smartpdf/ts/smartpdf.classes.smartpdf.ts

import * as plugins from './smartpdf.plugins.js';
import * as paths from './smartpdf.paths.js';
import { Server } from 'http';
import { PdfCandidate } from './smartpdf.classes.pdfcandidate.js';
import { type IPdf } from '@tsclass/tsclass/dist_ts/business/pdf.js';

declare const document: any;

export class SmartPdf {
  // STATIC
  public static async create() {
    const smartpdfInstance = new SmartPdf();
    return smartpdfInstance;
  }

  // INSTANCE
  htmlServerInstance: Server;
  serverPort: number;
  headlessBrowser: plugins.smartpuppeteer.puppeteer.Browser;
  externalBrowserBool: boolean = false;
  private _readyDeferred: plugins.smartpromise.Deferred<void>;
  private _candidates: { [key: string]: PdfCandidate } = {};

  constructor() {
    this._readyDeferred = new plugins.smartpromise.Deferred();
  }

  async start(headlessBrowserArg?: plugins.smartpuppeteer.puppeteer.Browser) {
    const done = plugins.smartpromise.defer();
    // lets set the external browser in case one is provided
    this.headlessBrowser = headlessBrowserArg;
    // setup puppeteer
    if (this.headlessBrowser) {
      this.externalBrowserBool = true;
    } else {
      this.headlessBrowser = await plugins.smartpuppeteer.getEnvAwareBrowserInstance({
        forceNoSandbox: true,
      });
    }

    // setup server
    const app = plugins.express();
    app.get('/:pdfId', (req, res) => {
      res.setHeader('PDF-ID', this._candidates[req.params.pdfId].pdfId);
      res.send(this._candidates[req.params.pdfId].htmlString);
    });
    this.htmlServerInstance = plugins.http.createServer(app);
    const smartnetworkInstance = new plugins.smartnetwork.SmartNetwork();
    const portAvailable = smartnetworkInstance.isLocalPortUnused(3210);
    this.htmlServerInstance.listen(3210, 'localhost');
    this.htmlServerInstance.on('listening', () => {
      this._readyDeferred.resolve();
      done.resolve();
    });
    await done.promise;
  }

  // stop
  async stop() {
    const done = plugins.smartpromise.defer<void>();
    this.htmlServerInstance.close(() => {
      done.resolve();
    });

    if (!this.externalBrowserBool) {
      await this.headlessBrowser.close();
    }

    await done.promise;
  }

  /**
   * returns a pdf for a given html string;
   */
  async getA4PdfResultForHtmlString(htmlStringArg: string): Promise<plugins.tsclass.business.IPdf> {
    await this._readyDeferred.promise;
    const pdfCandidate = new PdfCandidate(htmlStringArg);
    this._candidates[pdfCandidate.pdfId] = pdfCandidate;
    const page = await this.headlessBrowser.newPage();
    await page.setViewport({
      width: 794,
      height: 1122,
    });
    const response = await page.goto(`http://localhost:3210/${pdfCandidate.pdfId}`, {
      waitUntil: 'networkidle2',
    });
    // await plugins.smartdelay.delayFor(1000);
    const headers = response.headers();
    if (headers['pdf-id'] !== pdfCandidate.pdfId) {
      console.log('Error! Headers do not match. For security reasons no pdf is being emitted!');
      return;
    } else {
      console.log(`id security check passed for ${pdfCandidate.pdfId}`);
    }

    const pdfBuffer = await page.pdf({
      width: 794,
      height: 1122,
      printBackground: true,
      displayHeaderFooter: false,
    });
    await page.close();
    delete this._candidates[pdfCandidate.pdfId];
    pdfCandidate.doneDeferred.resolve();
    await pdfCandidate.doneDeferred.promise;
    return {
      id: pdfCandidate.pdfId,
      name: `${pdfCandidate.pdfId}.js`,
      metadata: {
        textExtraction: await this.extractTextFromPdfBuffer(pdfBuffer),
      },
      buffer: pdfBuffer,
    };
  }

  async getPdfResultForWebsite(websiteUrl: string): Promise<plugins.tsclass.business.IPdf> {
    const page = await this.headlessBrowser.newPage();
    await page.setViewport({
      width: 1980,
      height: 1200,
    });
    await page.emulateMediaType('screen');
    const response = await page.goto(websiteUrl, { waitUntil: 'networkidle2' });
    const pdfId = plugins.smartunique.shortId();
    const { documentHeight, documentWidth } = await page.evaluate(() => {
      return {
        documentHeight: document.height,
        documentWidth: document.width,
      };
    });
    const pdfBuffer = await page.pdf({
      height: documentHeight,
      width: documentWidth,
      printBackground: true,
      displayHeaderFooter: false,
    });
    await page.close();
    return {
      id: pdfId,
      name: `${pdfId}.js`,
      metadata: {
        textExtraction: await this.extractTextFromPdfBuffer(pdfBuffer),
      },
      buffer: pdfBuffer,
    };
  }

  async getFullWebsiteAsSinglePdf(websiteUrl: string): Promise<plugins.tsclass.business.IPdf> {
    const page = await this.headlessBrowser.newPage();
    await page.setViewport({
      width: 1920,
      height: 1200,
    });
    page.emulateMediaType('screen');
    const response = await page.goto(websiteUrl, { waitUntil: 'networkidle2' });
    const pdfId = plugins.smartunique.shortId();
    const { documentHeight, documentWidth } = await page.evaluate(() => {
      return {
        documentHeight: document.body.scrollHeight,
        documentWidth: document.body.clientWidth,
      };
    });
    await page.setViewport({
      width: 1920,
      height: documentHeight,
    });
    const pdfBuffer = await page.pdf({
      height: documentHeight,
      width: 1920,
      printBackground: true,
      displayHeaderFooter: false,
      scale: 1,
      pageRanges: '1',
    });
    await page.close();
    return {
      id: pdfId,
      name: `${pdfId}.js`,
      metadata: {
        textExtraction: await this.extractTextFromPdfBuffer(pdfBuffer),
      },
      buffer: pdfBuffer,
    };
  }

  public async mergePdfs(inputPdfBuffers: Uint8Array[]): Promise<Uint8Array> {
    const mergedPdf = await plugins.pdfLib.PDFDocument.create();
    for (const pdfBytes of inputPdfBuffers) {
      const pdfDoc = await plugins.pdfLib.PDFDocument.load(pdfBytes);
      const pages = await mergedPdf.copyPages(pdfDoc, pdfDoc.getPageIndices());
      pages.forEach((page) => mergedPdf.addPage(page));
    }

    const mergedPdfBytes = await mergedPdf.save();
    return mergedPdfBytes;
  }

  public async readFileToPdfObject(pathArg: string): Promise<plugins.tsclass.business.IPdf> {
    const path = plugins.smartpath.transform.makeAbsolute(pathArg);
    const parsedPath = plugins.path.parse(path);
    const buffer = await plugins.smartfile.fs.toBuffer(path);
    return {
      name: parsedPath.base,
      buffer,
      id: null,
      metadata: null,
    };
  }

  public async extractTextFromPdfBuffer(pdfBufferArg: Buffer): Promise<string> {
    const deferred = plugins.smartpromise.defer<string>();
    const pdfParser: any = new plugins.pdf2json();
    pdfParser.on('pdfParser_dataReady', (pdfData: any) => {
      let finalText = '';
      for (const page of pdfData.Pages) {
        for (const text of page.Texts) {
          for (const letter of text.R) {
            finalText = finalText + letter.T;
          }
        }
      }
      deferred.resolve(finalText);
    });
    pdfParser.parseBuffer(pdfBufferArg);
    return deferred.promise;
  }

  public async convertPDFToJPGBytes(
    pdfBytes: Uint8Array,
    options: {
      width?: number;
      height?: number;
      quality?: number;
    } = {}
  ): Promise<Uint8Array[]> {
    const { width = 1024, height = 768, quality = 100 } = options;

    // Load the PDF document
    const pdfDoc = await plugins.pdfLib.PDFDocument.load(pdfBytes);

    const converter = plugins.pdf2pic.fromBuffer(Buffer.from(pdfBytes), {
      density: 100, // Image density (DPI)
      format: 'jpg', // Image format
      width, // Output image width
      height, // Output image height
      quality, // Output image quality
    });

    // Get array promises that resolve to JPG buffers
    const imagePromises: Promise<Buffer>[] = [];
    const numPages = pdfDoc.getPageCount();

    for (let i = 0; i < numPages; i++) {
      imagePromises.push(converter(i + 1, {
        responseType: 'buffer',
      }).then((output) => output.buffer));
    }

    // Resolve all promises and return the array of buffers
    return Promise.all(imagePromises);
  }
}