ListingExtractor/index.js

const puppeteer = require('puppeteer')


var url = '';
process.argv.forEach((val, index, array) => {
    console.log(val)
    if(val.startsWith('url=')){
        url = val.substring(4);
    }
})
if(!url || url.length === 0) {
    console.log('Need to supply url');
    process.exit();
}
console.log("hello?")

const run = async () => {
    const browser = await puppeteer.launch({
        headless:false
    });
    
    const page = await browser.newPage();
  
    await page.goto(url);
    
    // Type into search box.
    // await page.type('.devsite-search-field', 'Headless Chrome');
  
    // Wait for suggest overlay to appear and click "show all results".
    const allResultsSelector = '.gallery';
    await page.waitForSelector(allResultsSelector);


    const vinSelector = 'body > main > div.listing > div:nth-child(3) > div.column.column-right.column-right-force > div.essentials > div:nth-child(5) > ul > li:nth-child(1) > a';
    await page.waitForSelector(vinSelector);
    let element = await page.$(vinSelector)
    let vin = await page.evaluate(el => el.textContent, element);
    console.log(vin);
    // await page.click(allResultsSelector);

    const client = await page.target().createCDPSession()
    await client.send('Page.setDownloadBehavior', {
    behavior: 'allow',
    downloadPath: './images',
    })
    const firstImageLinkSelector = '.gallery > a:nth-child(1)';
    await page.click(firstImageLinkSelector);
    await page.waitForSelector('.pswp__img')
    console.log('image ready?')
    await page.evaluate(downloadGallery, vin);
    console.log('all done');
    // Print all the files.
    // console.log(links.join('\n'));
  
    await browser.close();
  };

  downloadGallery = (vin) => {
    async function ensureCarouselVisible() {
        var imgWrap;
        do {
            imgWrap = document.elementFromPoint(100, 100);
            console.log(imgWrap.classList);
            await delay(50);
        } while(imgWrap.classList.contains('pswp__img--placeholder'));
        
      }
      
      function delay(ms) {
        return new Promise((resolve) => setTimeout(resolve, ms));
      }
      
      // Find the active image, surround it in an anchor tag, then click it.
      async function downloadImage(id) {
        const imgWrap = document.elementFromPoint(300, 300);
        const children = imgWrap.querySelectorAll('.pswp__img');
        const img = children[children.length - 1];
        console.log(img, imgWrap);
        // Full image hasn't loaded yet
        const src = imgWrap.src.split('?')[0]// get rid of querystring
        return downloadSrc(src, id);
      }
      
      function downloadSrc(src, id) {
        const a = document.createElement('a');
        a.href = src;
        a.download = `${id}`;
        console.log(src, a);
        a.click();
      }
      
      function nextImage() {
        document.querySelector('.pswp__button.pswp__button--arrow--right').click();
      }
      
      function getCounterValue() {
        const [position, total] = document.querySelector('.pswp__counter').textContent.split('/');
        return parseInt(position.trim(), 10);
      }
      
      async function run() {
        await ensureCarouselVisible();
        await delay(500);
        const firstValue = getCounterValue();
      
        async function imageLoop() {
          await downloadImage(`${vin}-${getCounterValue()}`);
          nextImage();
          await delay(100);
          // console.log('counter', getCounterValue());
          if (getCounterValue() !== firstValue) return imageLoop(); // recurse
        }
        return imageLoop();
      }
      
      return run().then(async () => await delay(1000));
  }

  run();