广州阿里云代理商：ajax+node+request爬取网络图片的实例(宅男福利)

核心代码实现

// 使用Express创建Web服务器
const express = require('express');
const request = require('request');
const fs = require('fs');
const app = express();

// 阿里云OSS配置
const OSS = require('ali-oss');
const client = new OSS({
  region: 'oss-cn-guangzhou',
  accessKeyId: 'YOUR_ACCESS_KEY',
  accessKeySecret: 'YOUR_SECRET_KEY',
  bucket: 'image-crawler-bucket'
});

// 图片爬取API
app.get('/crawl-images', (req, res) => {
  const targetUrl = req.query.url;
  
  // 使用request获取网页内容
  request(targetUrl, (error, response, body) => {
    if (error) {
      return res.status(500).json({ error: '网页请求失败' });
    }
    
    // 使用正则提取图片URL
    const imageRegex = /<img[^>]+src="([^">]+)"/g;
    const images = [];
    let match;
    
    while ((match = imageRegex.exec(body)) !== null) {
      images.push(match[1]);
    }
    
    // 下载并存储图片到阿里云OSS
    const uploadPromises = images.map((imgUrl, index) => {
      return new Promise((resolve) => {
        const filename = `image_${Date.now()}_${index}.jpg`;
        const writeStream = fs.createWriteStream(filename);
        
        request(imgUrl).pipe(writeStream).on('close', () => {
          // 上传到阿里云OSS
          client.put(filename, filename).then(result => {
            fs.unlinkSync(filename); // 删除本地临时文件
            resolve(result.url);
          });
        });
      });
    });
    
    // 返回结果
    Promise.all(uploadPromises).then(urls => {
      res.json({ 
        originalUrl: targetUrl,
        imageCount: urls.length,
        imageUrls: urls,
        storage: '阿里云OSS'
      });
    });
  });
});

// 启动服务器
app.listen(3000, () => {
  console.log('爬虫服务器运行在 http://localhost:3000');
});