
    (if (empty? vals)
      result
      (t/spy-let [
          out-first               (take 1 vals)
          [out-rest unprocessed]  (split-using pred (spyx (next vals)))
          out-vals                (glue out-first out-rest)
          new-result              (append result out-vals)]
        (recur unprocessed new-result)))))

给我们的输出如下:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
out-first => (1)
(next vals) => (2 3 6 7 8)
[out-rest unprocessed] => [[2] (3 6 7 8)]
out-vals => [1 2]
new-result => [[1 2]]
out-first => (3)
(next vals) => (6 7 8)
[out-rest unprocessed] => [[] [6 7 8]]
out-vals => [3]
new-result => [[1 2] [3]]
out-first => (6)
(next vals) => (7 8)
[out-rest unprocessed] => [[7 8] ()]
out-vals => [6 7 8]
new-result => [[1 2] [3] [6 7 8]]

(partition-using start-segment? [1 2 3 6 7 8]) => [[1 2] [3] [6 7 8]]

或更大的输入向量:

1
2
(partition-using start-segment? [1 2 3 6 7 8 9 12 13 15 16 17 18 18 18 3 4 5])
   => [[1 2] [3] [6 7 8] [9] [12 13] [15 16 17] [18] [18] [18] [3 4 5]]

您也可以使用嵌套的loop/recur创建解决方案,但是该解决方案已经在split-using函数中进行了编码:

1
2
3
4
5
6
7
8
9
10
11
12
(defn split-using  
 "Splits a collection based on a predicate with a collection argument.
  Finds the first index N such that (pred (drop N coll)) is true. Returns a length-2 vector
  of [ (take N coll) (drop N coll) ]. If pred is never satisified, [ coll [] ] is returned."
  [pred coll]
  (loop [left  []
         right (vec coll)]
    (if (or (empty? right) ; don't call pred if no more data
            (pred right))
      [left right]
      (recur  (append left (first right))
              (rest right)))))

实际上,上述功能似乎将来会很有用。 partition-using现在已添加到Tupelo库中。