Erlang 如何从任何复杂项中提取proplist{K，V}列表

有一个术语，它可能由任何东西组成，如命令、命令列表、命令列表、列表列表、命令列表或。。。例如：

Term1 = [ 
      {"AAA", [ 
               { "K1", "ka1" },
               { "K2", "ka2" }, 
               { "K3", "ka3" } 
              ]
      },
      {"BBB","one"},
      {"CCC", [ 
               { "K1", "kb1" },
               { "K2", "" }, 
               { "K3", "kb3" } 
              ]
      },
      {"DDD", [ 
                [
                 { "K1", "kc1" },
                 { "K2", "kc2" }
                ],
                [
                 { "K1", "kd1" },
                 { "K2", "kd2" }
                ],
                "CACETA",
                123
              ]
      },
      {"EEE", [ 
               { "K1", "kb1" },
               { "K2", 1122  }, 
               { "K3", "kb3" } 
              ]
      },
              { "T1", "T2", "T3" },
              123,
              { "X" },
              [ 1, 2, 3, { "K5", "V5" } ] 
],

我需要生成一个包含所有proplist[{K，V}，…]的列表

[
   { "AAA" , [ ...... ] },
   { "K1"  , "ka1" },
   { "K2" , "ka2" }, 
   ...
   { "BBB" ,"one"},
   { "CCC" , [ ... ] },
   { "K1"  , "kb1" },
   ...
   { "K5", "V5" } 
]

• 请注意，有些键必须在列表中重复，它们的值可以是字符串、列表、元组或数字，也可以是任何值

• 上面示例中数据中的最后一项，例如{“T1”、“T2”、“T3”}，不应该出现在结果中，因为它不是两个术语{K，V}的proplist，而是嵌套的{“K5”、“V5”}是并且应该是结果的一部分

我看了这个类似的例子，并试图对它进行一些调整，但对于我的新手erlang brain来说，要让它在上面的场景中工作变得越来越困难

下面是一个示例，我正试图使用它使其工作，但存在一些与模式匹配相关的错误：

extractPropList( [], ResultList ) -> ResultList;

extractPropList( [H|T], ResultList ) -> extractPropList(T, extractPropList(H, ResultList));

extractPropList( {K,V}, ResultList ) -> [ {K,V} | extractPropList(K, extractPropList(V, ResultList)) ].

---

在测试上述方法时，缺少的部分是处理没有其他匹配项（不是列表，不是{K，V}）的术语的最后一行：

给定与avove相似的数据，我得到的结果如下

[{"EEE",[{"K1","kb1"},{"K2",1122},{"K3","kb3"}]},
 {"K3","kb3"},
 {"K2",1122},
 {"K1","kb1"},
 {"DDD",
  [[{"K1","kc1"},{"K2","kc2"}],
   [{"K1","kd1"},{"K2","kd2"}],
   "CACETA",123]},
 {"K2","kd2"},
 {"K1","kd1"},
 {"K2","kc2"},
 {"K1","kc1"},
 {"CCC",[{"K1","kb1"},{"K2","kb2"},{"K3","kb3"}]},
 {"K3","kb3"},
 {"K2","kb2"},
 {"K1","kb1"},
 {"BBB","one"},
 {"AAA",[{"K1","ka1"},{"K2","ka2"},{"K3","ka3"}]},
 {"K3","ka3"},
 {"K2","ka2"},
 {"K1","ka1"}]

---

在测试上述方法时，缺少的部分是处理没有其他匹配项（不是列表，不是{K，V}）的术语的最后一行：

给定与avove相似的数据，我得到的结果如下

[{"EEE",[{"K1","kb1"},{"K2",1122},{"K3","kb3"}]},
 {"K3","kb3"},
 {"K2",1122},
 {"K1","kb1"},
 {"DDD",
  [[{"K1","kc1"},{"K2","kc2"}],
   [{"K1","kd1"},{"K2","kd2"}],
   "CACETA",123]},
 {"K2","kd2"},
 {"K1","kd1"},
 {"K2","kc2"},
 {"K1","kc1"},
 {"CCC",[{"K1","kb1"},{"K2","kb2"},{"K3","kb3"}]},
 {"K3","kb3"},
 {"K2","kb2"},
 {"K1","kb1"},
 {"BBB","one"},
 {"AAA",[{"K1","ka1"},{"K2","ka2"},{"K3","ka3"}]},
 {"K3","ka3"},
 {"K2","ka2"},
 {"K1","ka1"}]