video2document/services/modules/convert/convert.js

const fs = require("fs");
const path = require("path");
const puppeteer = require("puppeteer");
const htmlToDocx = require("html-to-docx");
const { execSync } = require("child_process");
const os = require("os");

const outputDir = path.join(__dirname, "../../../storage/documents");

if (!fs.existsSync(outputDir)) {
  fs.mkdirSync(outputDir, { recursive: true });
}

async function showSaveDialog(defaultName, format) {
  const platform = os.platform();

  if (platform === "darwin") {
    // macOS
    const applescript = `
            set defaultName to "${defaultName}.${format}"
            set theFile to choose file name with prompt "Dokument speichern als:" default name defaultName
            POSIX path of theFile
        `;

    try {
      const result = execSync(`osascript -e '${applescript}'`, {
        encoding: "utf8",
      });
      return result.trim();
    } catch (err) {
      if (err.status === 1) return null; // User canceled
      throw err;
    }
  } else if (platform === "win32") {
    const safeName = decodeURIComponent(defaultName);

    const powershell = `
        Add-Type -AssemblyName System.Windows.Forms;
        $dialog = New-Object System.Windows.Forms.SaveFileDialog;
        $dialog.FileName = '${safeName}.${format}';
        $dialog.Filter = '${format.toUpperCase()} Dateien (*.${format})|*.${format}|Alle Dateien (*.*)|*.*';
        $dialog.Title = 'Dokument speichern als';
        $result = $dialog.ShowDialog();
        if ($result -eq 'OK') { $dialog.FileName }
    `;

    try {
      const result = execSync(
        `powershell -NoProfile -Command "${powershell.replace(/\r?\n/g, " ")}"`,
        { encoding: "utf8" },
      );
      return result.trim() || null;
    } catch (err) {
      if (err.status === 1) return null; // User cancelled
      throw new Error("Save dialog failed: " + err.message);
    }
  } else {
    // Linux - zenity oder kdialog
    try {
      const result = execSync(
        `zenity --file-selection --save --confirm-overwrite --filename="${defaultName}.${format}"`,
        { encoding: "utf8" },
      );
      return result.trim();
    } catch (err) {
      try {
        const result = execSync(
          `kdialog --getsavefilename . "${defaultName}.${format}"`,
          { encoding: "utf8" },
        );
        return result.trim();
      } catch (err2) {
        // Fallback
        return path.join(os.homedir(), "Downloads", `${defaultName}.${format}`);
      }
    }
  }
}

const module_exports = {
  name: "htmlDocumentConverter",
  type: "converter",
  displayname: "HTML Document Converter",
  description: "Converts LLM-generated HTML to PDF, DOCX, TXT, or HTML",

  /**
   * Main conversion function
   * @param {Object} options
   * @param {string} options.inputPath - Path to the HTML input
   * @param {string} options.format - 'pdf' | 'docx' | 'html' | 'txt'
   * @param {string} [options.outputName] - Optional output filename (without extension)
   * @param {boolean} [options.showDialog] - Show save dialog (default: false in module mode, true in CLI mode)
   */
  async convert({ inputPath, format = "pdf", outputName, showDialog = false }) {
    format = format.toLowerCase().replace(".", ""); // <-- FIX

    if (!["pdf", "docx", "html", "txt"].includes(format)) {
      throw new Error(`Unsupported format: ${format}`);
    }
    if (!fs.existsSync(inputPath)) {
      throw new Error(`Input file not found: ${inputPath}`);
    }

    const ext = path.extname(inputPath).toLowerCase();
    const baseName = outputName || path.basename(inputPath, ext);

    let outputFile;

    if (showDialog) {
      // Zeige nativen Dialog
      outputFile = await showSaveDialog(baseName, format);
      if (!outputFile) {
        console.log("Speichervorgang abgebrochen.");
        return null;
      }
    } else {
      // Nutze Standard-Ausgabeverzeichnis
      outputFile = path.join(outputDir, `${baseName}.${format.toLowerCase()}`);
    }

    let htmlContent = fs.readFileSync(inputPath, "utf8");

    // Remove <think> tags if present
    htmlContent = htmlContent.replace(/<think>[\s\S]*?<\/think>/gi, "");

    switch (format.toLowerCase()) {
      case "html":
        fs.writeFileSync(outputFile, htmlContent, "utf8");
        break;
      case "pdf":
        await this.htmlToPDF(htmlContent, outputFile);
        break;
      case "docx":
        await this.htmlToDOCX(htmlContent, outputFile);
        break;
      case "txt":
        fs.writeFileSync(outputFile, this.htmlToTXT(htmlContent), "utf8");
        break;
      default:
        throw new Error(`Unsupported format: ${format}`);
    }

    console.log(`Erfolgreich gespeichert: ${outputFile}`);
    return outputFile;
  },

  // HTML → PDF
  async htmlToPDF(html, outputPath) {
    let browser;
    try {
      browser = await puppeteer.launch({
        headless: true,
        args: ["--no-sandbox", "--disable-setuid-sandbox"],
      });

      const page = await browser.newPage();
      await page.setContent(html, { waitUntil: "networkidle0" });
      await page.pdf({
        path: outputPath,
        format: "A4",
        printBackground: true,
        margin: {
          top: "20mm",
          right: "20mm",
          bottom: "20mm",
          left: "20mm",
        },
      });
    } finally {
      if (browser) {
        await browser.close();
      }
    }
  },

  // HTML → DOCX
  async htmlToDOCX(html, outputPath) {
    try {
      // html‑to‑docx library converts HTML string into a Word .docx buffer
      // Usage from html‑to‑docx docs:
      // await HTMLtoDOCX(htmlString, headerHTMLString, documentOptions, footerHTMLString)  [oai_citation:0‡GitHub](https://github.com/privateOmega/html-to-docx?utm_source=chatgpt.com)
      const buffer = await htmlToDocx(html, null, {
        table: { row: { cantSplit: true } },
      });
      fs.writeFileSync(outputPath, buffer);
    } catch (err) {
      throw new Error(`DOCX conversion failed: ${err.message}`);
    }
  },

  // HTML → TXT
  htmlToTXT(html) {
    // A decent plain text conversion: strip tags and collapse whitespace
    // If you want more advanced extraction consider using a library like `html-to-text` or `strip-html`  [oai_citation:1‡GitHub](https://github.com/html-to-text/node-html-to-text?utm_source=chatgpt.com)
    return (
      html
        // Remove all tags
        .replace(/<[^>]+>/g, "")
        // Convert multiple whitespace into single spaces
        .replace(/\s+/g, " ")
        .trim()
    );
  },
};

module.exports = module_exports;

// CLI usage mit Dialog
if (require.main === module) {
  (async () => {
    const args = process.argv.slice(2);
    if (args.length < 1) {
      console.log("Usage: node htmlDocumentConverter.js <input.html> [format]");
      console.log("Formats: pdf (default), docx, html, txt");
      console.log("");
      console.log(
        'Ein nativer "Speichern unter" Dialog wird automatisch geöffnet.',
      );
      process.exit(1);
    }

    const inputPath = args[0];
    const format = args[1] || "pdf";

    try {
      await module_exports.convert({
        inputPath,
        format,
        showDialog: true,
      });
    } catch (err) {
      console.error("Konvertierung fehlgeschlagen:", err.message);
      process.exit(1);
    }
  })();
}