ATO25 to Markdown

markdownit.js

// Made with GPT 4.1
// Run this in the console at https://2025.allthingsopen.org/schedule?dy=monday
// to get the full conference schedule as markdown.
//
// Note that this does scrap descriptions for each talk's page, so do be kind to the website
// and don't run this unnecessarily.

// 1. Load Turndown if not already loaded
if (typeof TurndownService === 'undefined') {
  await new Promise(resolve => {
    const script = document.createElement('script');
    script.src = 'https://unpkg.com/turndown/dist/turndown.js';
    script.onload = resolve;
    document.head.appendChild(script);
  });
}

// 2. Main extraction function
async function extractScheduleWithDescriptions() {
  function formatDate(dateStr) {
    const months = {
      January: "01", February: "02", March: "03", April: "04",
      May: "05", June: "06", July: "07", August: "08",
      September: "09", October: "10", November: "11", December: "12"
    };
    const [month, day, year] = dateStr.replace(',', '').split(' ');
    return `${year}-${months[month]}-${day.padStart(2, '0')}`;
  }

  const dateSpan = document.querySelector('.col.ar.c3 span');
  const date = formatDate(dateSpan.textContent.trim());
  let md = `## ${date}\n\n`;

  const turndownService = new TurndownService();

  async function fetchDescription(url) {
    try {
      const resp = await fetch(url);
      const html = await resp.text();
      const parser = new DOMParser();
      const doc = parser.parseFromString(html, "text/html");
      const frame = doc.querySelector('.page-frame.session-single');
      if (!frame) return '';

      const titleEl = frame.querySelector('.page-title');
      const speakersHeader = frame.querySelector('.session-speakers-header');
      if (!titleEl || !speakersHeader) return '';

      let htmlDesc = '';
      let node = titleEl.nextElementSibling;
      while (node && node !== speakersHeader) {
        htmlDesc += node.outerHTML || '';
        node = node.nextElementSibling;
      }

      return turndownService.turndown(htmlDesc).trim();
    } catch (e) {
      return '';
    }
  }

  const timeBlocks = Array.from(document.querySelectorAll('.time-block'));
  for (const timeBlock of timeBlocks) {
    const rangeDiv = timeBlock.querySelector('.time-block-range');
    if (!rangeDiv) continue;
    const timeRange = rangeDiv.childNodes[0].textContent.trim();
    const timezone = rangeDiv.querySelector('small')?.textContent.trim() || '';
    md += `### ${timeRange} ${timezone}\n\n`;

    const sessions = Array.from(timeBlock.querySelectorAll('.time-block-session-item'));
    for (const session of sessions) {
      const cat = session.querySelector('.time-block-session-item-cat')?.textContent.trim() || '';
      const linkEl = session.querySelector('h5 a');
      const title = linkEl?.textContent.trim() || '';
      const link = linkEl?.href || '';
      const speaker = session.querySelector('.time-block-session-item-speaker span')?.textContent.trim() || '';

      let description = '';
      if (link) {
        description = await fetchDescription(link);
      }

      md += `#### ${title}\n\n`;
      md += `category: ${cat}  \n`;
      md += `link: ${link}  \n`;
      if (speaker){
        md += `speaker: ${speaker}\n`;
      }
      if (description) {
        md += `\n${description}\n`;
      }
      md += `\n`;
    }
  }

  return md;
}

// 3. Run the extraction and print the result
(async () => {
  const markdown = await extractScheduleWithDescriptions();
  console.log(markdown);
})();