Metify/src/app/actions.ts

"use server";

import * as cheerio from "cheerio";

interface FaqItem {
  question: string;
  answer: string;
}

export interface HeadlineNode {
  tag: string;
  text: string;
  length: number;
  level: number;
  children: HeadlineNode[];
}

export interface ImageAltData {
  src: string;
  alt: string;
}

export async function extractMetaData(url: string, keyword?: string) {
  if (!url) {
    return { error: "URL is required." };
  }

  let formattedUrl = url;
  if (!/^https?:\/\//i.test(url)) {
    formattedUrl = `https://${url}`;
  }

  try {
    const response = await fetch(formattedUrl, {
      headers: {
        "User-Agent":
          "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36",
      },
    });

    if (!response.ok) {
      return { error: `Failed to fetch URL. Status: ${response.status}` };
    }

    const html = await response.text();
    const $ = cheerio.load(html);

    const title =
      $('meta[property="og:title"]').attr("content") ||
      $("title").text() ||
      "No title found";
    const description =
      $('meta[property="og:description"]').attr("content") ||
      $('meta[name="description"]').attr("content") ||
      "No description found";
    const image = $('meta[property="og:image"]').attr("content") || null;

    const faqData: FaqItem[] = [];
    $('script[type="application/ld+json"]').each((i, el) => {
      const jsonContent = $(el).html();
      if (!jsonContent) return;

      try {
        const data = JSON.parse(jsonContent);
        const graph = data["@graph"] || [data];

        for (const item of graph) {
          if (item["@type"] === "FAQPage" && Array.isArray(item.mainEntity)) {
            item.mainEntity.forEach((qa: any) => {
              if (
                qa["@type"] === "Question" &&
                qa.name &&
                qa.acceptedAnswer &&
                qa.acceptedAnswer.text
              ) {
                faqData.push({
                  question: qa.name,
                  answer: qa.acceptedAnswer.text,
                });
              }
            });
          }
        }
      } catch (e) {
        // Ignore parsing errors
      }
    });

    const headlines: HeadlineNode[] = [];
    const path: HeadlineNode[] = [];

    $("h1, h2, h3, h4, h5, h6").each((i, el) => {
      const tag = $(el).prop("tagName").toLowerCase();
      const text = $(el).text().trim();
      if (!text) return;

      const level = parseInt(tag.replace("h", ""), 10);

      const node: HeadlineNode = {
        tag,
        text,
        length: text.length,
        level,
        children: [],
      };

      while (path.length > 0 && path[path.length - 1].level >= level) {
        path.pop();
      }

      if (path.length === 0) {
        headlines.push(node);
      } else {
        path[path.length - 1].children.push(node);
      }

      path.push(node);
    });

    let keywordCount: number | null = null;
    const trimmedKeyword = keyword?.trim();
    if (trimmedKeyword) {
      $("script, style").remove();
      const bodyText = $("body").text();
      const regex = new RegExp(trimmedKeyword, "gi");
      const matches = bodyText.match(regex);
      keywordCount = matches ? matches.length : 0;
    }

    const imageAltData: ImageAltData[] = [];
    $("img").each((i, el) => {
      const src = $(el).attr("src");
      const alt = $(el).attr("alt") || "";

      if (src) {
        try {
          const absoluteSrc = new URL(src, formattedUrl).href;
          imageAltData.push({
            src: absoluteSrc,
            alt: alt.trim(),
          });
        } catch (e) {
          // Ignore invalid URLs
        }
      }
    });

    return {
      data: {
        title,
        description,
        image,
        faq: faqData.length > 0 ? faqData : null,
        headlines: headlines.length > 0 ? headlines : null,
        keyword: trimmedKeyword || null,
        keywordCount,
        images: imageAltData.length > 0 ? imageAltData : null,
      },
    };
  } catch (error) {
    console.error(error);
    if (error instanceof Error && error.message.includes("Invalid URL")) {
      return {
        error: "The provided URL is not valid. Please check and try again.",
      };
    }
    return { error: "An unexpected error occurred while fetching the URL." };
  }
}