Files
bidding_watcher/src/crawler/services/chng_target.spec.ts

153 lines
4.6 KiB
TypeScript
Raw Normal View History

import { ChngCrawler } from './chng_target';
import * as puppeteer from 'puppeteer';
// Increase timeout to 120 seconds for manual inspection and slow sites
jest.setTimeout(120000);
// 获取代理配置
const getProxyArgs = (): string[] => {
const proxyHost = process.env.PROXY_HOST;
const proxyPort = process.env.PROXY_PORT;
const proxyUsername = process.env.PROXY_USERNAME;
const proxyPassword = process.env.PROXY_PASSWORD;
if (proxyHost && proxyPort) {
const args = [`--proxy-server=${proxyHost}:${proxyPort}`];
if (proxyUsername && proxyPassword) {
args.push(`--proxy-auth=${proxyUsername}:${proxyPassword}`);
}
return args;
}
return [];
};
// 模拟人类鼠标移动
async function simulateHumanMouseMovement(page: puppeteer.Page) {
const viewport = page.viewport();
if (!viewport) return;
const movements = 5 + Math.floor(Math.random() * 5); // 5-10次随机移动
for (let i = 0; i < movements; i++) {
const x = Math.floor(Math.random() * viewport.width);
const y = Math.floor(Math.random() * viewport.height);
await page.mouse.move(x, y, {
steps: 10 + Math.floor(Math.random() * 20) // 10-30步使移动更平滑
});
// 随机停顿 100-500ms
await new Promise(r => setTimeout(r, 100 + Math.random() * 400));
}
}
// 模拟人类滚动
async function simulateHumanScrolling(page: puppeteer.Page) {
const scrollCount = 3 + Math.floor(Math.random() * 5); // 3-7次滚动
for (let i = 0; i < scrollCount; i++) {
const scrollDistance = 100 + Math.floor(Math.random() * 400); // 100-500px
await page.evaluate((distance) => {
window.scrollBy({
top: distance,
behavior: 'smooth'
});
}, scrollDistance);
// 随机停顿 500-1500ms
await new Promise(r => setTimeout(r, 500 + Math.random() * 1000));
}
// 滚动回顶部
await page.evaluate(() => {
window.scrollTo({ top: 0, behavior: 'smooth' });
});
await new Promise(r => setTimeout(r, 1000));
}
describe('ChngCrawler Real Site Test', () => {
let browser: puppeteer.Browser;
beforeAll(async () => {
const proxyArgs = getProxyArgs();
if (proxyArgs.length > 0) {
console.log('Using proxy:', proxyArgs.join(' '));
}
browser = await puppeteer.launch({
headless: false, // Run in non-headless mode
args: [
'--no-sandbox',
'--disable-setuid-sandbox',
'--disable-blink-features=AutomationControlled',
'--window-size=1920,1080',
"--disable-infobars",
...proxyArgs,
// "--headless=new",
// '--disable-dev-shm-usage',
// '--disable-accelerated-2d-canvas',
// '--no-first-run',
// '--no-zygote',
// '--disable-gpu',
// '--disable-features=VizDisplayCompositor',
// '--disable-webgl',
// '--disable-javascript',
],
defaultViewport: null
});
});
afterAll(async () => {
if (browser) {
// Keep open for a few seconds after test to see result
await new Promise(r => setTimeout(r, 50000));
await browser.close();
}
});
it('should visit the website and list all found bid information', async () => {
console.log(`
Starting crawl for: ${ChngCrawler.name}`);
console.log(`Target URL: ${ChngCrawler.url}`);
// 创建一个临时页面用于模拟人类行为
const tempPage = await browser.newPage();
await tempPage.setViewport({ width: 1920, height: 1080, deviceScaleFactor: 1 });
// 模拟人类鼠标移动
console.log('Simulating human mouse movements...');
await simulateHumanMouseMovement(tempPage);
// 模拟人类滚动
console.log('Simulating human scrolling...');
await simulateHumanScrolling(tempPage);
await tempPage.close();
const results = await ChngCrawler.crawl(browser);
console.log(`
Successfully found ${results.length} items:
`);
console.log('----------------------------------------');
results.forEach((item, index) => {
console.log(`${index + 1}. [${item.publishDate.toLocaleDateString()}] ${item.title}`);
console.log(` Link: ${item.url}`);
console.log('----------------------------------------');
});
expect(results).toBeDefined();
expect(Array.isArray(results)).toBeTruthy();
if (results.length === 0) {
console.warn('Warning: No items found. Observe the browser window to see if content is loading or if there is a verification challenge.');
} else {
const firstItem = results[0];
expect(firstItem.title).toBeTruthy();
expect(firstItem.url).toMatch(/^https?:\/\//);
expect(firstItem.publishDate).toBeInstanceOf(Date);
}
});
});