Automate Your Daily News Digest Summarize with Gemini and Get Daily Email Updates¶

/**
 * Summarizes the provided text using the Gemini API.
 * @param {string} text The text of the article to summarize.
 * @return {string} The summarized text.
 */
function summarizeWithGemini(text) {
  const GEMINI_API_KEY = PropertiesService.getScriptProperties().getProperty("GEMINI_API_KEY");
  const endpoint = `https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:generateContent?key=${GEMINI_API_KEY}`;

  const prompt = `Summarize the following news article in a single paragraph of no more than 200 words. Focus on the main events and key facts. The article is: \n\n${text}`;

  const requestBody = {
    contents: [
      {
        parts: [{ text: prompt }],
      },
    ],
  };

  const options = {
    method: "post",
    contentType: "application/json",
    payload: JSON.stringify(requestBody),
  };

  try {
    const response = UrlFetchApp.fetch(endpoint, options);
    const data = JSON.parse(response.getContentText());

    // Check if the response contains a candidate with parts
    if (
      data.candidates &&
      data.candidates.length > 0 &&
      data.candidates[0].content.parts.length > 0
    ) {
      return data.candidates[0].content.parts[0].text;
    } else {
      console.error(
        "Gemini API response did not contain a valid summary.",
        data
      );
      return "Could not generate a summary.";
    }
  } catch (e) {
    console.error("Error calling Gemini API: " + e.message);
    return "Could not generate a summary due to an error.";
  }
}

/**
 * Fetches news from an RSS feed, summarizes each article, and returns the results.
 * @return {Array<Object>} An array of objects, each containing an article's title, link, and summary.
 */
function getNewsSummaries() {
  const rssFeeds = [
    { name: "Adnan Kaya", url: "https://www.example.com.tr/en/rss/"},
    // add more news website rss
  ];

  let summaries = [];
  const MAX_NEWS_AMOUNT = 12; // you can change here
  const SLEEP_SECONDS = 4000; // you can change here

  for (const feed of rssFeeds) {
    const rssUrl = feed.url;
    const sourceName = feed.name;
    try {
      Logger.log(`------- Processing for ${rssUrl} -------`);

      const response = UrlFetchApp.fetch(rssUrl);
      if (response.getResponseCode() !== 200) {
        throw new Error(
          `Failed to fetch URL with status code: ${response.getResponseCode()}`
        );
      }

      const xml = response.getContentText();
      const document = XmlService.parse(xml);
      const root = document.getRootElement();

      if (!root) {
        throw new Error("Root element is null. Invalid XML structure.");
      }

      let items = [];
      let namespace = null;
      let titleElement = "title";
      let linkElement = "link";

      const rootName = root.getName();

      // Determine the feed format and adjust parsing logic
      if (rootName === "rss") {
        // RSS 2.0 format (most common)
        namespace = root.getNamespace();
        const channel = root.getChild("channel", namespace);
        if (channel) {
          items = channel.getChildren("item", namespace);
        }
      } else if (rootName === "feed") {
        // Atom format
        namespace = root.getNamespace();
        items = root.getChildren("entry", namespace);
        titleElement = "title";
        linkElement = "link";
      } else if (rootName === "RDF") {
        // RSS 1.0 (RDF) format
        const rss1Namespace = XmlService.getNamespace(
          "http://purl.org/rss/1.0/"
        );
        items = root.getChildren("item", rss1Namespace);
        // Set the namespace for title/link extraction to the RSS 1.0 namespace
        namespace = rss1Namespace;
      } else {
        throw new Error(`Unsupported RSS feed format for URL: ${rssUrl}`);
      }

      if (items.length === 0) {
        Logger.log(`No items found in feed: ${rssUrl}`);
        continue;
      }

      // Limit to a few articles
      for (let i = 0; i < Math.min(items.length, MAX_NEWS_AMOUNT); i++) {
        try {
          const item = items[i];
          const title = item.getChildText(titleElement, namespace);

          let link = item.getChildText(linkElement, namespace);
          const pubDate =
            item.getChildText("pubDate", namespace) ||
            item.getChildText("updated", namespace);

          // Handle Atom link format where link is an element with an attribute
          if (!link && rootName === "feed") {
            const linkElementNode = item.getChild(linkElement, namespace);
            if (linkElementNode) {
              link = linkElementNode.getAttribute("href")?.getValue();
            }
          }

          if (!link) {
            throw new Error("Article link is null or missing.");
          }

          const articleHtml = UrlFetchApp.fetch(link).getContentText();
          const articleText = extractTextFromHtml(articleHtml);

          if (!articleText || articleText.trim() === "") {
            throw new Error("Extracted article text is empty.");
          }

          // Assuming summarizeWithGemini is a function you have defined elsewhere
          const summary = summarizeWithGemini(articleText);

          summaries.push({
            title: title,
            link: link,
            summary: summary,
            source: sourceName,
            pubDate: pubDate,
          });

          // Add a sleep function to stay within the API quota
          Utilities.sleep(SLEEP_SECONDS);

        } catch (e) {
          console.error(
            `Error processing article from ${rssUrl}: ${e.message}`
          );
          continue;
        }
      }
    } catch (e) {
      console.error(`Error processing RSS feed ${rssUrl}: ${e.message}`);
    }
  }

  return summaries;
}

/**
 * A basic function to attempt to extract the main article text from HTML.
 * This will likely need to be customized for specific websites.
 * @param {string} html The HTML content of the webpage.
 * @return {string} The extracted text.
 */
function extractTextFromHtml(html) {
  // Use a regular expression to try and find the main article content within common tags
  // This is a very rough approach but can work for some sites.
  const regex = /<p[^>]*>(.*?)<\/p>/g;
  let matches = [];
  let match;
  while ((match = regex.exec(html)) !== null) {
    // Basic cleanup of HTML tags
    matches.push(match[1].replace(/<[^>]*>/g, ""));
  }
  return matches.join("\n\n");
}

/**
 * Fetches news summaries and sends them as an email digest.
 */
function sendDailyDigest() {
  const summaries = getNewsSummaries();
  const botName = "daily news bot;
  const recipient = "to-adnankaya@example.com"; // change here!!!!
  const mailFrom = "from-adnankaya@example.com"; //  change here!!!!!
  const subject = `Your Daily News Digest - ${new Date().toLocaleDateString()}`;

  // Start building the HTML email body with a container table
  let htmlEmailBody = `
    <html>
      <body style="font-family: Arial, sans-serif; background-color: #f4f4f4; margin: 0; padding: 20px;">
        <table width="100%" border="0" cellspacing="0" cellpadding="0">
          <tr>
            <td align="center">
              <table width="600" border="0" cellspacing="0" cellpadding="20" style="background-color: #ffffff; border-radius: 8px;">
                <tr>
                  <td>
                    <h1 style="color: #333333; font-size: 24px;">Daily News Digest</h1>
                    <p style="color: #666666; font-size: 14px;">Here is your daily dose of news summaries:</p>
                    <hr style="border: 0; height: 1px; background: #ddd; margin: 20px 0;">
  `;

  if (summaries.length === 0) {
    htmlEmailBody += `
      <p style="color: #666666; font-style: italic;">No news articles were summarized today.</p>
    `;
  } else {
    for (const item of summaries) {
      let formattedDate = item.pubDate
        ? new Date(item.pubDate).toLocaleString()
        : "";
      htmlEmailBody += `
        <table width="100%" border="0" cellspacing="0" cellpadding="0" style="margin-bottom: 20px;">
          <tr>
            <td>
              <h3 style="color: #0056b3; font-size: 18px; margin-top: 0; margin-bottom: 5px;">${
                item.title
              }</h3>
              <p style="color: #666666; font-size: 12px; margin-top: 0; margin-bottom: 10px;">
                <span style="font-weight: bold;">
                ${
                  formattedDate
                    ? ` | <span style="font-weight: bold;">Published:</span> ${formattedDate}`
                    : ""
                }
              </p>
              <p style="color: #333333; font-size: 14px; margin-top: 0;">${item.summary.replace(
                /\n/g,
                "<br>"
              )}</p>
              <a href="${
                item.link
              }" style="color: #007BFF; font-size: 12px; text-decoration: none;">Read the full article here.</a>
            </td>
          </tr>
        </table>
        <p style="color: #ccc; font-size: 12px; margin-top: 25px; margin-bottom: 5px;">${
          item.source
        }</p>
        <hr style="border: 0; height: 1px; background: #eee; margin: 20px 0;">
      `;
    }
  }

  // End the HTML body
  htmlEmailBody += `
                  </td>
                </tr>
              </table>
            </td>
          </tr>
        </table>
      </body>
    </html>
  `;
  var options = {
    name: botName,
    from: mailFrom
    htmlBody: htmlEmailBody,
    charset: "UTF-8",
  };
  GmailApp.sendEmail(recipient, subject, "", options);
  //Logger.log("Email sending attempt completed. Check your inbox and Apps Script Executions for status.");
}