recursivly convert html to markdown using nodejs

recursivly-convert-html-markdown.js

const fs = require("fs");
const path = require("path");
const glob = require("glob");
const TurndownService = require("turndown");

const turndownService = new TurndownService({
  // options
  headingStyle: "atx",
  bulletListMarker: "-",
  linkStyle: "referenced",
  linkReferenceStyle: "full",
});

glob("site/**/*.html", (err, files) => {
  if (err) {
    console.log(err);
  }
  if (files) {
      files.forEach((file) => {
        if (!fs.existsSync('out/' + path.dirname(file))) {
            fs.mkdirSync('out/' + path.dirname(file), { recursive: true })
        }
        console.log(file);
      const html = fs.readFileSync(file).toString();

      const textArray = turndownService.turndown(html).split("\n");
      const text = `${file.replace('site/', 'http://')}\r\n` + textArray
          .map((line) => {
            line = line.endsWith('.html') ? line.replace('.html', '') : line
            return line.includes('turn\\_client\\_track\\_id = "";', "") ? line.split('turn\\_client\\_track\\_id = "";')[1] : line
          })
          .filter((line) => {
            //   console.log(line);
            return (
              line !== "\r\n" &&
              line !== " " &&
              line.trim() !== "-   Search" &&
              line.trim() !== "Search" &&
              line.trim() !== "- Search" &&
              line.trim() !== "- Previous" &&
              line.trim() !== "- |" &&
              line.trim() !== "- Next" &&
              !/\[\]\[(.*?)\]/g.test(line) &&
              !line.startsWith("window.dataLayer") &&
              !line.startsWith("!function") &&
              !line.startsWith("![](") &&
              !line.startsWith("function ") &&
              !line.trim().startsWith("jQuery(function()") &&
              !line.startsWith("<img ") &&
              !line.startsWith("### _!") &&
              !line.startsWith("// ")
            );
          })
          .join("\r\n")
          .replace(/  \r\n    /g, " ")
          .replace(/]: /g, "]: https://www.goodwillcentraltexas.org/")
          .replace(/-   -   /g, "- ")
          .replace(/    -   /g, "  - ")
          .replace(/-   /g, "- ")
          // .replace(/ /g, "")
          // .replace(/ /g, "")
          // .replace(/ /g, "")
          .replace(/https:\/\/www.goodwillcentraltexas.org\/http/g, "http")
          .replace(
            "\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n\r\n",
            "\r\n"
          )
          .replace(/\r\n\r\n\r\n/g, "\r\n\r\n")
          .replace(/\r\n\r\n/g, "\r\n\r\n");
      // fs.writeFileSync("out.json", JSON.stringify(text, null, 2));
      fs.writeFileSync(`./out/${file.replace(".html", ".md")}`, text);
    });
  }
  console.log(`complete! processed: ${files.length} files`);
});