• Trie性能分析之敏感词过滤golang


     

    package util
    
    import (
        "strings"
    )
    
    type Node struct {
        //rune表示一个utf8字符
        char   rune
        Data   interface{}
        parent *Node
        Depth  int
        //childs 用来当前节点的所有孩子节点
        childs map[rune]*Node
        term   bool
    }
    
    type Trie struct {
        root *Node
        size int
    }
    
    func NewNode() *Node {
        return &Node{
            childs: make(map[rune]*Node, 32),
        }
    }
    
    func NewTrie() *Trie {
        return &Trie{
            root: NewNode(),
        }
    }
    
    //假如我要把 敏感词: “我操”
    // Add("我操", nil)
    // Add("色情片", nil)
    func (p *Trie) Add(key string, data interface{}) (err error) {
    
        key = strings.TrimSpace(key)
        node := p.root
        runes := []rune(key)
        for _, r := range runes {
            ret, ok := node.childs[r]
            if !ok {
                ret = NewNode()
                ret.Depth = node.Depth + 1
                ret.char = r
                node.childs[r] = ret
            }
    
            node = ret
        }
    
        node.term = true
        node.Data = data
        return
    }
    
    // findNode("色情片")
    func (p *Trie) findNode(key string) (result *Node) {
    
        node := p.root
        chars := []rune(key)
        for _, v := range chars {
            ret, ok := node.childs[v]
            if !ok {
                return
            }
    
            node = ret
        }
    
        result = node
        return
    }
    
    func (p *Trie) collectNode(node *Node) (result []*Node) {
    
        if node == nil {
            return
        }
    
        if node.term {
            result = append(result, node)
            return
        }
    
        var queue []*Node
        queue = append(queue, node)
    
        for i := 0; i < len(queue); i++ {
            if queue[i].term {
                result = append(result, queue[i])
                continue
            }
    
            for _, v1 := range queue[i].childs {
                queue = append(queue, v1)
            }
        }
    
        return
    }
    
    func (p *Trie) PrefixSearch(key string) (result []*Node) {
    
        node := p.findNode(key)
        if node == nil {
            return
        }
    
        result = p.collectNode(node)
        return
    }
    
    // text = "我们都喜欢王八蛋"
    // replace = "***"
    func (p *Trie) Check(text, replace string) (result string, hit bool) {
    
        chars := []rune(text)
        if p.root == nil {
            return
        }
    
        var left []rune
        node := p.root
        start := 0
        for index, v := range chars {
            ret, ok := node.childs[v]
            if !ok {
                left = append(left, chars[start:index+1]...)
                start = index + 1
                node = p.root
                continue
            }
    
            node = ret
            if ret.term {
                hit = true
                node = p.root
                left = append(left, ([]rune(replace))...)
                start = index + 1
                continue
            }
        }
    
        result = string(left)
        return
    }
    package util
    
    import (
        "fmt"
        "testing"
    )
    
    func TestTrie(t *testing.T) {
    
        trie := NewTrie()
        trie.Add("黄色", nil)
        trie.Add("绿色", nil)
        trie.Add("蓝色", nil)
    
        result, str := trie.Check("我们这里有一个黄色的灯泡,他存在了很久。他是蓝色的。", "***")
    
        fmt.Printf("result:%#v, str:%v
    ", result, str)
    
    }
  • 相关阅读:
    核心动画03-图标抖动.mp4
    核心动画02-CAKeyframeAnimation帧动画
    核心动画01-CABasicAnimation
    CALayer参考文献
    CALayer04-自定义图层
    CALayer03-隐式动画
    CALayer02-创建新图层
    iOS开发中的网络请求
    IT教育课程考评系统开发-06
    IT教育课程考评系统开发-05
  • 原文地址:https://www.cnblogs.com/sunlong88/p/11980046.html
Copyright © 2020-2023  润新知