Parsing 要映射的Clojure惯用读取文件

我需要读取文件以映射{：v'[]：f'[]}。我分割每一行，若第一个元素是v，那个么我将剩余的部分添加到v数组中，对于f数组也是如此

例如：

v 1.234 3.234 4.2345234
v 2.234 4.235235 6.2345
f 1 1 1

预期结果：

{:v [("1.234" "3.234" "4.2345234"), ("2.234" "4.235235" "6.2345")]
 :f [("1" "1" "1")]}

我的结果是：

{:v [("2.234" "4.235235" "6.2345")]
 :f [("1" "1" "1")]}

问题:

如何修复错误？只有最后一行被添加到地图中 我能避免全局变量模型和副作用吗？ 代码：

---

您似乎正在删除模型的更改状态，而不是使用两个空向量将所有行的数据附加到原始模型。您可以在读取文件时保持模型的状态，例如使用reduce：

(def model
  {:v '[]
   :f '[]})

(defn- file-lines
  [filename]
  (line-seq (io/reader filename)))

(defn- lines-with-data
  [filename]
  (->>
    (file-lines filename)
    (filter not-empty)
    (filter #(not (str/starts-with? % "#")))))

(defn- to-item [data]
  (let [[type & remaining] data]
    (case type
      "v" [:v (conj (:v model) remaining)]
      "f" [:f (conj (:f model) remaining)])))

(defn- fill-model
  [lines]
  (into model
        (for [data lines] (to-item data))))

(defn parse
  [filename]
  (->>
    (lines-with-data filename)
    (map #(str/split % #"\s+"))
    (fill-model)))

(defn- add-item [m data]
  (let [[type & remaining] data]
    ;; Updates the model `m` by appending data under the given key.
    ;; In the line below, `(keyword type)` will be :v or :f depending on `type`.
    ;; Returns the updated model.
    (update m (keyword type) conj remaining)))

(defn fill-model [lines]
  ;; Applies `add-item` to the original model and each line of the file
  (reduce add-item model lines)))