[Erlang]AC自动机过滤屏蔽词

Erlang实现的AC自动机，通过预处理屏蔽词能够实现接近O(n)的时间复杂度
实现了屏蔽词的检查、匹配、替换功能
先挖个坑，贴个代码，后续有时间再讲讲aho_corasick算法的原理
在文件根目录运行命令行程序，由于模块统一使用了utf8编码，windows命令行默认使用的是gbk，需要使用chcp 65001切换到unicode编码，然后使用命令erl打开erlang程序
c(aho_corasick)函数进行编译

aho_corasick:test()运行测试用例，可以看到成功的屏蔽了题目和作者名

%%%-------------------------------------------------------------------
%%% @author Huangcanxin cx2298545090@outlook.com
%%% @copyright (C) 2022
%%% @doc
%%% AC自动机
%%% @end
%%% Created : 23. 6月 2022 10:22
%%%-------------------------------------------------------------------
-module(aho_corasick).

-define(AC_TRIE_ROOT, 0).
-record(ac_trie, {
   
   
    mod,
    success,      % 成功跳转表
    failure,      % 回溯表
    output,       % 输出表
    child,
    next_node = 1
}).


%% API
-export([build/2, rebuild/2, rebuild/3]). % 生成树
-export([add_word/2, check/2, match/2, replace/3, replace/4, make_child/1, append_word/2]).
-export([init_trie/0, get/3, set/3, clean/1, foldl/3, erase/2]). % maps 结构的ac_trie
-export([merge/1, sub/2]).
-export([test/0]).


%%%% -----------------------------------
%%%% 添加关键词创建前缀树
%%%% -----------------------------------
% output 为了节省空间和加速替换存的是关键词长度
add_word(Word, AcTrie) ->
    add_word(Word, bit_size(Word), ?AC_TRIE_ROOT, AcTrie).
add_word(<<>>, OutputWord, Node, AcTrie) ->
    if
        OutputWord == 0 orelse OutputWord == <<>> ->
            AcTrie;
        true ->
            #ac_trie{
   
   output = Output, mod = Mod} = AcTrie,
            NOutput = Mod:set(Node, OutputWord, Output),
            AcTrie#ac_trie{
   
   output = NOutput}
    end;
add_word(<<Char/utf8, Left/binary>>, OutputWord, Node, AcTrie = #ac_trie{
   
   next_node = NextNode, child = Child, success = Success, mod = Mod}) ->
    case Mod:get({
   
   Node, Char}, Success, ?AC_TRIE_ROOT) of
        ?AC_TRIE_ROOT ->
            NSuccess = Mod:set({
   
   Node, Char}, NextNode, Success),
            ChildChar = Mod:get(Node, Child, []),
            NChild = Mod:set(Node, [Char | ChildChar], Child),
            NewAcTrie = AcTrie#ac_trie{
   
   next_node = NextNode + 1, child = NChild, success = NSuccess},
            add_word(Left, OutputWord, NextNode, NewAcTrie);
        GoNode ->
            add_word(Left, OutputWord, GoNode, AcTrie)
    end.

%%%% -----------------------------------
%%%% 通过success生成child
%%%% -----------------------------------
make_child(AcTrie) ->
    #ac_trie{
   
   success = Success, mod = Mod, child = Child} = AcTrie,
    NChild = Mod:foldl(
        fun({
   
   {
   
   Node, Char}, _V}, C) ->
            ChildChar = Mod:get(Node, C, []),
            Mod:set(Node, [Char | ChildChar], C)
        end,
        Child,
        Success
    ),
    AcTrie#ac_trie{
   
   child = NChild}.


%%%% -----------------------------------
%%%% 需要确保child存在
%%%% 追加关键词 会重新建立failure 需要添加多个词的话最好一次性插入
%%%% -----------------------------------
append_word([], AcTrie) ->
    #ac_trie{
   
   failure = Failure, mod = Mod} = AcTrie,
    NFailure