a sample code for puppeteer

index.mjs

import { createClient, destroyPage } from './puppeteerClient.mjs';

const waitMs = (msec) => new Promise(res => setTimeout(() => res(), msec));

const getLastPageFromPage = async (page) => {
    return await page.evaluate(() => {
        const urls = Array.from(document.querySelectorAll('.pagination a[href]'))
        if (urls.length) {
            const lastUrl = urls.pop().href;
            return lastUrl.match(/\d+/g).map(Number)[0];
        }
        return 1;
    });
};

const extractUrl = (u) => {
    if (u) {
        const p = u.indexOf('"');
        const l = u.lastIndexOf('"');
        if (p >= 0 && l >= 0) {
            return u.substring(p + 1, l);
        }
    }
    return undefined;
}

const getAlbumsInfosFromPage = async (page) => {
    const titles = await page.evaluate(() =>
        Array.from(document.querySelectorAll('.info')).map(x => x.innerText)
    );

    if (!titles?.length) return {};

    const res = await page.evaluate(() =>
        Array.from(document.querySelectorAll('a.albumbgphoto'))
            .map(x => ({
                url: x.href,
                thumbnailUrl: x.style['background-image'],
            }))
    );
    const urls = res.map(x => ({
        ...x,
        thumbnailUrl: extractUrl(x.thumbnailUrl),
    }));

    return titles.map((title, idx) => ({
        title,
        ...urls[idx],
    }));
}

const client = await createClient({ debug: process.env.DEBUG });

const LOGIN_URL = '.../login.html';
const LOGINED_URL = '.../main.html';

try {
  const page = client.page;
  await page.goto(LOGIN_URL);
  if (page.url() === LOGINED_URL) {
     console.info('has been logined');
  } else {
    await page.waitForSelector('#loginBtn', { timeout: 5000 });
    await waitMs(2000);
    await page.type('input#account', user, { delay: 180 });
    await page.type('input#password', pass, { delay: 180 });
    await page.click('#loginBtn', { delay: 1000 });
    await page.waitForNavigation();
    if (page.url() !== LOGINED_URL) {
      throw new Error(`logined url not match ${page.url()} != ${LOGINED_URL}`);
    }
  }
  const lastPage = await getLastPageFromPage(page) || 1;
  console.log('last page', lastPage);
  const infos = await getAlbumsInfosFromPage(page);
  console.log(JSON.stringify(infos, null, 2));
} catch (error) {
  throw error;
}
finally {
  porgressBar?.stop();
  destroyPage(client);
}

puppeteerClient.mjs

import puppeteer from 'puppeteer';
import path from 'node:path';
import { fileURLToPath } from 'url';

export const createClient = async (config = {}) => {
    const { debug } = config;
    const args = [
        '--no-sandbox',
        '--disable-setuid-sandbox',
        '--disable-dev-shm-usage',
        '--disable-gpu',
        '--mute-audio',
        '--disable-notifications',
    ];

    // __dirname not defined in mjs
    const __filename = fileURLToPath(import.meta.url);
    const __dirname = path.dirname(__filename);
    console.debug(__dirname)

    const browser = await puppeteer.launch({
        headless: debug ? false : 'new',
        userDataDir: path.join(__dirname, '..', 'puppeteerCache'),
        args,
    });

    const page = await browser.newPage();

    await page.setViewport({
        width: 1024,
        height: 768,
    });

    return { browser, page, args, debug };
};

export const destroyPage = async (client = {}) => {
    const { browser, page } = client;
    if (!browser) {
        console.warn('not found browser');
        return;
    }

    if (page) {
        await page.close();
    } else {
        console.warn('not found page');
    }

    await browser.close();
};