go语言之正则表达式

go语言之正则表达式

//regexp
package main

import (
    "fmt"
    "regexp"
)

func main()  {
    //需要匹配的模式
    //reg := regexp.MustCompile("\\w+") 正则表达式中的\需要转义
    reg := regexp.MustCompile(`^z.*1$`)
    //返回匹配到的结果
    result := reg.FindAllString("zhangsan1", -1)
    fmt.Printf("%v\n", result)

    //演示一个匹配不到的例子
    reg1 := regexp.MustCompile(`^z(.*)1$`)
    result1 := reg1.FindAllString("zhangsand", -1)
    fmt.Printf("%v\n", result1)

    reg2 := regexp.MustCompile(`^z(.{1})(.{1})(.*)1$`)
    result2 := reg2.FindAllString("zhangsan1", -1)
    fmt.Printf("%v\n", result2)
}

// simple_crawler
package main

import (
    "fmt"
    "io/ioutil"
    "net/http"
    "regexp"
)

func main() {
    url := "https://movie.douban.com/subject/24751763/"
    resp, err := http.Get(url)
    if err != nil {
        panic(err)
    }

    defer resp.Body.Close()

    sHtml, _ := ioutil.ReadAll(resp.Body)

    reg := regexp.MustCompile(`<span\s*property="v:itemreviewed">(.*)</span>`)
    result := reg.FindAllStringSubmatch(string(sHtml), -1)

    fmt.Println(result[0][1])

    reg1 := regexp.MustCompile(`<strong\s*class="ll\s*rating_num"\s*property="v:average">(.*)</strong>`)
    result1 := reg1.FindAllStringSubmatch(string(sHtml), -1)

    fmt.Println(result1[0][1])
}