• node爬虫之图片下载


    背景:针对一些想换头像的玩家,而又不知道用什么头像的,作为一名代码爱好者,能用程序解决的,就不用程序来换头像,说干就干,然后就整理了一下。

    效果图

    环境配置

    • 安装node环境
    • node -v
    • node版本最好在8.11.1以上

    项目结构

    assets是存放所下载的图片

    static是静态资源页面

    eg.js是下载图片示例(node eg.js)

    img.json是网页所获取的json数据

    index.js属于服务端

    安装依赖

    npm init ( 会生成一个package.json)
    
    npm i express --save-dev
    
    npm i cheerio--save-dev
    
    npm i superagent--save-dev
    
    npm i superagent-charset--save-dev
    
    npm i request--save-dev

    代码区

    1. eg.js

    var fs = require('fs');
    var request = require("request");
    var path = require('path');
    var src = "https://pic.qqtn.com/up/2019-6/2019061811092772406.jpg";
    var writeStream = fs.createWriteStream('./assets/aa.png');
    var readStream = request(src)
    readStream.pipe(writeStream);
    
    readStream.on('end', function() {
        console.log('文件下载成功');
    });
    readStream.on('error', function() {
        console.log("错误信息:" + err)
    })
    writeStream.on("finish", function() {
        console.log("文件写入成功");
        writeStream.end();
    });

    2.index.js

    var superagent = require('superagent');
    var charset = require('superagent-charset');
    charset(superagent);
    var express = require('express');
    var baseUrl = 'https://www.qqtn.com/';
    const cheerio = require('cheerio');
    var request = require("request");
    var fs = require('fs')
    var path = require('path')
    
    var checkDir = fs.existsSync("assets");
    var app = express();
    app.use(express.static('static'))
    app.get('/index', function (req, res) {
        //设置请求头
        res.header("Access-Control-Allow-Origin", "*");
        res.header('Access-Control-Allow-Methods', 'PUT, GET, POST, DELETE, OPTIONS');
        res.header("Access-Control-Allow-Headers", "X-Requested-With");
        res.header('Access-Control-Allow-Headers', 'Content-Type');
        //类型
        console.log(req.query, '类型')
        var type = req.query.type;
        //页码
        var page = req.query.page;
        type = type || 'weixin';
        page = page || '1';
        var route = `tx/${type}tx_${page}.html`
        //网页页面信息是gb2312,所以chaeset应该为.charset('gb2312'),一般网页则为utf-8,可以直接使用.charset('utf-8')
        superagent.get(baseUrl + route)
            .charset('gb2312')
            .end(function (err, sres) {
                var items = [];
                if (err) {
                    console.log('ERR: ' + err);
                    res.json({ code: 400, msg: err, sets: items });
                    return;
                }
                var $ = cheerio.load(sres.text);
                $('div.g-main-bg ul.g-gxlist-imgbox li a').each(function (idx, element) {
                    var $element = $(element);
                    var $subElement = $element.find('img');
                    var thumbImgSrc = $subElement.attr('src');
                    items.push({
                        title: $(element).attr('title'),
                        href: $element.attr('href'),
                        thumbSrc: thumbImgSrc
                    });
                });
                if (!checkDir) {
                    fs.mkdir('assets', function (error) {
                        if (error) {
                            console.log(error);
                            return false;
                        }
                        console.log('创建目录成功');
                    })
                }
                fs.access(path.join(__dirname, '/img.json'), fs.constants.F_OK, err => {
                    if (err) { // 文件不存在
                        fs.writeFile(path.join(__dirname, '/img.json'), JSON.stringify([
                            {
                                route,
                                items
                            }
                        ]), err => {
                            if (err) {
                                console.log(err)
                                return false
                            }
                            console.log('保存成功')
                        })
                    } else {
                        fs.readFile(path.join(__dirname, '/img.json'), (err, data) => {
                            if (err) {
                                return false
                            }
                            data = JSON.parse(data.toString())
                            let exist = data.some((page, index) => {
                                return page.route == route
                            })
                            if (!exist) {
                                fs.writeFile(path.join(__dirname, 'img.json'), JSON.stringify([
                                    ...data,
                                    {
                                        route,
                                        items
                                    },
                                ]), err => {
                                    if (err) {
                                        return false
                                    }
                                })
                            }
                        })
                    }
                    res.json({ code: 200, msg: "", data: items });
                })
                try {
                    fs.readFile(path.join(__dirname, '/img.json'), (err, data) => {
                        if (err) {
                            return false
                        }else{
                            data = JSON.parse(data.toString());
                            data.map((v, i) => {
                                v.items.map((v,i) => {
                                    i = request(v.thumbSrc)
                                    // 后缀.jpg可用正则匹配
                                    i.pipe(fs.createWriteStream('./assets/' + v.title + '.jpg'));
                                })
        
                            })
                        }
                    })
                } catch(err){}
            })
    });
    app.get('/show', (req, res) => {
        fs.readFile(path.join(__dirname, 'img.json'), (err, data) => {
            if (err) {
                console.log(err)
                return false
            }
            res.json(data.toString())
        })
    })
    var server = app.listen(8081, function () {
        var host = server.address().address
        var port = server.address().port
    })
    View Code

    3.static文件夹下index.html

    <!DOCTYPE html>
    <html lang="en">
    <head>
      <meta charset="UTF-8">
      <meta name="viewport" content="width=device-width, initial-scale=1.0">
      <meta http-equiv="X-UA-Compatible" content="ie=edge">
      <title>Document</title>
    </head>
    <body>
      <script src="./index.js"></script>
    </body>
    </html>

    4.static文件夹下index.js

    fetch('/index', {
      method: 'GET'
    }).then(res => {
      return res.json()
    }).then(res => {
    
      if (res.code == 200) {
        fetch('/show', {
          method: 'GET'
        }).then(res => {
          return res.json()
        }).then(res => {
          res = JSON.parse(res)
          console.log(res, res.length)
          document.body.innerHTML = res.map((page, index) => {
            console.log(page)
            return page.items.map((item, itemIndex) => {
              return `<a href="${item.thumbSrc}" ><img src="${item.thumbSrc}" width="200" height="200"/></a>`
            }).join('')
          }).join('')
        })
      }
    
    })

    总结

    写到这里基本是结束了,对于node我还是怀着一个敬畏的心,摸摸索索终于把这个demo写完了,项目也传到gitHub了如有需要可私信

  • 相关阅读:
    商业研究(10):比呀比-海淘电商-海外正品直邮-高品质跨境购物平台
    商业研究(10):比呀比-海淘电商-海外正品直邮-高品质跨境购物平台
    POJ 2923 Relocation
    ZJU 2425 Inversion
    HDU 1394 Minimum Inversion Number
    POJ 1787 Charlie's Change
    HDU 2191 悼念512汶川大地震
    Gym
    HDU 2069 Coin Change
    UVA 674 Coin Change
  • 原文地址:https://www.cnblogs.com/gaoht/p/11303611.html
Copyright © 2020-2023  润新知