pdf-merger-js/PDFMergerBase.js

merge multiple PDF documents, or parts of them, to a new PDF document

import { PDFDocument } from 'pdf-lib'

import { parsePagesString } from './parsePagesString.js'

/**
 * @typedef {Object} Metadata
 * @property {string} [producer]
 * @property {string} [author]
 * @property {string} [title]
 * @property {string} [creator]
 */

/**
 * @typedef {Uint8Array | ArrayBuffer | Blob | URL} PdfInput
 */

export default class PDFMergerBase {
  /**
   * The internal pdf-lib document.
   *
   * @protected
   * @type {PDFDocument | undefined}
   */
  _doc = undefined

  /**
   * The load options for pdf-lib.
   *
   * @type { import('pdf-lib').LoadOptions }
   * @protected
   */
  _loadOptions = {
    // allow merging of encrypted pdfs (issue #88)
    ignoreEncryption: true
  }

  constructor () {
    this.reset()
  }

  /**
   * Resets the internal state of the document, to start again.
   *
   * @returns {void}
   */
  reset () {
    this._doc = undefined
  }

  /**
   * Set the metadata of the merged PDF.
   *
   * @async
   * @param {Metadata} metadata
   * @returns {Promise<void>}
   */
  async setMetadata (metadata) {
    await this._ensureDoc()
    if (metadata.producer) this._doc.setProducer(metadata.producer)
    if (metadata.author) this._doc.setAuthor(metadata.author)
    if (metadata.title) this._doc.setTitle(metadata.title)
    if (metadata.creator) this._doc.setCreator(metadata.creator)
  }

  /**
   * Add pages from a PDF document to the end of the merged document.
   *
   * @async
   * @param {PdfInput} input - a pdf source
   * @param {string | string[] | number | number[] | undefined | null} [pages]
   * @returns {Promise<void>}
   */
  async add (input, pages) {
    await this._ensureDoc()
    if (typeof pages === 'undefined' || pages === null || pages === 'all') {
      // of no pages are given, add the entire document
      await this._addPagesFromDocument(input)
    } else if (typeof pages === 'number') {
      // e.g. 2
      await this._addPagesFromDocument(input, [pages])
    } else if (Array.isArray(pages)) {
      // e.g. [2,3,6] or ["2","3","6"]
      const pagesAsNumbers = pages.map(p => typeof p === 'string' ? parseInt(p.trim()) : p)
      await this._addPagesFromDocument(input, pagesAsNumbers)
    } else if (typeof pages === 'string' || pages instanceof String) {
      // e.g. "2,3,6" or "2-6" or "2to6,8,10-12"
      const pagesArray = parsePagesString(pages)
      await this._addPagesFromDocument(input, pagesArray)
    } else {
      throw new Error([
        'Invalid parameter "pages".',
        'Must be a string like "1,2,3" or "1-3" or an Array of numbers.'
      ].join(' '))
    }
  }

  /**
   * Creates a new PDFDocument and sets the metadata
   * if this.#doc does not exist yet
   *
   * @protected
   * @async
   * @returns {Promise<void>}
   */
  async _ensureDoc () {
    if (!this._doc) {
      this._doc = await PDFDocument.create()
      this._doc.setProducer('pdf-merger-js')
      this._doc.setCreationDate(new Date())
    }
  }

  /**
   * Get the merged PDF as a Uint8Array.
   *
   * @async
   * @protected
   * @returns {Promise<Uint8Array>}
   */
  async _saveAsUint8Array () {
    await this._ensureDoc()
    return await this._doc.save()
  }

  /**
   * Get the merged PDF as a Base64 encoded string.
   *
   * @async
   * @protected
   * @returns {Promise<string>}
   */
  async _saveAsBase64 () {
    await this._ensureDoc()
    return await this._doc.saveAsBase64({ dataUri: true })
  }

  /**
   * Converts the input to a Uint8Array.
   * If the input is a string, it is treated as a URL and the document gets fetched.
   *
   * @async
   * @protected
   * @param {PdfInput} input
   * @returns {Uint8Array}
   */
  async _getInputAsUint8Array (input) {
    // see https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Uint8Array
    if (input instanceof Uint8Array) {
      return input
    }

    // see https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/ArrayBuffer
    if (input instanceof ArrayBuffer || Object.prototype.toString.call(input) === '[object ArrayBuffer]') {
      return new Uint8Array(input)
    }

    // see https://developer.mozilla.org/en-US/docs/Web/API/Blob
    if (typeof Blob !== 'undefined' && input instanceof Blob) {
      const aBuffer = await input.arrayBuffer()
      return new Uint8Array(aBuffer)
    }

    // see https://developer.mozilla.org/en-US/docs/Web/API/URL
    if (input instanceof URL) {
      if (typeof fetch === 'undefined') {
        throw new Error('fetch is not defined. You need to use a polyfill for this to work.')
      }
      const res = await fetch(input)
      const aBuffer = await res.arrayBuffer()
      return new Uint8Array(aBuffer)
    }

    // throw a meaningful error if input type is unknown or invalid
    const allowedTypes = ['Uint8Array', 'ArrayBuffer', 'File', 'Blob', 'URL']
    let errorMsg = `pdf-input must be of type ${allowedTypes.join(', ')}, a valid filename or url!`
    if (typeof input === 'string' || input instanceof String) {
      errorMsg += ` Input was "${input}" wich is not an existing file, nor a valid URL!`
    } else {
      errorMsg += ` Input was of type "${typeof input}" instead.`
    }
    throw new Error(errorMsg)
  }

  /**
   * @async
   * @protected
   * @param {PdfInput} input
   * @param {number[] | undefined} pages - array of page numbers to add (starts at 1)
   * @returns {Promise<void>}
   */
  async _addPagesFromDocument (input, pages = undefined) {
    const src = await this._getInputAsUint8Array(input)
    const srcDoc = await PDFDocument.load(src, this._loadOptions)

    let indices = []
    if (pages === undefined) {
      // add the whole document
      indices = srcDoc.getPageIndices()
    } else {
      // add selected pages switching to a 0-based index
      indices = pages.map(p => p - 1)
    }

    const copiedPages = await this._doc.copyPages(srcDoc, indices)
    copiedPages.forEach((page) => {
      this._doc.addPage(page)
    })
  }
}