在Common Lisp中用独立包充当哈希表是否可行？有何优劣？

是否可以在独立于项目包的专用包中存储大量符号，将其当作哈希表（用符号作为键来表示集合成员关系）供项目访问？

代码示例 #

* (defpackage :project (:use :cl))
#<PACKAGE "PROJECT">

* (defpackage :temp (:use :cl))
#<PACKAGE "TEMP">

* (in-package :project)
#<PACKAGE "PROJECT">

* (intern "ABC" :temp)
TEMP::ABC
NIL

* (find-symbol "ABC" :temp)
TEMP::ABC
:INTERNAL

;The following is not needed for this project, but is available

* (setf (symbol-value (find-symbol "ABC" :temp)) 123)
123

* (symbol-value (find-symbol "ABC" :temp))
123

这种方式看似可行，但有没有充分理由应该避免这么做，转而使用标准哈希表？我最初的想法是避免项目包被大量杂项键符号充斥——这些符号会在运行时由字符串生成（用于检查集合成员），涉及在运行时intern大量可能存在或不存在于"表"中的符号。我还想知道find-symbol是否比gethash更高效。另外，我也可以使用基于equal的字符串哈希表，但大量查找场景下似乎eq比对equal更合适。

补充测试：equal哈希表 vs 包符号查找的性能对比 #

哈希表测试代码 #

* (defun random-string (n)
  "Generate a random string of length n."
  (let ((charset "ABCDEFGHIJKLMNOPQRSTUVWXYZ"))
    (iter (repeat n)
          (collect (char charset (random (length charset)))
                   result-type string))))
RANDOM-STRING

* (defparameter *ht* (make-hash-table :test #'equal :size 10000))
*HT*

* (iter (for i from 0 to 5000)
      (setf (gethash (random-string 5) *ht*)
            t))
NIL

* (time (dotimes (i 1000000)
        (gethash (random-string 5) *ht*)))

Evaluation took:
  0.150 seconds of real time
  0.109375 seconds of total run time (0.109375 user, 0.000000 system)
  72.67% CPU
  541,959,942 processor cycles
  127,937,760 bytes consed

包查找测试代码 #

* (defpackage :temp1)
#<PACKAGE "TEMP1">

* (iter (for i from 0 to 5000)
      (intern (random-string 5) :temp1))
NIL

* (time (dotimes (i 1000000)
        (find-symbol (random-string 5) :temp1)))

Evaluation took:
  0.224 seconds of real time
  0.171875 seconds of total run time (0.156250 user, 0.015625 system)
  [ Run times consist of 0.015 seconds GC time, and 0.157 seconds non-GC time. ]
  76.79% CPU
  807,944,162 processor cycles
  127,954,624 bytes consed

测试结果 #

测试显示，使用equal哈希表的查找速度比包符号查找快约1.5倍。

内容的提问来源于stack exchange，提问作者davypough