基于golang的文章信息抓取

学习golang爬虫，实现广度爬取，抓取特定的网页地址：测试站点新笔趣阁（https://www.xsbiquge.com/）

主要学习golang的goroutine和channel之间的协作，无限爬取站点小说的地址仅限书目录地址，不进行文章内容爬取

在学习中遇到了一些问题：例如共享变量map写入竞争、连接提前关闭问题等等

福利彩蛋：没有好玩的 API 接口？上百款免费接口等你来，免费 API，免费 API 大全

为了解决map类型共享变量竞争问题，封装map类型添加读写锁限制，防止不同的goroutine之间出现写入竞争（虽然只有goroutine比较多的时候才会出现）

package utilimport "sync"//封装一个map结构体，主要用来过滤相同的url地址
type VisitMap struct {sync.RWMutexvisited map[string]bool
}func (vis *VisitMap) ReadMap(url string) bool {vis.RLock()value := vis.visited[url]vis.RUnlock()return value
}func (vis *VisitMap) WriteMap(url string) {vis.Lock()vis.visited[url] = truevis.Unlock()
}//channel封装
type Pool struct {VisitMap *VisitMapqueue    chan string
}func New(size int) *Pool {if size < 1 {size = 1}visitMap := new(VisitMap)visitMap.visited = make(map[string]bool)return &Pool{VisitMap: visitMap,queue:    make(chan string, size),wg:       &sync.WaitGroup{},}
}

接下来是main函数处理

package mainimport ("fmt""github.com/thinkeridea/go-extend/exstrings""golang.org/x/net/html""net/http""net/url""regexp""reptile/demo/queue/dao""reptile/demo/queue/model""reptile/tools""strings""time"
)func main() {//1,初始化Orm_, err := tools.OrmEngine()if err != nil {fmt.Println(err)panic(err)}bookUrl := "https://www.xsbiquge.com/"//bookUrl = "https://www.xsbiquge.com/68_68470/"pool := New(100)pool.queue <- bookUrl//根据主页面爬取子页面u, err := url.Parse(bookUrl)if err != nil {fmt.Println(err)return}hostName := u.Hostname()for uri := range pool.queue {go DownLoad(hostName, uri, pool)fmt.Println("range pool.queue  ： " + uri)}fmt.Println("结束啦：" + bookUrl)}

html抓取和信息处理，匹配小说列表链接

/**
下载
*/
func DownLoad(host, bookUrl string, pool *Pool) {showTime("download start")pool.VisitMap.WriteMap(bookUrl)//http 客户端client := http.Client{}//创建请求req, err := http.NewRequest("GET", bookUrl, nil)if err != nil {fmt.Println(err)}//设置请求headerreq.Header.Set("User-Agent", "Mozilla/5.0 (Macintosh; Intel Mac OS X 11_0_0) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/86.0.4240.198 Safari/537.36")//执行请求showTime("client.Do start")resp, err := client.Do(req)showTime("client.Do end")if err != nil {fmt.Println(err)}//这里判断下，如果响应关闭。则直接返回，实测是存在这种情况的if resp == nil || resp.Close {return}defer resp.Body.Close()showTime("links start")body, err := html.Parse(resp.Body)links := visit(nil, body)showTime("links end")for _, link := range links {absolute := urlJoin(link, bookUrl)//fmt.Println(runtime.NumGoroutine())//匹配是否是文章页面rh, _ := regexp.Compile(host)host := rh.MatchString(absolute)//匹配是否是文章页面r, _ := regexp.Compile(`\.html`)html := r.MatchString(absolute)if bookUrl != " " && host && !html {len := strings.Index(absolute, "#")if len != -1 {absolute = exstrings.SubString(absolute, 0, len)}fmt.Println("current url: " + absolute)if !pool.VisitMap.ReadMap(absolute) {fmt.Println("add url: " + absolute)go urlQueue(absolute, pool)}}}
}func urlQueue(url string, pool *Pool) {pool.queue <- url
}func urlJoin(href, base string) string {uri, err := url.Parse(href)if err != nil {return " "}baseUrl, err := url.Parse(base)if err != nil {return " "}return baseUrl.ResolveReference(uri).String()
}func showTime(action string) {//fmt.Println(fmt.Sprintf("%s :%s", action, time.Now().String()))
}

在页面链接处理过程中遇到了空指针问题：所以在处理之前先判断下 *html.Node指针是否是空指针

//文档链接处理
func visit(links []string, n *html.Node) []string {if n != nil {if n.Type == html.ElementNode && n.Data == "a" {for _, a := range n.Attr {if a.Key == "href" {links = append(links, a.Val)}}}for c := n.FirstChild; c != nil; c = c.NextSibling {links = visit(links, c)}}return links
}