golang 使用去并发下载图片资源来源：HTTP：//my.oschina.net/qbit/blog/189928

Posted 2021-05-24

package img

import (
	"database/sql"
	"fmt"
	_ "github.com/go-sql-driver/mysql"
	"io"
	"log"
	"net/http"
	"os"
	"strconv"
	"sync"
	"time"
)

const (
	DataRoot     = "./tmp/" // 存放封面图的根目录
	TimeoutLimit = 5        // 设置超时时间
)

// 表示章节ID和封面URL的对应关系
type VolumeCover struct {
	VolumeID int
	Url      string
	Lock     sync.Mutex
	Msg      chan string
}

// 将图片下载并保存到本地
func SaveImage(vc *VolumeCover) {
	res, err := http.Get(vc.Url)
	defer res.Body.Close()
	if err != nil {
		vc.Msg <- (strconv.Itoa(vc.VolumeID) + " HTTP_ERROR")
	}
	// 创建文件
	dst, err := os.Create(DataRoot + strconv.Itoa(vc.VolumeID) + ".jpg")
	if err != nil {
		vc.Msg <- (strconv.Itoa(vc.VolumeID) + " OS_ERROR")
	}
	// 生成文件
	_, err := io.Copy(dst, res.Body)
	if err != nil {
		vc.Msg <- (strconv.Itoa(vc.VolumeID) + " COPY_ERROR")
	}
	// goroutine通信
	vc.Lock.Lock()
	vc.Msg <- "in"
	vc.Lock.Unlock()
}

func Start(name string, password string, limit int) error {
	runtime.GOMAXPROCS(4)
	sl, err := sql.Open("mysql", name+":"+password+"@/xxx?charset=utf8")
	defer sl.Close()
	if err != nil {
		return err
	}
	// 构造SELECT语句并检索
	queryStr := "SELECT VolumeID, ImageUrl FROM volume "
	if limit > 0 {
		queryStr += "limit " + strconv.Itoa(limit)
	}
	rows, err := sl.Query(queryStr)
	defer rows.Close()
	if err != nil {
		return err
	}

	// 构建列表
	i := 0
	vclist := make([]*VolumeCover, limit)
	for rows.Next() {
		vc := &VolumeCover{}
		if err := rows.Scan(&vc.VolumeID, &vc.Url); err != nil {
			return err
		}
		vc.Msg = make(chan string, 1)
		vc.Lock = *new(sync.Mutex)
		vclist[i] = vc
		i++
	}

	// start goroutines
	for i := 0; i < len(vclist); i++ {
		go SaveImage(vclist[i])
		// set timeout
		go func(i int) {
			time.Sleep(TimeoutLimit * time.Second)
			vclist[i].Lock.Lock()
			vclist[i].Msg <- "out"
			vclist[i].Lock.Unlock()
		}(i)
	}

	// 阻塞地获取结果
	for i := 0; i < len(vclist); i++ {
		func(c *VolumeCover) {
			select {
			case <-c.Msg:
				if m, ok := <-c.Msg; ok {
					fmt.Println(m)
				}
				close(c.Msg)
			}
		}(vclist[i])
	}
	return nil
}

HTTP的并发下载 Accept-Ranges

        大家都用过迅雷等下载工具，特点就是支持并发下载，断点续传。

        主要讲三个方面，如何HTTP的并发下载、通过Golang进行多协程开发、如何断点续传。

        想要并发下载，就是把下载内容分块，然后并行下载这些块。这就要求服务器能够支持分块获取数据。大迅雷、电驴这种都有自己的协议，thunder://这种，我们只研究原理，就说说HTTP协议对于并发的支持。

	HTTP头		对应值		含义
	Content-Length		14247		HTTP响应的Body大小，下载的时候，Body就是文件，也可以认为是文件大小，单位是比特
	Content-Disposition		inline; filename=”bryce.jpg”		是MIME协议的扩展，MIME协议指示MIME用户代理如何显示附加的文件。当浏览器接收到头时，它会激活文件下载。这里还包含了文件名
	Accept-Ranges		bytes		允许客户端以bytes的形式获取文件
	Range		bytes=0-511		分块获取数据，这里表示获取第0到第511的数据，共512字节

        Nginx官网ngx_http_slice_module: Module ngx_http_slice_module

         Nginx的ngx_http_slice_module模块是用来支持Range回源的。 ngx_http_slice_module从Nginx的1.9.8版本开始有的。启用ngx_http_slice_module模块需要在编译Nginx时，加参数--with-http_slice_module。

location / 
    slice             1m;
    proxy_cache       cache;
    proxy_cache_key   $uri$is_args$args$slice_range;
    proxy_set_header  Range $slice_range;
    proxy_cache_valid 200 206 1h;
    proxy_pass        http://localhost:8000;

  Module ngx_http_slice_module 缓存配置文件

#user  nobody;
worker_processes  1;

events 
    worker_connections  1024;



http 
    log_format  main  '$remote_addr - $remote_user [$time_local] "$request" '
                      '$status $body_bytes_sent "$http_referer" '
                      '"$http_user_agent" "$http_x_forwarded_for"';

    access_log  logs/access.log  main;

    sendfile        on;

    keepalive_timeout  65;

    #cache
    proxy_cache_path /data/cache
             keys_zone=cache_my:100m
                levels=1:1
                inactive=12d
                max_size=200m;

    server 
        listen       80;
        server_name  localhost;

        location / 
                #slice
                slice 1k;
                proxy_cache cache_my;
                proxy_cache_key $uri$is_args$args$slice_range;
                add_header X-Cache-Status $upstream_cache_status;
                proxy_set_header Range $slice_range;
                proxy_cache_valid 200 206 3h;
                proxy_pass http://192.168.1.10:80;

                proxy_cache_purge PURGE from 127.0.0.1;
        
    

三、运行结果

查看的是源站的日志

index.html文件大小为5196

curl  www.*****.com/index.html -r 0-1024

四  GO实例代码

package main

import (
	"flag"
	"fmt"
	"io"
	"log"
	"math"
	"mime"
	"net/http"
	"net/http/httputil"
	"os"
	"strings"
	"sync"
	"time"
)

const (
	DEFAULT_DOWNLOAD_BLOCK int64 = 4096
)

type GoGet struct 
	Url           string
	Cnt           int
	DownloadBlock int64
	CostomCnt     int
	Latch         int
	Header        http.Header
	MediaType     string
	MediaParams   map[string]string
	FilePath      string // 包括路径和文件名
	GetClient     *http.Client
	ContentLength int64
	DownloadRange [][]int64
	File          *os.File
	TempFiles     []*os.File
	WG            sync.WaitGroup


func NewGoGet() *GoGet 
	get := new(GoGet)
	get.FilePath = "./"
	get.GetClient = new(http.Client)

	flag.Parse()
	get.Url = *urlFlag
	get.DownloadBlock = DEFAULT_DOWNLOAD_BLOCK

	return get


var urlFlag = flag.String("u", "http://7b1h1l.com1.z0.glb.clouddn.com/bryce.jpg", "Fetch file url")

// var cntFlag = flag.Int("c", 1, "Fetch concurrently counts")

func main() 
	get := NewGoGet()

	download_start := time.Now()

	req, err := http.NewRequest("HEAD", get.Url, nil)
	resp, err := get.GetClient.Do(req)
	get.Header = resp.Header
	if err != nil 
		log.Panicf("Get %s error %v.\\n", get.Url, err)
	
	get.MediaType, get.MediaParams, _ = mime.ParseMediaType(get.Header.Get("Content-Disposition"))
	get.ContentLength = resp.ContentLength
	get.Cnt = int(math.Ceil(float64(get.ContentLength / get.DownloadBlock)))
	if strings.HasSuffix(get.FilePath, "/") 
		get.FilePath += get.MediaParams["filename"]
	
	get.File, err = os.Create(get.FilePath)
	if err != nil 
		log.Panicf("Create file %s error %v.\\n", get.FilePath, err)
	
	log.Printf("Get %s MediaType:%s, Filename:%s, Size %d.\\n", get.Url, get.MediaType, get.MediaParams["filename"], get.ContentLength)
	if get.Header.Get("Accept-Ranges") != "" 
		log.Printf("Server %s support Range by %s.\\n", get.Header.Get("Server"), get.Header.Get("Accept-Ranges"))
	 else 
		log.Printf("Server %s doesn't support Range.\\n", get.Header.Get("Server"))
	

	log.Printf("Start to download %s with %d thread.\\n", get.MediaParams["filename"], get.Cnt)
	var range_start int64 = 0
	for i := 0; i < get.Cnt; i++ 
		if i != get.Cnt-1 
			get.DownloadRange = append(get.DownloadRange, []int64range_start, range_start + get.DownloadBlock - 1)
		 else 
			// 最后一块
			get.DownloadRange = append(get.DownloadRange, []int64range_start, get.ContentLength - 1)
		
		range_start += get.DownloadBlock
	
	// Check if the download has paused.
	for i := 0; i < len(get.DownloadRange); i++ 
		range_i := fmt.Sprintf("%d-%d", get.DownloadRange[i][0], get.DownloadRange[i][1])
		temp_file, err := os.OpenFile(get.FilePath+"."+range_i, os.O_RDONLY|os.O_APPEND, 0)
		if err != nil 
			temp_file, _ = os.Create(get.FilePath + "." + range_i)
		 else 
			fi, err := temp_file.Stat()
			if err == nil 
				get.DownloadRange[i][0] += fi.Size()
			
		
		get.TempFiles = append(get.TempFiles, temp_file)
	

	go get.Watch()
	get.Latch = get.Cnt
	for i, _ := range get.DownloadRange 
		get.WG.Add(1)
		go get.Download(i)
	

	get.WG.Wait()

	for i := 0; i < len(get.TempFiles); i++ 
		temp_file, _ := os.Open(get.TempFiles[i].Name())
		cnt, err := io.Copy(get.File, temp_file)
		if cnt <= 0 || err != nil 
			log.Printf("Download #%d error %v.\\n", i, err)
		
		temp_file.Close()
	
	get.File.Close()
	log.Printf("Download complete and store file %s with %v.\\n", get.FilePath, time.Now().Sub(download_start))
	defer func() 
		for i := 0; i < len(get.TempFiles); i++ 
			err := os.Remove(get.TempFiles[i].Name())
			if err != nil 
				log.Printf("Remove temp file %s error %v.\\n", get.TempFiles[i].Name(), err)
			 else 
				log.Printf("Remove temp file %s.\\n", get.TempFiles[i].Name())
			
		
	()


func (get *GoGet) Download(i int) 
	defer get.WG.Done()
	if get.DownloadRange[i][0] > get.DownloadRange[i][1] 
		return
	
	range_i := fmt.Sprintf("%d-%d", get.DownloadRange[i][0], get.DownloadRange[i][1])
	log.Printf("Download #%d bytes %s.\\n", i, range_i)

	defer get.TempFiles[i].Close()

	req, err := http.NewRequest("GET", get.Url, nil)
	req.Header.Set("Range", "bytes="+range_i)
	resp, err := get.GetClient.Do(req)
	defer resp.Body.Close()
	if err != nil 
		log.Printf("Download #%d error %v.\\n", i, err)
	 else 
		cnt, err := io.Copy(get.TempFiles[i], resp.Body)
		if cnt == int64(get.DownloadRange[i][1]-get.DownloadRange[i][0]+1) 
			log.Printf("Download #%d complete.\\n", i)
		 else 
			req_dump, _ := httputil.DumpRequest(req, false)
			resp_dump, _ := httputil.DumpResponse(resp, true)
			log.Panicf("Download error %d %v, expect %d-%d, but got %d.\\nRequest: %s\\nResponse: %s\\n", resp.StatusCode, err, get.DownloadRange[i][0], get.DownloadRange[i][1], cnt, string(req_dump), string(resp_dump))
		
	


// http://stackoverflow.com/questions/15714126/how-to-update-command-line-output
func (get *GoGet) Watch() 
	fmt.Printf("[=================>]\\n")

  单元测试 

package tests

import (
	"fmt"
	"testing"

	"github.com/mnhkahn/go_code/goget"
)

func TestProcess(t *testing.T) 
	schedule := goget.NewGoGetSchedules(2)
	schedule.SetDownloadBlock(1)
	job := schedule.NextJob()
	fmt.Println(job)
	schedule.FinishJob(job)

	job = schedule.NextJob()
	fmt.Println(job)
	schedule.FinishJob(job)

判断是否支持 Accept-Ranges ： bytes

 curl -i -X HEAD https://img-blog.csdnimg.cn/30f7496053534e77a036730ac40da861.png

参考资料：

ngx_http_slice_module (ngx_http_slice_module) - Nginx 中文开发手册 - 开发者手册 - 云+社区 - 腾讯云

尝鲜：Nginx-1.9.8 推出的切片模块 - 纯真年代

CDN如何使用nginx负载均衡实现回源请求_BigChen_up的博客-CSDN博客_nginx 回源

​​​​​​Nginx进行Range缓存_zzhongcy的博客-CSDN博客_nginx range

Golang实现多线程并发下载 - Go语言中文网 - Golang中文社区