# @cantoo/pdf-lib

> Create and modify PDF files with JavaScript

Latest version **2.11.1** (published 2026-09-15) · MIT license · 297.9K weekly downloads

## Install

```sh
npm install @cantoo/pdf-lib
pnpm add @cantoo/pdf-lib
yarn add @cantoo/pdf-lib
bun add @cantoo/pdf-lib
```

## Health

**Score 85/100 (A)** — status: active.

Positive: moderate downloads; has types; esm support; no vulnerabilities; recently updated; high maintenance score; high quality score.

Warnings: large bundle.

## Facts

| | |
|---|---|
| Version | 2.11.1 |
| Published | 2026-09-15 |
| First published | 2021-03-12 |
| Weekly downloads | 297.9K |
| License | MIT |
| TypeScript types | bundled |
| Module format | ESM + CommonJS |
| Dependencies | 4 |
| Unpacked size | 25.1 MB |
| Known vulnerabilities | 0 |
| Install scripts | no |
| GitHub stars | 349 |
| Author | Andrew Dillon |
| Maintainers | sharcoux, ribeirofmatheus, matheusrdsantos, gabrielnsd, felipecantoo, nicodemos234 |
| Keywords | pdf-lib, pdf, document, create, modify, creation, modification, edit, editing, typescript, javascript, library |

## Links

- npm: https://www.npmjs.com/package/@cantoo/pdf-lib
- Repository: https://github.com/cantoo-scribe/pdf-lib
- Homepage: https://pdf-lib.js.org
- Issues: https://github.com/cantoo-scribe/pdf-lib/issues
- npm.io page: https://npm.io/package/@cantoo/pdf-lib

## Dependencies (4)

- [tslib](https://npm.io/package/tslib.md) >=2
- [culori](https://npm.io/package/culori.md) >=4
- [fflate](https://npm.io/package/fflate.md) ^0.8.2
- [node-html-better-parser](https://npm.io/package/node-html-better-parser.md) >=1.5.9

## Alternatives

- [datatables.net-buttons](https://npm.io/package/datatables.net-buttons.md) — 200.1K weekly downloads
- [@ckeditor/ckeditor5-export-pdf](https://npm.io/package/@ckeditor/ckeditor5-export-pdf.md) — 167.0K weekly downloads
- [scanbot-web-sdk](https://npm.io/package/scanbot-web-sdk.md) — 15.0K weekly downloads
- [@syncfusion/ej2-angular-pdfviewer](https://npm.io/package/@syncfusion/ej2-angular-pdfviewer.md) — 8.8K weekly downloads
- [@regulaforensics/document-reader-webclient](https://npm.io/package/@regulaforensics/document-reader-webclient.md) — 7.8K weekly downloads

## Recent versions

- 2.11.1 (latest) — 2026-09-15
- 2.2.3-0 (next) — 2024-08-10
- 2.11.0 — 2026-09-11
- 2.9.2 — 2026-09-07
- 2.9.1 — 2026-08-18
- 2.9.0 — 2026-08-18
- 2.8.4 — 2026-08-16
- 2.8.3 — 2026-08-14
- 2.8.2 — 2026-08-12
- 2.8.1 — 2026-07-30
- 2.8.0 — 2026-07-30
- 2.7.4 — 2026-07-16
- 2.7.3 — 2026-07-14
- 2.7.2 — 2026-07-12
- 2.7.1 — 2026-05-27
- … 107 more at https://npm.io/package/@cantoo/pdf-lib/versions

## README

## Forked by [Cantoo](https://cantoo.fr)

[`@cantoo/pdf-lib`](https://www.npmjs.com/package/@cantoo/pdf-lib) is an actively maintained fork of [`Hopding/pdf-lib`](https://github.com/Hopding/pdf-lib). We keep the original API and docs, and ship the features below that upstream does not (or no longer) provide. Install with:

```bash
npm install @cantoo/pdf-lib
# or
yarn add @cantoo/pdf-lib
```

We maintain this project for our own product needs and cannot guarantee support for every issue outside that roadmap — but contributions that fit the library are welcome.

### Benefits vs original `pdf-lib`

**Drawing & content**

- **Full SVG drawing** (`page.drawSvg` / `embedSvg`) — not just `drawSvgPath`: shapes, text, transforms, viewBox cropping, blend modes, rounded rects, CSS colors via [`culori`](https://culorijs.org/)
- **`PDFPage.extractContents()`** — pull typed assets (text, images, approximate vector graphics as SVG) out of page content streams
- **`characterSpacing`** on `drawText`

**Documents & save**

- **Incremental updates** (`forIncrementalUpdate`, `save({ useObjectStreams… })`, `commit()`) — append changes without rewriting the whole file (signing / multi-increment workflows)
- **PDF header preserved** on save (writers no longer always force `%PDF-1.7`)
- **Recover truncated PDFs** instead of failing hard on a broken trailer object
- More robust parsing (hex-string whitespace, object streams on older PDFs, HTML closing tags spanning lines, case-insensitive PDF names, …)

**Forms & scripts**

- **XFA helpers** — `preserveXFA`, read/modify XFA JavaScript, signature-field discovery (static / government-style XFA)
- Extract / work with **document-level JavaScript**
- **`PDFForm.flatten()`** also flattens orphaned widget annotations not listed in `AcroForm.Fields`
- **Attachments** — add, extract (including unsaved), and detach embedded files; name trees kept in PDF lexical order

**Standards & e-invoicing**

- **`convertToPDFA()`** — PDF/A-1/2/3 structural pieces (OutputIntent + sRGB ICC, `/ID`, XMP synced with Info)
- **`embedFacturX()`** — Factur-X / ZUGFeRD PDF/A-3 hybrid invoices

**Security & crypto**

- **`PDFDocument.encrypt()`** — AES-256 revision 6 by default; AES-128 / RC4 via `algorithm` (RC4 needs `allowWeakCryptography`)
- Decrypt PDFs whose crypt filters are **indirect objects**
- No `crypto-js`: in-tree AES/RC4/MD5/SHA-256 + Web Crypto `getRandomValues`

**Fonts & layers**

- First-class **[`@cantoo/fontkit`](https://www.npmjs.com/package/@cantoo/fontkit)** support (types, default export, UMD); still compatible with upstream `fontkit` / `@pdf-lib/fontkit`
- **`postscriptName`** to pick a face from `.ttc` / `.dfont` collections
- **Optional content (PDF layers)** — `getOptionalContentGroups()` / `setOptionalContentGroupVisibility()`

**Packaging & deps**

- Vendored **[`@cantoo/upng`](./src/packages/upng)** and **[`@cantoo/standard-fonts`](./src/packages/standard-fonts)** (also as `@cantoo/pdf-lib/upng` and `@cantoo/pdf-lib/standard-fonts`)
- Compression via **[`fflate`](https://www.npmjs.com/package/fflate)** instead of `pako` (modular TS UPNG with Photopea `quantize` / `framize` / lossy `cnum`)
- Lighter color parsing with **`culori`**; `html-entities` as a peer (pulled in by the HTML/SVG parser)

See [CHANGELOG.md](CHANGELOG.md) for version-by-version detail.

<hr/>

<a href="https://pdf-lib.js.org">
<h1 align="center">
<img alt="pdf-lib" height="300" src="https://raw.githubusercontent.com/Hopding/pdf-lib-docs/master/assets/logo-full.svg?sanitize=true">
</h1>
</a>

<div align="center">
  <strong>Create and modify PDF documents in any JavaScript environment.</strong>
</div>
<div align="center">
  Designed to work in any modern JavaScript runtime. Tested in Node, Browser, Deno, and React Native environments.
</div>

<br />

<div align="center">
  <!-- NPM Version -->
  <a href="https://www.npmjs.com/package/pdf-lib">
    <img
      src="https://img.shields.io/npm/v/pdf-lib.svg?style=flat-square"
      alt="NPM Version"
    />
  </a>
  <!-- Build Status -->
  <a href="https://circleci.com/gh/Hopding/pdf-lib">
    <img
      src="https://img.shields.io/circleci/project/github/Hopding/pdf-lib/master.svg?style=flat-square&label=CircleCI"
      alt="CircleCI Build Status"
    />
  </a>
  <!-- Prettier Badge -->
  <a href="https://prettier.io/">
    <img
      src="https://img.shields.io/badge/code_style-prettier-ff69b4.svg?style=flat-square"
      alt="Prettier Badge"
    />
  </a>
  <!-- Discord Badge -->
  <a href="https://discord.gg/Y7uuVMc">
    <img
      src="https://img.shields.io/static/v1?label=discord&message=pdf-lib&color=566fbb&style=flat-square"
      alt="Discord Badge"
    />
  </a>
</div>

<br />

> **Learn more at [pdf-lib.js.org](https://pdf-lib.js.org)**

## Table of Contents

- [Benefits vs original `pdf-lib`](#benefits-vs-original-pdf-lib)
- [Features](#features)
- [Motivation](#motivation)
- [Usage Examples](#usage-examples)
  - [Create Document](#create-document)
  - [Modify Document](#modify-document)
  - [Incremental Document Modification](#incremental-document-modification)
  - [Consecutive Incremental Updates](#consecutive-incremental-updates)
  - [Create Form](#create-form)
  - [Fill Form](#fill-form)
  - [Flatten Form](#flatten-form)
  - [Work with XFA Forms](#work-with-xfa-forms)
  - [Extract XFA JavaScript](#extract-xfa-javascript)
  - [Modify XFA JavaScript](#modify-xfa-javascript)
  - [Extract Document JavaScript](#extract-document-javascript)
  - [Copy Pages](#copy-pages)
  - [Embed PNG and JPEG Images](#embed-png-and-jpeg-images)
  - [Embed PDF Pages](#embed-pdf-pages)
  - [Embed Font and Measure Text](#embed-font-and-measure-text)
  - [Add Attachments](#add-attachments)
  - [Extract Attachments](#extract-attachments)
  - [Create PDF/A Documents](#create-pdfa-documents)
  - [Embed Factur-X / ZUGFeRD Invoices](#embed-factur-x--zugferd-invoices)
  - [Set Document Metadata](#set-document-metadata)
  - [Read Document Metadata](#read-document-metadata)
  - [Set Viewer Preferences](#set-viewer-preferences)
  - [Read Viewer Preferences](#read-viewer-preferences)
  - [Draw SVG Paths](#draw-svg-paths)
  - [Draw SVG](#draw-svg)
- [Deno Usage](#deno-usage)
- [Complete Examples](#complete-examples)
- [Installation](#installation)
  - [NPM Module](#npm-module)
  - [Vendored PNG & standard fonts](#vendored-png--standard-fonts)
  - [UMD Module](#umd-module)
  - [Fontkit Installation](#fontkit-installation)
- [Documentation](#documentation)
- [Fonts and Unicode](#fonts-and-unicode)
- [Creating and Filling Forms](#creating-and-filling-forms)
- [Limitations](#limitations)
- [Help and Discussion](#help-and-discussion)
- [Encryption Handling](#encryption-handling)
- [Migrating to v1.0.0](docs/MIGRATION.md)
- [Contributing](#contributing)
- [Maintainership](#maintainership)
- [Tutorials and Cool Stuff](#tutorials-and-cool-stuff)
- [Prior Art](#prior-art)
- [Git History Rewrite](#git-history-rewrite)
- [Changelog](CHANGELOG.md)
- [License](#license)

## Features

- Create new PDFs
- Modify existing PDFs
- Incremental document updates (append-only saves / multi-commit)
- Encrypt / decrypt PDFs (AES-256 by default)
- Create forms
- Fill forms
- Flatten forms
- Preserve XFA forms
- Extract XFA JavaScript
- Modify XFA JavaScript
- Extract document-level JavaScript
- Add Pages
- Insert Pages
- Remove Pages
- Copy pages between PDFs
- Draw Text (including character spacing)
- Draw Images
- Draw PDF Pages
- Draw Vector Graphics
- Draw SVG Paths
- Draw full SVG documents (`drawSvg` / `embedSvg`)
- Extract page contents (text, images, vector graphics)
- Optional content / PDF layer visibility
- Measure width and height of text
- Embed Fonts (supports UTF-8 and UTF-16 character sets; TTC/DFont face selection)
- Set document metadata
- Read document metadata
- Set viewer preferences
- Read viewer preferences
- Add attachments
- Extract / detach attachments
- Create PDF/A documents (parts 1–3)
- Embed Factur-X / ZUGFeRD e-invoices (PDF/A-3)

## Motivation

`pdf-lib` was created to address the JavaScript ecosystem's lack of robust support for PDF manipulation (especially for PDF _modification_).

Two of `pdf-lib`'s distinguishing features are:

1. Supporting modification (editing) of existing documents.
2. Working in all JavaScript environments - not just in Node or the Browser.

There are [other](#prior-art) good open source JavaScript PDF libraries available. However, most of them can only _create_ documents, they cannot _modify_ existing ones. And many of them only work in particular environments.

## Usage Examples

### Create Document

_This example produces [this PDF](assets/pdfs/examples/create_document.pdf)._

[Try the JSFiddle demo](https://jsfiddle.net/Hopding/rxwsc8f5/13/)

<!-- prettier-ignore -->
```js
import { PDFDocument, StandardFonts, rgb } from 'pdf-lib'

// Create a new PDFDocument
const pdfDoc = await PDFDocument.create()

// Embed the Times Roman font
const timesRomanFont = await pdfDoc.embedFont(StandardFonts.TimesRoman)

// Add a blank page to the document
const page = pdfDoc.addPage()

// Get the width and height of the page
const { width, height } = page.getSize()

// Draw a string of text toward the top of the page
const fontSize = 30
page.drawText('Creating PDFs in JavaScript is awesome!', {
  x: 50,
  y: height - 4 * fontSize,
  size: fontSize,
  font: timesRomanFont,
  color: rgb(0, 0.53, 0.71),
})

// Serialize the PDFDocument to bytes (a Uint8Array)
const pdfBytes = await pdfDoc.save()

// For example, `pdfBytes` can be:
//   • Written to a file in Node
//   • Downloaded from the browser
//   • Rendered in an <iframe>
```

### Modify Document

_This example produces [this PDF](assets/pdfs/examples/modify_document.pdf)_ (when [this PDF](assets/pdfs/with_update_sections.pdf) is used for the `existingPdfBytes` variable).

[Try the JSFiddle demo](https://jsfiddle.net/Hopding/64zajhge/1/)

<!-- prettier-ignore -->
```js
import { degrees, PDFDocument, rgb, StandardFonts } from 'pdf-lib';

// This should be a Uint8Array or ArrayBuffer
// This data can be obtained in a number of different ways
// If your running in a Node environment, you could use fs.readFile()
// In the browser, you could make a fetch() call and use res.arrayBuffer()
const existingPdfBytes = ...

// Load a PDFDocument from the existing PDF bytes
const pdfDoc = await PDFDocument.load(existingPdfBytes)

// Embed the Helvetica font
const helveticaFont = await pdfDoc.embedFont(StandardFonts.Helvetica)

// Get the first page of the document
const pages = pdfDoc.getPages()
const firstPage = pages[0]

// Get the width and height of the first page
const { width, height } = firstPage.getSize()

// Draw a string of text diagonally across the first page
firstPage.drawText('This text was added with JavaScript!', {
  x: 5,
  y: height / 2 + 300,
  size: 50,
  font: helveticaFont,
  color: rgb(0.95, 0.1, 0.1),
  rotate: degrees(-45),
})


// Serialize the PDFDocument to bytes (a Uint8Array)
const pdfBytes = await pdfDoc.save()

// For example, `pdfBytes` can be:
//   • Written to a file in Node
//   • Downloaded from the browser
//   • Rendered in an <iframe>
```

### Incremental Document Modification

You can load a PDF for incremental update, generating the original document plus the increment, on save. You can also handle incremental update manually.
The incremental modification saving is designed to be used for pdf signing. The signature is added to an existing page, then only the 'incremental' PDF is generated and concatenated to the initial version.

_This example produces [this PDF](assets/pdfs/examples/incremental_document_modification.pdf)_ (when [this PDF](assets/pdfs/simple.pdf) is used for the `existingPdfBytes` variable).

<!-- prettier-ignore -->
```js
import { PDFDocument, StandardFonts } from 'pdf-lib';

// This should be a Uint8Array or ArrayBuffer
// This data can be obtained in a number of different ways
// If your running in a Node environment, you could use fs.readFile()
// In the browser, you could make a fetch() call and use res.arrayBuffer()
const existingPdfBytes = ...

// Load a PDFDocument from the existing PDF bytes
const pdfDoc = await PDFDocument.load(existingPdfBytes)

// Take a snapshot of the document
const snapshot = pdfDoc.takeSnapshot();

// Get the first page of the document
const pages = pdfDoc.getPages()
const firstPage = pages[0]

// Mark the page as modified
snapshot.markRefForSave(firstPage.ref)

// Draw a string of text diagonally across the first page
firstPage.drawText('Incremental saving is also awesome!', {
  x: 50,
  y: 4 * fontSize,
  size: fontSize
})

// Serialize the PDFDocument to bytes (a Uint8Array)
const pdfIncrementalBytes = await pdfDoc.saveIncremental(snapshot)
const pdfBytes = Buffer.concatenate([ existingPdfBytes, pdfIncrementalBytes ])

// For example, `pdfBytes` can be:
//   • Written to a file in Node
//   • Downloaded from the browser
//   • Rendered in an <iframe>
```

Loading an existing PDF forIncrementalUpdate, makes things easier:

<!-- prettier-ignore -->
```js
import { PDFDocument, StandardFonts } from 'pdf-lib';

// This should be a Uint8Array or ArrayBuffer
const existingPdfBytes = ...

// Load a PDFDocument from the existing PDF bytes, for incremental update
const pdfDoc = await PDFDocument.load(existingPdfBytes,{forIncrementalUpdate:true})

// Get the first page of the document
const pages = pdfDoc.getPages()
const firstPage = pages[0]

// Draw a string of text diagonally across the first page
firstPage.drawText('Incremental saving is also awesome!', {
  x: 50,
  y: 4 * fontSize,
  size: fontSize
})

// Serialize the PDFDocument to bytes (a Uint8Array), using incremental updates
const pdfBytes = await pdfDoc.save()

// For example, `pdfBytes` can be:
//   • Written to a file in Node
//   • Downloaded from the browser
//   • Rendered in an <iframe>
```

You can force a rewrite of a PDF that was open for incremental update with the right parameter on save:

<!-- prettier-ignore -->
```js
import { PDFDocument } from 'pdf-lib';

// This should be a Uint8Array or ArrayBuffer
const existingPdfBytes = ...

// Load a PDFDocument from the existing PDF bytes, for incremental update
const pdfDoc = await PDFDocument.load(existingPdfBytes,{forIncrementalUpdate:true})

// Do something..

// Serialize the PDFDocument to bytes (a Uint8Array), NOT using incremental updates
const pdfBytes = await pdfDoc.save({rewrite: true})
```

#### Using pdf-lib to generate a placeholder for an electronic signature

@signpdf includes a pdf-lib-placeholder component, but it is based on the pdf-lib package that has no incremental update functionality. This code is taken from that library and modified to use incremental update for not invalidating previous file signatures. Is an example, that can be seen in integration test #20, some @signpdf constants has been changed to arbitrary values for the example to work "out of the box".

<!-- prettier-ignore -->
```js
import { PDFDocument, StandardFonts, rgb, PDFArray, PDFNumber, PDFName, PDFHexString, PDFString, PDFInvalidObject } from 'pdf-lib';
const pdfBytes = ...

const pdfDoc = await PDFDocument.load(pdfBytes, {forIncrementalUpdate: true});
// visual representation of the signature
const page = pdfDoc.addPage([500, 200]);
const font = pdfDoc.embedStandardFont(StandardFonts.Helvetica);
page.drawRectangle({
  x: 10,
  y: 30,
  width: 280,
  height: 150,
  borderWidth: 2,
  borderColor: rgb(0.45, 0.45, 0.45),
});
page.drawText(`Electronic Signature Example\nSigned on ${(new Date()).toIsoString()}\nThis is not the real signature!!`, {
  x: 20,
  y: 200,
  size: 14,
  font,
});
// Add an AcroForm or update the existing one
let acroForm = pdfDoc.catalog.getOrCreateAcroForm();
// Create a placeholder where the the last 3 parameters of the
// actual range will be replaced when signing is done.
const byteRange = PDFArray.withContext(pdfDoc.context);
byteRange.push(PDFNumber.of(0));
byteRange.push(PDFName.of('*********'));
byteRange.push(PDFName.of('*********'));
byteRange.push(PDFName.of('*********'));
// Fill the contents of the placeholder with 00s.
const placeholder = PDFHexString.of(String.fromCharCode(0).repeat(8096));
// Create a signature dictionary to be referenced in the signature widget.
const appBuild = { App: { Name: 'Signature Example pdf-lib' } };
const signatureDict = pdfDoc.context.obj({
    Type: 'Sig',
    Filter: 'Adobe.PPKLite',
    SubFilter: 'adbe.pkcs7.detached',
    ByteRange: byteRange,
    Contents: placeholder,
    Reason: PDFString.of('Example Signature'),
    M: PDFString.fromDate(new Date()),
    ContactInfo: PDFString.of('me@pdf-lib.org'),
    Name: PDFString.of('Example Signer'),
    Location: PDFString.of('In a far away galaxy..'),
    Prop_Build: {
      Filter: { Name: 'Adobe.PPKLite' },
      ...appBuild,
    },
});
// Register signatureDict as a PDFInvalidObject to prevent PDFLib from serializing it
// in an object stream.
const signatureBuffer = new Uint8Array(signatureDict.sizeInBytes());
signatureDict.copyBytesInto(signatureBuffer, 0);
const signatureObj = PDFInvalidObject.of(signatureBuffer);
const signatureDictRef = pdfDoc.context.register(signatureObj);
// Create the signature widget
const widgetRect = [0, 0, 0, 0];
const rect = PDFArray.withContext(pdfDoc.context);
widgetRect.forEach((c) => rect.push(PDFNumber.of(c)));
const apStream = pdfDoc.context.formXObject([], {
  BBox: widgetRect,
  Resources: {}, // Necessary to avoid Acrobat bug (see https://stackoverflow.com/a/73011571)
});
const widgetDict = pdfDoc.context.obj({
  Type: 'Annot',
  Subtype: 'Widget',
  FT: 'Sig',
  Rect: rect,
  V: signatureDictRef,
  T: PDFString.of('TestSig'),
  TU: PDFString.of('Electronic Signature Example'),
  F: 2,
  P: page.ref,
  AP: { N: pdfDoc.context.register(apStream) }, // Required for PDF/A compliance
});
const widgetDictRef = pdfDoc.context.register(widgetDict);
// Annotate the widget on the given page
let annotations = page.node.lookupMaybe(PDFName.of('Annots'), PDFArray);
if (typeof annotations === 'undefined') {
  annotations = pdfDoc.context.obj([]);
}
annotations.push(widgetDictRef);
page.node.set(PDFName.of('Annots'), annotations);
let sigFlags: PDFNumber;
if (acroForm.dict.has(PDFName.of('SigFlags'))) {
  // Already has some flags, will merge
  sigFlags = acroForm.dict.get(PDFName.of('SigFlags')) as PDFNumber;
} else {
  // Create blank flags
  sigFlags = PDFNumber.of(0);
}
const updatedFlags = PDFNumber.of(sigFlags!.asNumber() | 1 | 2);
acroForm.dict.set(PDFName.of('SigFlags'), updatedFlags);
let fields = acroForm.dict.get(PDFName.of('Fields'));
if (!(fields instanceof PDFArray)) {
  fields = pdfDoc.context.obj([]);
  acroForm.dict.set(PDFName.of('Fields'), fields);
}
(fields as PDFArray).push(widgetDictRef);
// Serialize the PDFDocument to bytes (a Uint8Array), using incremental updates
const pdfBytes = await pdfDoc.save()

// `pdfBytes` should be handled to the signing library to calculate the
//  file hash and fill in the generated placeholder for the signature
```

### Consecutive Incremental Updates

You can load a PDF for incremental update, and then generate multiple increments, over the original document, with commit() method.  
This method simplifies the sequence:

<!-- prettier-ignore -->
```js
import { PDFDocument, StandardFonts } from 'pdf-lib';

// This should be a Uint8Array or ArrayBuffer
const existingPdfBytes = ...
// Load a PDFDocument from the existing PDF bytes, for incremental update
const pdfDoc = await PDFDocument.load(existingPdfBytes,{forIncrementalUpdate:true})
// modify pdf
...
// Serialize the PDFDocument to bytes (a Uint8Array), using incremental updates
const firstUpdatedDoc = await pdfDoc.save()
const pdfDoc2 = await PDFDocument.load(firstUpdatedDoc,{forIncrementalUpdate:true})
// modify pdf
...
// Serialize the PDFDocument to bytes (a Uint8Array), using incremental updates
const secondUpdateDoc = await pdfDoc2.save()
// etc, etc
```

Allowing this:

<!-- prettier-ignore -->
```js
import { PDFDocument, StandardFonts } from 'pdf-lib';

// This should be a Uint8Array or ArrayBuffer
const existingPdfBytes = ...
// Load a PDFDocument from the existing PDF bytes, for incremental update
const pdfDoc = await PDFDocument.load(existingPdfBytes,{forIncrementalUpdate:true})
// modify pdf
...
// Serialize the PDFDocument to bytes (a Uint8Array), using incremental updates
const firstUpdatedDoc = await pdfDoc.commit();
// modify pdf
...
const secondUpdateDoc = await pdfDoc.commit();
// etc, etc
```

The _commit_ method has the same parameters than _save_ method. If document is not loaded **forIncrementalUpdate**, an exception is raised. After calling _commit_ an update section is added to the original document, and this replaces the original document, and a new snapshot is taken.

### Create Form

_This example produces [this PDF](assets/pdfs/examples/create_form.pdf)._

[Try the JSFiddle demo](https://jsfiddle.net/Hopding/bct7vngL/4/)

> See also [Creating and Filling Forms](#creating-and-filling-forms)

<!-- prettier-ignore -->
```js
import { PDFDocument } from 'pdf-lib'

// Create a new PDFDocument
const pdfDoc = await PDFDocument.create()

// Add a blank page to the document
const page = pdfDoc.addPage([550, 750])

// Get the form so we can add fields to it
const form = pdfDoc.getForm()

// Add the superhero text field and description
page.drawText('Enter your favorite superhero:', { x: 50, y: 700, size: 20 })

const superheroField = form.createTextField('favorite.superhero')
superheroField.setText('One Punch Man')
superheroField.addToPage(page, { x: 55, y: 640 })

// Add the rocket radio group, labels, and description
page.drawText('Select your favorite rocket:', { x: 50, y: 600, size: 20 })

page.drawText('Falcon Heavy', { x: 120, y: 560, size: 18 })
page.drawText('Saturn IV', { x: 120, y: 500, size: 18 })
page.drawText('Delta IV Heavy', { x: 340, y: 560, size: 18 })
page.drawText('Space Launch System', { x: 340, y: 500, size: 18 })

const rocketField = form.createRadioGroup('favorite.rocket')
rocketField.addOptionToPage('Falcon Heavy', page, { x: 55, y: 540 })
rocketField.addOptionToPage('Saturn IV', page, { x: 55, y: 480 })
rocketField.addOptionToPage('Delta IV Heavy', page, { x: 275, y: 540 })
rocketField.addOptionToPage('Space Launch System', page, { x: 275, y: 480 })
rocketField.select('Saturn IV')

// Add the gundam check boxes, labels, and description
page.drawText('Select your favorite gundams:', { x: 50, y: 440, size: 20 })

page.drawText('Exia', { x: 120, y: 400, size: 18 })
page.drawText('Kyrios', { x: 120, y: 340, size: 18 })
page.drawText('Virtue', { x: 340, y: 400, size: 18 })
page.drawText('Dynames', { x: 340, y: 340, size: 18 })

const exiaField = form.createCheckBox('gundam.exia')
const kyriosField = form.createCheckBox('gundam.kyrios')
const virtueField = form.createCheckBox('gundam.virtue')
const dynamesField = form.createCheckBox('gundam.dynames')

exiaField.addToPage(page, { x: 55, y: 380 })
kyriosField.addToPage(page, { x: 55, y: 320 })
virtueField.addToPage(page, { x: 275, y: 380 })
dynamesField.addToPage(page, { x: 275, y: 320 })

exiaField.check()
dynamesField.check()

// Add the planet dropdown and description
page.drawText('Select your favorite planet*:', { x: 50, y: 280, size: 20 })

const planetsField = form.createDropdown('favorite.planet')
planetsField.addOptions(['Venus', 'Earth', 'Mars', 'Pluto'])
planetsField.select('Pluto')
planetsField.addToPage(page, { x: 55, y: 220 })

// Add the person option list and description
page.drawText('Select your favorite person:', { x: 50, y: 180, size: 18 })

const personField = form.createOptionList('favorite.person')
personField.addOptions([
  'Julius Caesar',
  'Ada Lovelace',
  'Cleopatra',
  'Aaron Burr',
  'Mark Antony',
])
personField.select('Ada Lovelace')
personField.addToPage(page, { x: 55, y: 70 })

// Just saying...
page.drawText(`* Pluto should be a planet too!`, { x: 15, y: 15, size: 15 })

// Serialize the PDFDocument to bytes (a Uint8Array)
const pdfBytes = await pdfDoc.save()

// For example, `pdfBytes` can be:
//   • Written to a file in Node
//   • Downloaded from the browser
//   • Rendered in an <iframe>
```

### Fill Form

_This example produces [this PDF](assets/pdfs/examples/fill_form.pdf)_ (when [this PDF](assets/pdfs/dod_character.pdf) is used for the `formPdfBytes` variable, [this image](assets/images/small_mario.png) is used for the `marioImageBytes` variable, and [this image](assets/images/mario_emblem.png) is used for the `emblemImageBytes` variable).

[Try the JSFiddle demo](https://jsfiddle.net/Hopding/0mwfqkv6/3/)

> See also [Creating and Filling Forms](#creating-and-filling-forms)

<!-- prettier-ignore -->
```js
import { PDFDocument } from 'pdf-lib'

// These should be Uint8Arrays or ArrayBuffers
// This data can be obtained in a number of different ways
// If your running in a Node environment, you could use fs.readFile()
// In the browser, you could make a fetch() call and use res.arrayBuffer()
const formPdfBytes = ...
const marioImageBytes = ...
const emblemImageBytes = ...

// Load a PDF with form fields
const pdfDoc = await PDFDocument.load(formPdfBytes)

// Embed the Mario and emblem images
const marioImage = await pdfDoc.embedPng(marioImageBytes)
const emblemImage = await pdfDoc.embedPng(emblemImageBytes)

// Get the form containing all the fields
const form = pdfDoc.getForm()

// Get all fields in the PDF by their names
const nameField = form.getTextField('CharacterName 2')
const ageField = form.getTextField('Age')
const heightField = form.getTextField('Height')
const weightField = form.getTextField('Weight')
const eyesField = form.getTextField('Eyes')
const skinField = form.getTextField('Skin')
const hairField = form.getTextField('Hair')

const alliesField = form.getTextField('Allies')
const factionField = form.getTextField('FactionName')
const backstoryField = form.getTextField('Backstory')
const traitsField = form.getTextField('Feat+Traits')
const treasureField = form.getTextField('Treasure')

const characterImageField = form.getButton('CHARACTER IMAGE')
const factionImageField = form.getTextField('Faction Symbol Image')

// Fill in the basic info fields
nameField.setText('Mario')
ageField.setText('24 years')
heightField.setText(`5' 1"`)
weightField.setText('196 lbs')
eyesField.setText('blue')
skinField.setText('white')
hairField.setText('brown')

// Fill the character image field with our Mario image
characterImageField.setImage(marioImage)

// Fill in the allies field
alliesField.setText(
  [
    `Allies:`,
    `  • Princess Daisy`,
    `  • Princess Peach`,
    `  • Rosalina`,
    `  • Geno`,
    `  • Luigi`,
    `  • Donkey Kong`,
    `  • Yoshi`,
    `  • Diddy Kong`,
    ``,
    `Organizations:`,
    `  • Italian Plumbers Association`,
  ].join('\n'),
)

// Fill in the faction name field
factionField.setText(`Mario's Emblem`)

// Fill the faction image field with our emblem image
factionImageField.setImage(emblemImage)

// Fill in the backstory field
backstoryField.setText(
  `Mario is a fictional character in the Mario video game franchise, owned by Nintendo and created by Japanese video game designer Shigeru Miyamoto. Serving as the company's mascot and the eponymous protagonist of the series, Mario has appeared in over 200 video games since his creation. Depicted as a short, pudgy, Italian plumber who resides in the Mushroom Kingdom, his adventures generally center upon rescuing Princess Peach from the Koopa villain Bowser. His younger brother and sidekick is Luigi.`,
)

// Fill in the traits field
traitsField.setText(
  [
    `Mario can use three basic three power-ups:`,
    `  • the Super Mushroom, which causes Mario to grow larger`,
    `  • the Fire Flower, which allows Mario to throw fireballs`,
    `  • the Starman, which gives Mario temporary invincibility`,
  ].join('\n'),
)

// Fill in the treasure field
treasureField.setText(['• Gold coins', '• Treasure chests'].join('\n'))

// Serialize the PDFDocument to bytes (a Uint8Array)
const pdfBytes = await pdfDoc.save()

// For example, `pdfBytes` can be:
//   • Written to a file in Node
//   • Downloaded from the browser
//   • Rendered in an <iframe>
```

### Flatten Form

_This example produces [this PDF](assets/pdfs/examples/flatten_form.pdf)_ (when [this PDF](assets/pdfs/form_to_flatten.pdf) is used for the `formPdfBytes` variable).

[Try the JSFiddle demo](https://jsfiddle.net/Hopding/skevywdz/2/)

<!-- prettier-ignore -->
```js
import { PDFDocument } from 'pdf-lib'

// This should be a Uint8Array or ArrayBuffer
// This data can be obtained in a number of different ways
// If your running in a Node environment, you could use fs.readFile()
// In the browser, you could make a fetch() call and use res.arrayBuffer()
const formPdfBytes = ...

// Load a PDF with form fields
const pdfDoc = await PDFDocument.load(formPdfBytes)

// Get the form containing all the fields
const form = pdfDoc.getForm()

// Fill the form's fields
form.getTextField('Text1').setText('Some Text');

form.getRadioGroup('Group2').select('Choice1');
form.getRadioGroup('Group3').select('Choice3');
form.getRadioGroup('Group4').select('Choice1');

form.getCheckBox('Check Box3').check();
form.getCheckBox('Check Box4').uncheck();

form.getDropdown('Dropdown7').select('Infinity');

form.getOptionList('List Box6').select('Honda');

// Flatten the form's fields
form.flatten();

// Serialize the PDFDocument to bytes (a Uint8Array)
const pdfBytes = await pdfDoc.save()

// For example, `pdfBytes` can be:
//   • Written to a file in Node
//   • Downloaded from the browser
//   • Rendered in an <iframe>
```

### Work with XFA Forms

XFA (XML Forms Architecture) forms are complex, dynamic PDF forms commonly used for government forms, tax documents, and enterprise applications. Unlike standard AcroForms, XFA forms embed their structure and JavaScript in XML format.

**Important:** To preserve XFA forms when loading and saving PDFs, use the `preserveXFA` option:

<!-- prettier-ignore -->
```js
import { PDFDocument } from 'pdf-lib'

const xfaPdfBytes = ... // Load your XFA PDF

// Load with XFA preservation
const pdfDoc = await PDFDocument.load(xfaPdfBytes, { 
  preserveXFA: true 
})

// Make modifications...

// Save the document
const pdfBytes = await pdfDoc.save()
```

**Note:** The `preserveXFA` option must be set to `true` when loading to preserve XFA data. XFA preservation during save happens automatically if it was preserved during load.

**Scope and limitations (v1):** XFA support in pdf-lib is intentionally narrow and targets the common government/tax "static" XFA layout:

- Only the array form of `/XFA` (alternating name/stream pairs) is supported; the single-stream packaging is not read or written.
- Only the `template` packet is inspected — dynamic packaging and other packets (e.g. `datasets`, `config`, `xdp` wrappers) are not (re)generated. Editing JavaScript does not re-render or repackage the form.
- XFA signature detection reads `<signature>`/`<manifest>` entries from the template; it does not validate or create cryptographic signatures.
- `getForm()` removes XFA data unless the document was loaded with `preserveXFA: true`, so call the XFA helpers before `getForm()` (or load with `preserveXFA: true`).
- Template XML is decoded as UTF-8 with a latin1 fallback; exotic encodings may not round-trip cleanly.

### Extract XFA JavaScript

XFA forms often contain JavaScript for validation, calculations, and data import/export. You can extract all JavaScript from an XFA form:

<!-- prettier-ignore -->
```js
import { PDFDocument } from 'pdf-lib'

const xfaPdfBytes = ... // Load your XFA PDF

// Load the PDF with XFA preservation
const pdfDoc = await PDFDocument.load(xfaPdfBytes, { 
  preserveXFA: true 
})

// Extract all XFA JavaScript
const scripts = pdfDoc.getXFAJavaScripts()

// Each script contains:
// - field: The field name (e.g., 'Button1', 'TextField2')
// - event: The event name (e.g., 'event__click', 'event__change')
// - script: The JavaScript code

console.log(`Found ${scripts.length} scripts`)

scripts.forEach((script) => {
  console.log(`Field: ${script.field}`)
  console.log(`Event: ${script.event}`)
  console.log(`Code: ${script.script}`)
})

// Find specific scripts
const clickHandlers = scripts.filter(s => 
  s.event.includes('click')
)

const validationScripts = scripts.filter(s => 
  s.script.includes('validate') || s.script.includes('Validate')
)
```

### Modify XFA JavaScript

You can modify JavaScript in XFA forms to customize behavior, add logging, or fix issues:

<!-- prettier-ignore -->
```js
import { PDFDocument } from 'pdf-lib'

const xfaPdfBytes = ... // Load your XFA PDF

// Load the PDF with XFA preservation
const pdfDoc = await PDFDocument.load(xfaPdfBytes, { 
  preserveXFA: true 
})

// Extract scripts to find what you want to modify
const scripts = pdfDoc.getXFAJavaScripts()
const importButton = scripts.find(s => s.field === 'ImportButton')

if (importButton) {
  // Modify the import button's click handler
  const newScript = `
    // Custom import handler
    try {
      console.println("Starting import...");
      ${importButton.script}
      console.println("Import completed!");
    } catch(e) {
      xfa.host.messageBox("Error: " + e.message);
    }
  `
  
  try {
    pdfDoc.setXFAJavaScript(
      'ImportButton',         // field name
      importButton.event,     // event name (e.g., 'event__click')
      newScript               // new JavaScript code
    )
    console.log('Successfully modified XFA JavaScript')
  } catch (error) {
    console.error('Failed to modify script:', error.message)
  }
}

// Save the document
const pdfBytes = await pdfDoc.save()

// The modified PDF will have the updated JavaScript
```

**Use Cases:**

- Add error handling to existing scripts
- Modify validation rules
- Add logging for debugging
- Customize import/export behavior
- Fix compatibility issues

### Extract Document JavaScript

PDF documents can contain document-level JavaScript that executes when the document is opened. You can extract these scripts:

<!-- prettier-ignore -->
```js
import { PDFDocument } from 'pdf-lib'

const pdfBytes = ... // Load your PDF

const pdfDoc = await PDFDocument.load(pdfBytes)

// Extract all document-level JavaScript
const scripts = pdfDoc.getDocumentJavaScripts()

// Each script contains:
// - name: The script name
// - script: The JavaScript code

console.log(`Found ${scripts.length} document scripts`)

scripts.forEach((script) => {
  console.log(`Script name: ${script.name}`)
  console.log(`Code: ${script.script}`)
})

// Find specific scripts
const initScripts = scripts.filter(s => 
  s.name.toLowerCase().includes('init')
)
```

**Note:** Document-level JavaScript is different from XFA JavaScript. Document-level scripts are stored in the document's Names dictionary and execute when the PDF is opened. XFA JavaScript is embedded in XFA form templates.

### Copy Pages

_This example produces [this PDF](assets/pdfs/examples/copy_pages.pdf)_ (when [this PDF](assets/pdfs/with_update_sections.pdf) is used for the `firstDonorPdfBytes` variable and [this PDF](assets/pdfs/with_large_page_count.pdf) is used for the `secondDonorPdfBytes` variable).

[Try the JSFiddle demo](https://jsfiddle.net/Hopding/ybank8s9/2/)

<!-- prettier-ignore -->
```js
import { PDFDocument } from 'pdf-lib'

// Create a new PDFDocument
const pdfDoc = await PDFDocument.create()

// These should be Uint8Arrays or ArrayBuffers
// This data can be obtained in a number of different ways
// If your running in a Node environment, you could use fs.readFile()
// In the browser, you could make a fetch() call and use res.arrayBuffer()
const firstDonorPdfBytes = ...
const secondDonorPdfBytes = ...

// Load a PDFDocument from each of the existing PDFs
const firstDonorPdfDoc = await PDFDocument.load(firstDonorPdfBytes)
const secondDonorPdfDoc = await PDFDocument.load(secondDonorPdfBytes)

// Copy the 1st page from the first donor document, and
// the 743rd page from the second donor document
const [firstDonorPage] = await pdfDoc.copyPages(firstDonorPdfDoc, [0])
const [secondDonorPage] = await pdfDoc.copyPages(secondDonorPdfDoc, [742])

// Add the first copied page
pdfDoc.addPage(firstDonorPage)

// Insert the second copied page to index 0, so it will be the
// first page in `pdfDoc`
pdfDoc.insertPage(0, secondDonorPage)

// Serialize the PDFDocument to bytes (a Uint8Array)
const pdfBytes = await pdfDoc.save()

// For example, `pdfBytes` can be:
//   • Written to a file in Node
//   • Downloaded from the browser
//   • Rendered in an <iframe>
```

### Embed PNG and JPEG Images

_This example produces [this PDF](assets/pdfs/examples/embed_png_and_jpeg_images.pdf)_ (when [this image](assets/images/cat_riding_unicorn.jpg) is used for the `jpgImageBytes` variable and [this image](assets/images/minions_banana_alpha.png) is used for the `pngImageBytes` variable).

[Try the JSFiddle demo](https://jsfiddle.net/Hopding/bcya43ju/5/)

<!-- prettier-ignore -->
```js
import { PDFDocument } from 'pdf-lib'

// These should be Uint8Arrays or ArrayBuffers
// This data can be obtained in a number of different ways
// If your running in a Node environment, you could use fs.readFile()
// In the browser, you could make a fetch() call and use res.arrayBuffer()
const jpgImageBytes = ...
const pngImageBytes = ...

// Create a new PDFDocument
const pdfDoc = await PDFDocument.create()

// Embed the JPG image bytes and PNG image bytes
const jpgImage = await pdfDoc.embedJpg(jpgImageBytes)
const pngImage = await pdfDoc.embedPng(pngImageBytes)

// Get the width/height of the JPG image scaled down to 25% of its original size
const jpgDims = jpgImage.scale(0.25)

// Get the width/height of the PNG image scaled down to 50% of its original size
const pngDims = pngImage.scale(0.5)

// Add a blank page to the document
const page = pdfDoc.addPage()

// Draw the JPG image in the center of the page
page.drawImage(jpgImage, {
  x: page.getWidth() / 2 - jpgDims.width / 2,
  y: page.getHeight() / 2 - jpgDims.height / 2,
  width: jpgDims.width,
  height: jpgDims.height,
})

// Draw the PNG image near the lower right corner of the JPG image
page.drawImage(pngImage, {
  x: page.getWidth() / 2 - pngDims.width / 2 + 75,
  y: page.getHeight() / 2 - pngDims.height,
  width: pngDims.width,
  height: pngDims.height,
})

// Serialize the PDFDocument to bytes (a Uint8Array)
const pdfBytes = await pdfDoc.save()

// For example, `pdfBytes` can be:
//   • Written to a file in Node
//   • Downloaded from the browser
//   • Rendered in an <iframe>
```

### Embed PDF Pages

_This example produces [this PDF](assets/pdfs/examples/embed_pdf_pages.pdf)_ (when [this PDF](assets/pdfs/american_flag.pdf) is used for the `americanFlagPdfBytes` variable and [this PDF](assets/pdfs/us_constitution.pdf) is used for the `usConstitutionPdfBytes` variable).

[Try the JSFiddle demo](https://jsfiddle.net/Hopding/Lyb16ocj/13/)

<!-- prettier-ignore -->
```js
import { PDFDocument } from 'pdf-lib'

// These should be Uint8Arrays or ArrayBuffers
// This data can be obtained in a number of different ways
// If your running in a Node environment, you could use fs.readFile()
// In the browser, you could make a fetch() call and use res.arrayBuffer()
const americanFlagPdfBytes = ...
const usConstitutionPdfBytes = ...

// Create a new PDFDocument
const pdfDoc = await PDFDocument.create()

// Embed the American flag PDF bytes
const [americanFlag] = await pdfDoc.embedPdf(americanFlagPdfBytes)

// Load the U.S. constitution PDF bytes
const usConstitutionPdf = await PDFDocument.load(usConstitutionPdfBytes)

// Embed the second page of the constitution and clip the preamble
const preamble = await pdfDoc.embedPage(usConstitutionPdf.getPages()[1], {
  left: 55,
  bottom: 485,
  right: 300,
  top: 575,
})

// Get the width/height of the American flag PDF scaled down to 30% of
// its original size
const americanFlagDims = americanFlag.scale(0.3)

// Get the width/height of the preamble clipping scaled up to 225% of
// its original size
const preambleDims = preamble.scale(2.25)

// Add a blank page to the document
const page = pdfDoc.addPage()

// Draw the American flag image in the center top of the page
page.drawPage(americanFlag, {
  ...americanFlagDims,
  x: page.getWidth() / 2 - americanFlagDims.width / 2,
  y: page.getHeight() - americanFlagDims.height - 150,
})

// Draw the preamble clipping in the center bottom of the page
page.drawPage(preamble, {
  ...preambleDims,
  x: page.getWidth() / 2 - preambleDims.width / 2,
  y: page.getHeight() / 2 - preambleDims.height / 2 - 50,
})

// Serialize the PDFDocument to bytes (a Uint8Array)
const pdfBytes = await pdfDoc.save()

// For example, `pdfBytes` can be:
//   • Written to a file in Node
//   • Downloaded from the browser
//   • Rendered in an <iframe>
```

### Embed Font and Measure Text

`pdf-lib` relies on [`@cantoo/fontkit`](https://www.npmjs.com/package/@cantoo/fontkit) to support embedding custom fonts. You must add the `@cantoo/fontkit` module to your project and register it using `pdfDoc.registerFontkit(...)` before embedding custom fonts. Upstream [`fontkit`](https://www.npmjs.com/package/fontkit) and the older [`@pdf-lib/fontkit`](https://www.npmjs.com/package/@pdf-lib/fontkit) package still work if you already depend on them.

> **[See below for detailed installation instructions on installing `@cantoo/fontkit` as an NPM or UMD module.](#fontkit-installation)**

_This example produces [this PDF](assets/pdfs/examples/embed_font_and_measure_text.pdf)_ (when [this font](assets/fonts/ubuntu/Ubuntu-R.ttf) is used for the `fontBytes` variable).

[Try the JSFiddle demo](https://jsfiddle.net/Hopding/rgu6ca59/2/)

<!-- prettier-ignore -->
```js
import { PDFDocument, rgb } from 'pdf-lib'
import fontkit from '@cantoo/fontkit'

// This should be a Uint8Array or ArrayBuffer
// This data can be obtained in a number of different ways
// If you're running in a Node environment, you could use fs.readFile()
// In the browser, you could make a fetch() call and use res.arrayBuffer()
const fontBytes = ...

// Create a new PDFDocument
const pdfDoc = await PDFDocument.create()

// Register the `fontkit` instance
pdfDoc.registerFontkit(fontkit)

// Embed our custom font in the document
const customFont = await pdfDoc.embedFont(fontBytes)

// Add a blank page to the document
const page = pdfDoc.addPage()

// Create a string of text and measure its width and height in our custom font
const text = 'This is text in an embedded font!'
const textSize = 35
const textWidth = customFont.widthOfTextAtSize(text, textSize)
const textHeight = customFont.heightAtSize(textSize)

// Draw the string of text on the page
page.drawText(text, {
  x: 40,
  y: 450,
  size: textSize,
  font: customFont,
  color: rgb(0, 0.53, 0.71),
})

// Draw a box around the string of text
page.drawRectangle({
  x: 40,
  y: 450,
  width: textWidth,
  height: textHeight,
  borderColor: rgb(1, 0, 0),
  borderWidth: 1.5,
})

// Serialize the PDFDocument to bytes (a Uint8Array)
const pdfBytes = await pdfDoc.save()

// For example, `pdfBytes` can be:
//   • Written to a file in Node
//   • Downloaded from the browser
//   • Rendered in an <iframe>
```

### Add Attachments

_This example produces [this PDF](assets/pdfs/examples/add_attachments.pdf)_ (when [this image](assets/images/cat_riding_unicorn.jpg) is used for the `jpgAttachmentBytes` variable and [this PDF](assets/pdfs/us_constitution.pdf) is used for the `pdfAttachmentBytes` variable).

[Try the JSFiddle demo](https://jsfiddle.net/Hopding/9snL63wj/5/)

<!-- prettier-ignore -->
```js
import { PDFDocument } from 'pdf-lib'

// These should be Uint8Arrays or ArrayBuffers
// This data can be obtained in a number of different ways
// If your running in a Node environment, you could use fs.readFile()
// In the browser, you could make a fetch() call and use res.arrayBuffer()
const jpgAttachmentBytes = ...
const pdfAttachmentBytes = ...

// Create a new PDFDocument
const pdfDoc = await PDFDocument.create()

// Add the JPG attachment
await pdfDoc.attach(jpgAttachmentBytes, 'cat_riding_unicorn.jpg', {
  mimeType: 'image/jpeg',
  description: 'Cool cat riding a unicorn! 🦄🐈🕶️',
  creationDate: new Date('2019/12/01'),
  modificationDate: new Date('2020/04/19'),
})

// Add the PDF attachment
await pdfDoc.attach(pdfAttachmentBytes, 'us_constitution.pdf', {
  mimeType: 'application/pdf',
  description: 'Constitution of the United States 🇺🇸🦅',
  creationDate: new Date('1787/09/17'),
  modificationDate: new Date('1992/05/07'),
})

// Add a page with some text
const page = pdfDoc.addPage();
page.drawText('This PDF has two attachments', { x: 135, y: 415 })

// Serialize the PDFDocument to bytes (a Uint8Array)
const pdfBytes = await pdfDoc.save()

// For example, `pdfBytes` can be:
//   • Written to a file in Node
//   • Downloaded from the browser
//   • Rendered in an <iframe>
```

### Extract Attachments

If you load a PDF that has `cars.csv` as an attachment, you can use the
following to extract the attachments:

<!-- prettier-ignore -->
```js
const pdfDoc = await PDFDocument.load(...)
const attachments = pdfDoc.getAttachments()
const csv = attachments.find(({ name }) => name === 'cars.csv')
fs.writeFileSync(csv.name, csv.data)
```

> NOTE: The method also finds attachments added after the last call to
> `save()`.

### Create PDF/A Documents

Convert a document to PDF/A by calling `convertToPDFA()`. This adds the
structural pieces PDF/A requires (document `/ID`, OutputIntent with an embedded
sRGB ICC profile, uncompressed XMP metadata with `pdfaid`, and the appropriate
PDF header version). It does **not** rewrite arbitrary content to make it
compliant — in particular, drawn text must use an **embedded** font (the 14
standard fonts are not PDF/A compliant). Validate the result with a tool such as
[veraPDF](https://verapdf.org/).

<!-- prettier-ignore -->
```js
import { PDFDocument } from '@cantoo/pdf-lib'
import fontkit from '@cantoo/fontkit'

const fontBytes = ... // e.g. fs.readFileSync('Roboto-Regular.ttf')

const pdfDoc = await PDFDocument.create()
pdfDoc.registerFontkit(fontkit)

const font = await pdfDoc.embedFont(fontBytes)
const page = pdfDoc.addPage()
page.drawText('Hello PDF/A', { x: 50, y: 700, size: 24, font })

pdfDoc.setTitle('Archival document')
pdfDoc.setAuthor('ACME GmbH')

// '1B' | '2B' | '2U' | '3B' | '3U' — defaults to '3B'
pdfDoc.convertToPDFA({ conformance: '3B' })

const pdfBytes = await pdfDoc.save()
```

After conversion, pdf-lib keeps the Info dictionary and XMP metadata in sync on
`save()`, while preserving any _strictly foreign_ XMP `rdf:Description` blocks
(for example Factur-X schemas passed via `extensions`). Descriptions that mix
owned namespaces (`dc` / `xmp` / `pdf` / `pdfaid`) with custom ones are not
preserved — re-supply them via `extensions` or `embedFacturX()`. Loading an
existing PDF/A file does not enable that sync by itself: call `convertToPDFA()`
(or `embedFacturX()`) again if you change Info fields and need them mirrored.

### Embed Factur-X / ZUGFeRD Invoices

`embedFacturX()` wraps a human-readable PDF and a machine-readable Factur-X /
ZUGFeRD XML into a PDF/A-3 hybrid invoice: it ensures PDF/A-3 (converting to 3B
if needed, or keeping an existing 3U/3B level), writes the required `fx:` XMP
properties (plus the PDF/A extension schema), and attaches the XML with an
associated-file relationship.

It does **not** generate or validate the Cross Industry Invoice XML — pass a
complete `factur-x.xml` from your invoicing stack.

<!-- prettier-ignore -->
```js
import { PDFDocument, embedFacturX } from '@cantoo/pdf-lib'
import fontkit from '@cantoo/fontkit'

const fontBytes = ...
const invoiceXmlBytes = ... // Factur-X / ZUGFeRD XML (Uint8Array)

const pdfDoc = await PDFDocument.create()
pdfDoc.registerFontkit(fontkit)

const font = await pdfDoc.embedFont(fontBytes)
const page = pdfDoc.addPage()
page.drawText('Invoice 2026-0001', { x: 50, y: 700, size: 18, font })
// ... draw the rest of the human-readable invoice with the embedded font ...

pdfDoc.setTitle('Invoice 2026-0001')
pdfDoc.setAuthor('ACME GmbH')

await embedFacturX(pdfDoc, invoiceXmlBytes, {
  // MINIMUM | BASIC WL | BASIC | EN 16931 | EXTENDED | XRECHNUNG
  conformanceLevel: 'EN 16931',
})

const pdfBytes = await pdfDoc.save()
```

### Set Document Metadata

_This example produces [this PDF](assets/pdfs/examples/set_document_metadata.pdf)_.

[Try the JSFiddle demo](https://jsfiddle.net/Hopding/vcwmfnbe/2/)

<!-- prettier-ignore -->
```js
import { PDFDocument, StandardFonts } from 'pdf-lib'

// Create a new PDFDocument
const pdfDoc = await PDFDocument.create()

// Embed the Times Roman font
const timesRomanFont = await pdfDoc.embedFont(StandardFonts.TimesRoman)

// Add a page and draw some text on it
const page = pdfDoc.addPage([500, 600])
page.setFont(timesRomanFont)
page.drawText('The Life of an Egg', { x: 60, y: 500, size: 50 })
page.drawText('An Epic Tale of Woe', { x: 125, y: 460, size: 25 })

// Set all available metadata fields on the PDFDocument. Note that these fields
// are visible in the "Document Properties" section of most PDF readers.
pdfDoc.setTitle('🥚 The Life of an Egg 🍳')
pdfDoc.setAuthor('Humpty Dumpty')
pdfDoc.setSubject('📘 An Epic Tale of Woe 📖')
pdfDoc.setKeywords(['eggs', 'wall', 'fall', 'king', 'horses', 'men'])
pdfDoc.setProducer('PDF App 9000 🤖')
pdfDoc.setCreator('pdf-lib (https://github.com/Hopding/pdf-lib)')
pdfDoc.setCreationDate(new Date('2018-06-24T01:58:37.228Z'))
pdfDoc.setModificationDate(new Date('2019-12-21T07:00:11.000Z'))

// Serialize the PDFDocument to bytes (a Uint8Array)
const pdfBytes = await pdfDoc.save()

// For example, `pdfBytes` can be:
//   • Written to a file in Node
//   • Downloaded from the browser
//   • Rendered in an <iframe>
```

### Read Document Metadata

[Try the JSFiddle demo](https://jsfiddle.net/Hopding/eg8rfz3k/16/)

<!-- prettier-ignore -->
```js
import { PDFDocument } from 'pdf-lib'

// This should be a Uint8Array or ArrayBuffer
// This data can be obtained in a number of different ways
// If your running in a Node environment, you could use fs.readFile()
// In the browser, you could make a fetch() call and use res.arrayBuffer()
const existingPdfBytes = ...

// Load a PDFDocument without updating its existing metadata
const pdfDoc = await PDFDocument.load(existingPdfBytes, {
  updateMetadata: false
})

// Print all available metadata fields
console.log('Title:', pdfDoc.getTitle())
console.log('Author:', pdfDoc.getAuthor())
console.log('Subject:', pdfDoc.getSubject())
console.log('Creator:', pdfDoc.getCreator())
console.log('Keywords:', pdfDoc.getKeywords())
console.log('Producer:', pdfDoc.getProducer())
console.log('Creation Date:', pdfDoc.getCreationDate())
console.log('Modification Date:', pdfDoc.getModificationDate())
```

This script outputs the following (_when [this PDF](assets/pdfs/with_cropbox.pdf) is used for the `existingPdfBytes` variable_):

```
Title: Microsoft Word - Basic Curriculum Vitae example.doc
Author: Administrator
Subject: undefined
Creator: PScript5.dll Version 5.2
Keywords: undefined
Producer: Acrobat Distiller 8.1.0 (Windows)
Creation Date: 2010-07-29T14:26:00.000Z
Modification Date: 2010-07-29T14:26:00.000Z
```

### Set Viewer Preferences

<!-- prettier-ignore -->
```js
import {
  PDFDocument,
  StandardFonts,
  NonFullScreenPageMode,
  ReadingDirection,
  PrintScaling,
  Duplex,
  PDFName,
} from 'pdf-lib'

// Create a new PDFDocument
const pdfDoc = await PDFDocument.create()

// Embed the Times Roman font
const timesRomanFont = await pdfDoc.embedFont(StandardFonts.TimesRoman)

// Add a page and draw some text on it
const page = pdfDoc.addPage([500, 600])
page.setFont(timesRomanFont)
page.drawText('The Life of an Egg', { x: 60, y: 500, size: 50 })
page.drawText('An Epic Tale of Woe', { x: 125, y: 460, size: 25 })

// Set all available viewer preferences on the PDFDocument:
const viewerPrefs = pdfDoc.catalog.getOrCreateViewerPreferences()
viewerPrefs.setHideToolbar(true)
viewerPrefs.setHideMenubar(true)
viewerPrefs.setHideWindowUI(true)
viewerPrefs.setFitWindow(true)
viewerPrefs.setCenterWindow(true)
viewerPrefs.setDisplayDocTitle(true)

// Set the PageMode (otherwise setting NonFullScreenPageMode has no meaning)
pdfDoc.catalog.set(PDFName.of('PageMode'), PDFName.of('FullScreen'))

// Set what happens when fullScreen is closed
viewerPrefs.setNonFullScreenPageMode(NonFullScreenPageMode.UseOutlines)

viewerPrefs.setReadingDirection(ReadingDirection.L2R)
viewerPrefs.setPrintScaling(PrintScaling.None)
viewerPrefs.setDuplex(Duplex.DuplexFlipLongEdge)
viewerPrefs.setPickTrayByPDFSize(true)

// We can set the default print range to only the first page
viewerPrefs.setPrintPageRange({ start: 0, end: 0 })

// Or we can supply noncontiguous ranges (e.g. pages 1, 3, and 5-7)
viewerPrefs.setPrintPageRange([
  { start: 0, end: 0 },
  { start: 2, end: 2 },
  { start: 4, end: 6 },
])

viewerPrefs.setNumCopies(2)

// Serialize the PDFDocument to bytes (a Uint8Array)
const pdfBytes = await pdfDoc.save()

// For example, `pdfBytes` can be:
//   • Written to a file in Node
//   • Downloaded from the browser
//   • Rendered in an <iframe>
```

### Read Viewer Preferences

<!-- prettier-ignore -->
```js
import { PDFDocument } from 'pdf-lib'

// This should be a Uint8Array or ArrayBuffer
// This data can be obtained in a number of different ways
// If your running in a Node environment, you could use fs.readFile()
// In the browser, you could make a fetch() call and use res.arrayBuffer()
const existingPdfBytes = ...

// Load a PDFDocument without updating its existing metadata
const pdfDoc = await PDFDocument.load(existingPdfBytes)
const viewerPrefs = pdfDoc.catalog.getOrCreateViewerPreferences()

// Print all available viewer preference fields
console.log('HideToolbar:', viewerPrefs.getHideToolbar())
console.log('HideMenubar:', viewerPrefs.getHideMenubar())
console.log('HideWindowUI:', viewerPrefs.getHideWindowUI())
console.log('FitWindow:', viewerPrefs.getFitWindow())
console.log('CenterWindow:', viewerPrefs.getCenterWindow())
console.log('DisplayDocTitle:', viewerPrefs.getDisplayDocTitle())
console.log('NonFullScreenPageMode:', viewerPrefs.getNonFullScreenPageMode())
console.log('ReadingDirection:', viewerPrefs.getReadingDirection())
console.log('PrintScaling:', viewerPrefs.getPrintScaling())
console.log('Duplex:', viewerPrefs.getDuplex())
console.log('PickTrayByPDFSize:', viewerPrefs.getPickTrayByPDFSize())
console.log('PrintPageRange:', viewerPrefs.getPrintPageRange())
console.log('NumCopies:', viewerPrefs.getNumCopies())
```

This script outputs the following (_when [this PDF](assets/pdfs/with_viewer_prefs.pdf) is used for the `existingPdfBytes` variable_):

```
HideToolbar: true
HideMenubar: true
HideWindowUI: false
FitWindow: true
CenterWindow: true
DisplayDocTitle: true
NonFullScreenPageMode: UseNone
ReadingDirection: R2L
PrintScaling: None
Duplex: DuplexFlipLongEdge
PickTrayByPDFSize: true
PrintPageRange: [ { start: 1, end: 1 }, { start: 3, end: 4 } ]
NumCopies: 2
```

### Draw SVG Paths

_This example produces [this PDF](assets/pdfs/examples/draw_svg_paths.pdf)_.

[Try the JSFiddle demo](https://jsfiddle.net/Hopding/bwaomr9h/2/)

<!-- prettier-ignore -->
```js
import { PDFDocument, rgb } from 'pdf-lib'

// SVG path for a wavy line
const svgPath =
  'M 0,20 L 100,160 Q 130,200 150,120 C 190,-40 200,200 300,150 L 400,90'

// Create a new PDFDocument
const pdfDoc = await PDFDocument.create()

// Add a blank page to the document
const page = pdfDoc.addPage()
page.moveTo(100, page.getHeight() - 5)

// Draw the SVG path as a black line
page.moveDown(25)
page.drawSvgPath(svgPath)

// Draw the SVG path as a thick green line
page.moveDown(200)
page.drawSvgPath(svgPath, { borderColor: rgb(0, 1, 0), borderWidth: 5 })

// Draw the SVG path and fill it with red
page.moveDown(200)
page.drawSvgPath(svgPath, { color: rgb(1, 0, 0) })

// Draw the SVG path at 50% of its original size
page.moveDown(200)
page.drawSvgPath(svgPath, { scale: 0.5 })

// Serialize the PDFDocument to bytes (a Uint8Array)
const pdfBytes = await pdfDoc.save()

// For example, `pdfBytes` can be:
//   • Written to a file in Node
//   • Downloaded from the browser
//   • Rendered in an <iframe>
```

### Draw SVG

```js
import { PDFDocument, rgb } from 'pdf-lib';

// SVG of a square inside a square
const svg = `<svg width="100" height="100">
  <rect y="0" x="0" width="100" height="100" fill="none" stroke="black"/>
  <rect y="25" x="25" width="50" height="50" fill="black"/>
</svg>`;
const svg2 =
  '<svg><image href="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAQAAAC1HAwCAAAAC0lEQVQYV2NgYAAAAAMAAWgmWQ0AAAAASUVORK5CYII="/></svg>';

// Create a new PDFDocument
const pdfDoc = await PDFDocument.create();

// Add a blank page to the document
const page = pdfDoc.addPage();

// drawSvg can accept the svg as a string, as long as there are no images in it
page.moveTo(100, 10);
page.drawSvg(svg);

// If the svg has images, or if you don't know if it does, you should call embedSVG first
page.moveTo(200, 10);
const pdfSvg = await pdfDoc.embedSvg(svg2);
page.drawSvg(pdfSvg);

// Serialize the PDFDocument to bytes (a Uint8Array)
const pdfBytes = await pdfDoc.save();
```

## Deno Usage

`pdf-lib` fully supports the exciting new [Deno](https://deno.land/) runtime! All of the [usage examples](#usage-examples) work in Deno. The only thing you need to do is change the imports for `pdf-lib` and `fontkit` to use CDN URLs, because Deno requires all modules to be referenced via URLs.

> **See also [How to Create and Modify PDF Files in Deno With pdf-lib](https://medium.com/swlh/how-to-create-and-modify-pdf-files-in-deno-ffaad7099b0?source=friends_link&sk=3da183bb776d059df428eaea52102f19)**

### Creating a Document with Deno

Below is the [**create document**](#create-document) example modified for Deno:

```js
import {
  PDFDocument,
  StandardFonts,
  rgb,
} from 'https://cdn.skypack.dev/pdf-lib@^1.11.1?dts';

const pdfDoc = await PDFDocument.create();
const timesRomanFont = await pdfDoc.embedFont(StandardFonts.TimesRoman);

const page = pdfDoc.addPage();
const { width, height } = page.getSize();
const fontSize = 30;
page.drawText('Creating PDFs in JavaScript is awesome!', {
  x: 50,
  y: height - 4 * fontSize,
  size: fontSize,
  font: timesRomanFont,
  color: rgb(0, 0.53, 0.71),
});

const pdfBytes = await pdfDoc.save();

await Deno.writeFile('out.pdf', pdfBytes);
```

If you save this script as `create-document.ts`, you can execute it using Deno with the following command:

```
deno run --allow-write create-document.ts
```

The resulting `out.pdf` file will look like [this PDF](assets/pdfs/examples/create_document.pdf).

### Embedding a Font with Deno

Here's a slightly more complicated example demonstrating how to embed a font and measure text in Deno:

```js
import {
  degrees,
  PDFDocument,
  rgb,
  StandardFonts,
} from 'https://cdn.skypack.dev/pdf-lib@^1.11.1?dts';
import fontkit from 'https://esm.sh/@cantoo/fontkit@2.0.9';

const url = 'https://pdf-lib.js.org/assets/ubuntu/Ubuntu-R.ttf';
const fontBytes = await fetch(url).then((res) => res.arrayBuffer());

const pdfDoc = await PDFDocument.create();

pdfDoc.registerFontkit(fontkit);
const customFont = await pdfDoc.embedFont(fontBytes);

const page = pdfDoc.addPage();

const text = 'This is text in an embedded font!';
const textSize = 35;
const textWidth = customFont.widthOfTextAtSize(text, textSize);
const textHeight = customFont.heightAtSize(textSize);

page.drawText(text, {
  x: 40,
  y: 450,
  size: textSize,
  font: customFont,
  color: rgb(0, 0.53, 0.71),
});
page.drawRectangle({
  x: 40,
  y: 450,
  width: textWidth,
  height: textHeight,
  borderColor: rgb(1, 0, 0),
  borderWidth: 1.5,
});

const pdfBytes = await pdfDoc.save();

await Deno.writeFile('out.pdf', pdfBytes);
```

If you save this script as `custom-font.ts`, you can execute it with the following command:

```
deno run --allow-write --allow-net custom-font.ts
```

The resulting `out.pdf` file will look like [this PDF](assets/pdfs/examples/embed_font_and_measure_text.pdf).

## Complete Examples

The [usage examples](#usage-examples) provide code that is brief and to the point, demonstrating the different features of `pdf-lib`. You can find complete working examples in the [`apps/`](apps/) directory. These apps are used to do manual testing of `pdf-lib` before every release (in addition to the [automated tests](tests/)).

There are currently four apps:

- [**`node`**](apps/node/) - contains [tests](apps/node/tests/) for `pdf-lib` in Node environments. These tests are a handy reference when trying to save/load PDFs, fonts, or images with `pdf-lib` from the filesystem. They also allow you to quickly open your PDFs in different viewers (Acrobat, Preview, Foxit, Chrome, Firefox, etc...) to ensure compatibility.
- [**`web`**](apps/web/) - contains [tests](apps/web/) for `pdf-lib` in browser environments. These tests are a handy reference when trying to save/load PDFs, fonts, or images with `pdf-lib` in a browser environment.
- [**`rn`**](apps/rn) - contains [tests](apps/rn/src/tests/) for `pdf-lib` in React Native environments. These tests are a handy reference when trying to save/load PDFs, fonts, or images with `pdf-lib` in a React Native environment.
- [**`deno`**](apps/deno) - contains [tests](apps/deno/tests/) for `pdf-lib` in Deno environments. These tests are a handy reference when trying to save/load PDFs, fonts, or images with `pdf-lib` from the filesystem.

## Installation

### NPM Module

To install the latest stable Cantoo fork:

```bash
# With npm
npm install --save @cantoo/pdf-lib

# With yarn
yarn add @cantoo/pdf-lib
```

(`pdf-lib` without the scope is still the original Hopding package on npm.)

### Vendored PNG & standard fonts

`@cantoo/pdf-lib` vendors [`@cantoo/upng`](./src/packages/upng) and [`@cantoo/standard-fonts`](./src/packages/standard-fonts) under `src/packages/` (same pattern as [`@cantoo/fontkit`](https://github.com/cantoo-scribe/fontkit)). Compression uses [`fflate`](https://www.npmjs.com/package/fflate) (no `pako`).

Install the subpackages alone if needed:

```bash
npm install @cantoo/upng @cantoo/standard-fonts
# or via pdf-lib subpath exports:
# import UPNG from '@cantoo/pdf-lib/upng'
# import { FontNames } from '@cantoo/pdf-lib/standard-fonts'
```

### UMD Module

You can also download `pdf-lib` as a UMD module from [unpkg](https://unpkg.com/#/) or [jsDelivr](https://www.jsdelivr.com/). The UMD builds have been compiled to ES5, so they should work [in any modern browser](https://caniuse.com/#feat=es5). UMD builds are useful if you aren't using a package manager or module bundler. For example, you can use them directly in the `<script>` tag of an HTML page.

The following builds are available:

- https://unpkg.com/pdf-lib/dist/pdf-lib.js
- https://unpkg.com/pdf-lib/dist/pdf-lib.min.js
- https://cdn.jsdelivr.net/npm/pdf-lib/dist/pdf-lib.js
- https://cdn.jsdelivr.net/npm/pdf-lib/dist/pdf-lib.min.js

> **NOTE:** if you are using the CDN scripts in production, you should include a specific version number in the URL, for example:
>
> - https://unpkg.com/pdf-lib@1.4.0/dist/pdf-lib.min.js
> - https://cdn.jsdelivr.net/npm/pdf-lib@1.4.0/dist/pdf-lib.min.js

When using a UMD build, you will have access to a global `window.PDFLib` variable. This variable contains all of the classes and functions exported by `pdf-lib`. For example:

```javascript
// NPM module
import { PDFDocument, rgb } from 'pdf-lib';

// UMD module
var PDFDocument = PDFLib.PDFDocument;
var rgb = PDFLib.rgb;
```

## Fontkit Installation

`pdf-lib` relies upon [`@cantoo/fontkit`](https://www.npmjs.com/package/@cantoo/fontkit) to support embedding custom fonts. You must add a fontkit-compatible module to your project and register it using `pdfDoc.registerFontkit(...)` before embedding custom fonts (see the [font embedding example](#embed-font-and-measure-text)). This module is not included by default because not all users need it, and it increases bundle size.

Prefer Cantoo's maintained [`@cantoo/fontkit`](https://www.npmjs.com/package/@cantoo/fontkit) fork (based on fontkit v2). It ships TypeScript types and UMD builds (including a minified file). Upstream [`fontkit`](https://www.npmjs.com/package/fontkit) and the older [`@pdf-lib/fontkit`](https://www.npmjs.com/package/@pdf-lib/fontkit) package remain supported for compatibility.

### Fontkit NPM Module

```bash
# With npm
npm install --save @cantoo/fontkit

# With yarn
yarn add @cantoo/fontkit
```

Register the `fontkit` instance before embedding custom fonts. With `@cantoo/fontkit` ≥2.0.9 you can use a default import (recommended). A namespace import also works. Upstream `fontkit` still has no default export under Node ESM, so use `import * as fontkit from 'fontkit'` if you depend on that package.

<!-- prettier-ignore -->
```js
import { PDFDocument } from 'pdf-lib'
import fontkit from '@cantoo/fontkit'

const pdfDoc = await PDFDocument.create()
pdfDoc.registerFontkit(fontkit)
```

### Fontkit UMD Module

You can also download `@cantoo/fontkit` as a UMD module from [unpkg](https://unpkg.com/#/) or [jsDelivr](https://www.jsdelivr.com/):

- https://unpkg.com/@cantoo/fontkit/dist/fontkit.umd.min.js
- https://unpkg.com/@cantoo/fontkit/dist/fontkit.umd.js
- https://cdn.jsdelivr.net/npm/@cantoo/fontkit/dist/fontkit.umd.min.js
- https://cdn.jsdelivr.net/npm/@cantoo/fontkit/dist/fontkit.umd.js

> **NOTE:** if you are using the CDN scripts in production, you should include a specific version number in the URL, for example:
>
> - https://unpkg.com/@cantoo/fontkit@2.0.9/dist/fontkit.umd.min.js
> - https://cdn.jsdelivr.net/npm/@cantoo/fontkit@2.0.9/dist/fontkit.umd.min.js

When using a UMD build, you will have access to a global `window.fontkit` variable. To register the `fontkit` instance:

<!-- prettier-ignore -->
```js
var pdfDoc = await PDFLib.PDFDocument.create()
pdfDoc.registerFontkit(fontkit)
```

## Documentation

API documentation is available on the project site at https://pdf-lib.js.org/docs/api/.

The repo for the project site (and generated documentation files) is
located here: https://github.com/Hopding/pdf-lib-docs.

## Fonts and Unicode

When working with PDFs, you will frequently come across the terms "character encoding" and "font". If you have experience in web development, you may wonder why these are so prevalent. Aren't they just annoying details that you shouldn't need to worry about? Shouldn't PDF libraries and readers be able to handle all of this for you like web browsers can? Unfortunately, this is not the case. The nature of the PDF file format makes it very difficult to avoid thinking about character encodings and fonts when working with PDFs.

`pdf-lib` does its best to simplify things for you. But it can't perform magic. This means you should be aware of the following:

- **There are 14 standard fonts** defined in the PDF specification. They are as follows: _Times Roman_ (normal, bold, and italic), _Helvetica_ (normal, bold, and italic), _Courier_ (normal, bold, and italic), _ZapfDingbats_ (normal), and _Symbol_ (normal). These 14 fonts are guaranteed to be available in PDF readers. As such, you do not need to embed any font data if you wish to use one of these fonts. You can use a standard font like so:
  <!-- prettier-ignore -->
  ```js
  import { PDFDocument, StandardFonts } from 'pdf-lib'
  const pdfDoc = await PDFDocument.create()
  const courierFont = await pdfDoc.embedFont(StandardFonts.Courier)
  const page = pdfDoc.addPage()
  page.drawText('Some boring latin text in the Courier font', {
    font: courierFont,
  })
  ```
- **The standard fonts do not support all characters** available in Unicode. The Times Roman, Helvetica, and Courier fonts use WinAnsi encoding (aka [Windows-1252](https://en.wikipedia.org/wiki/Windows-1252)). The WinAnsi character set only supports 218 characters in the Latin alphabet. For this reason, many users will find the standard fonts insufficient for their use case. This is unfortunate, but there's nothing that PDF libraries can do to change this. This is a result of the PDF specification and its age. Note that the [ZapfDingbats](https://en.wikipedia.org/wiki/Zapf_Dingbats) and [Symbol](<https://en.wikipedia.org/wiki/Symbol_(typeface)>) fonts use their own specialized encodings that support 203 and 194 characters, respectively. However, the characters they support are not useful for most use cases. See [here](assets/pdfs/standard_fonts_demo.pdf) for an example of all 14 standard fonts.
- **You can use characters outside the Latin alphabet** by embedding your own fonts. Embedding your own font requires to you load the font data (from a file or via a network request, for example) and pass it to the `embedFont` method. When you embed your own font, you can use any Unicode characters that it supports. This capability frees you from the limitations imposed by the standard fonts. Most PDF files use embedded fonts. You can embed and use a custom font like so ([see also](#embed-font-and-measure-text)):
  <!-- prettier-ignore -->
  ```js
  import { PDFDocument } from 'pdf-lib'
  import fontkit from '@cantoo/fontkit'

  const url = 'https://pdf-lib.js.org/assets/ubuntu/Ubuntu-R.ttf'
  const fontBytes = await fetch(url).then((res) => res.arrayBuffer())

  const pdfDoc = await PDFDocument.create()

  pdfDoc.registerFontkit(fontkit)
  const ubuntuFont = await pdfDoc.embedFont(fontBytes)

  const page = pdfDoc.addPage()
  page.drawText('Some fancy Unicode text in the ŪЬȕǹƚü font', {
    font: ubuntuFont,
  })
  ```

Note that encoding errors will be thrown if you try to use a character with a font that does not support it. For example, `Ω` is not in the WinAnsi character set. So trying to draw it on a page with the standard Helvetica font will throw the following error:

```
Error: WinAnsi cannot encode "Ω" (0x03a9)
    at Encoding.encodeUnicodeCodePoint
```

### Font Subsetting

Embedding a font in a PDF document will typically increase the file's size. You can reduce the amount a file's size is increased by subsetting the font so that only the necessary characters are embedded. You can subset a font by setting the [`subset` option](https://pdf-lib.js.org/docs/api/interfaces/embedfontoptions#optional-subset) to `true`. For example:

```js
const font = await pdfDoc.embedFont(fontBytes, { subset: true });
```

Note that subsetting does not work for all fonts. See https://github.com/Hopding/pdf-lib/issues/207#issuecomment-537210471 for additional details.

## Creating and Filling Forms

`pdf-lib` can create, fill, and read PDF form fields. The following field types are supported:

- [Buttons](https://pdf-lib.js.org/docs/api/classes/pdfbutton)
- [Check Boxes](https://pdf-lib.js.org/docs/api/classes/pdfcheckbox)
- [Dropdowns](https://pdf-lib.js.org/docs/api/classes/pdfdropdown)
- [Option Lists](https://pdf-lib.js.org/docs/api/classes/pdfoptionlist)
- [Radio Groups](https://pdf-lib.js.org/docs/api/classes/pdfradiogroup)
- [Text Fields](https://pdf-lib.js.org/docs/api/classes/pdftextfield)

See the [form creation](#create-form) and [form filling](#fill-form) usage examples for code samples. Tests 1, 14, 15, 16, and 17 in the [complete examples](#complete-examples) contain working example code for form creation and filling in a variety of different JS environments.

**IMPORTANT:** The default font used to display text in buttons, dropdowns, option lists, and text fields is the standard Helvetica font. This font only supports characters in the latin alphabet (see [Fonts and Unicode](#fonts-and-unicode) for details). This means that if any of these field types are created or modified to contain text outside the latin alphabet (as is often the case), you will need to embed and use a custom font to update the field appearances. Otherwise an error will be thrown (likely when you save the `PDFDocument`).

You can use an embedded font when filling form fields as follows:

```js
import { PDFDocument } from 'pdf-lib';
import fontkit from '@cantoo/fontkit';

// Fetch the PDF with form fields
const formUrl = 'https://pdf-lib.js.org/assets/dod_character.pdf';
const formBytes = await fetch(formUrl).then((res) => res.arrayBuffer());

// Fetch the Ubuntu font
const fontUrl = 'https://pdf-lib.js.org/assets/ubuntu/Ubuntu-R.ttf';
const fontBytes = await fetch(fontUrl).then((res) => res.arrayBuffer());

// Load the PDF with form fields
const pdfDoc = await PDFDocument.load(formBytes);

// Embed the Ubuntu font
pdfDoc.registerFontkit(fontkit);
const ubuntuFont = await pdfDoc.embedFont(fontBytes);

// Get two text fields from the form
const form = pdfDoc.getForm();
const nameField = form.getTextField('CharacterName 2');
const ageField = form.getTextField('Age');

// Fill the text fields with some fancy Unicode characters (outside
// the WinAnsi latin character set)
nameField.setText('Ӎӑȑїõ');
ageField.setText('24 ŷȇȁŗš');

// **Key Step:** Update the field appearances with the Ubuntu font
form.updateFieldAppearances(ubuntuFont);

// Save the PDF with filled form fields
const pdfBytes = await pdfDoc.save();
```

### Handy Methods for Filling, Creating, and Reading Form Fields

Existing form fields can be accessed with the following methods of [`PDFForm`](https://pdf-lib.js.org/docs/api/classes/pdfform):

- [`PDFForm.getButton`](https://pdf-lib.js.org/docs/api/classes/pdfform#getbutton)
- [`PDFForm.getCheckBox`](https://pdf-lib.js.org/docs/api/classes/pdfform#getcheckbox)
- [`PDFForm.getDropdown`](https://pdf-lib.js.org/docs/api/classes/pdfform#getdropdown)
- [`PDFForm.getOptionList`](https://pdf-lib.js.org/docs/api/classes/pdfform#getoptionlist)
- [`PDFForm.getRadioGroup`](https://pdf-lib.js.org/docs/api/classes/pdfform#getradiogroup)
- [`PDFForm.getTextField`](https://pdf-lib.js.org/docs/api/classes/pdfform#gettextfield)

New form fields can be created with the following methods of [`PDFForm`](https://pdf-lib.js.org/docs/api/classes/pdfform):

- [`PDFForm.createButton`](https://pdf-lib.js.org/docs/api/classes/pdfform#createbutton)
- [`PDFForm.createCheckBox`](https://pdf-lib.js.org/docs/api/classes/pdfform#createcheckbox)
- [`PDFForm.createDropdown`](https://pdf-lib.js.org/docs/api/classes/pdfform#createdropdown)
- [`PDFForm.createOptionList`](https://pdf-lib.js.org/docs/api/classes/pdfform#createoptionlist)
- [`PDFForm.createRadioGroup`](https://pdf-lib.js.org/docs/api/classes/pdfform#createradiogroup)
- [`PDFForm.createTextField`](https://pdf-lib.js.org/docs/api/classes/pdfform#createtextfield)

Below are some of the most commonly used methods for reading and filling the aforementioned subclasses of [`PDFField`](https://pdf-lib.js.org/docs/api/classes/pdffield):

- [`PDFCheckBox.check`](https://pdf-lib.js.org/docs/api/classes/pdfcheckbox#check)
- [`PDFCheckBox.uncheck`](https://pdf-lib.js.org/docs/api/classes/pdfcheckbox#uncheck)
- [`PDFCheckBox.isChecked`](https://pdf-lib.js.org/docs/api/classes/pdfcheckbox#ischecked)

---

- [`PDFDropdown.select`](https://pdf-lib.js.org/docs/api/classes/pdfdropdown#select)
- [`PDFDropdown.clear`](https://pdf-lib.js.org/docs/api/classes/pdfdropdown#clear)
- [`PDFDropdown.getSelected`](https://pdf-lib.js.org/docs/api/classes/pdfdropdown#getselected)
- [`PDFDropdown.getOptions`](https://pdf-lib.js.org/docs/api/classes/pdfdropdown#getoptions)
- [`PDFDropdown.addOptions`](https://pdf-lib.js.org/docs/api/classes/pdfdropdown#addoptions)

---

- [`PDFOptionList.select`](https://pdf-lib.js.org/docs/api/classes/pdfoptionlist#select)
- [`PDFOptionList.clear`](https://pdf-lib.js.org/docs/api/classes/pdfoptionlist#clear)
- [`PDFOptionList.getSelected`](https://pdf-lib.js.org/docs/api/classes/pdfoptionlist#getselected)
- [`PDFOptionList.getOptions`](https://pdf-lib.js.org/docs/api/classes/pdfoptionlist#getoptions)
- [`PDFOptionList.addOptions`](https://pdf-lib.js.org/docs/api/classes/pdfoptionlist#addoptions)

---

- [`PDFRadioGroup.select`](https://pdf-lib.js.org/docs/api/classes/pdfradiogroup#select)
- [`PDFRadioGroup.clear`](https://pdf-lib.js.org/docs/api/classes/pdfradiogroup#clear)
- [`PDFRadioGroup.getSelected`](https://pdf-lib.js.org/docs/api/classes/pdfradiogroup#getselected)
- [`PDFRadioGroup.getOptions`](https://pdf-lib.js.org/docs/api/classes/pdfradiogroup#getoptions)
- [`PDFRadioGroup.addOptionToPage`](https://pdf-lib.js.org/docs/api/classes/pdfradiogroup#addoptiontopage)

---

- [`PDFTextField.setText`](https://pdf-lib.js.org/docs/api/classes/pdftextfield#settext)
- [`PDFTextField.getText`](https://pdf-lib.js.org/docs/api/classes/pdftextfield#gettext)
- [`PDFTextField.setMaxLength`](https://pdf-lib.js.org/docs/api/classes/pdftextfield#setmaxlength)
- [`PDFTextField.getMaxLength`](https://pdf-lib.js.org/docs/api/classes/pdftextfield#getmaxlength)
- [`PDFTextField.removeMaxLength`](https://pdf-lib.js.org/docs/api/classes/pdftextfield#removemaxlength)

## Limitations

- `pdf-lib` **can** extract the content of text fields (see [`PDFTextField.getText`](https://pdf-lib.js.org/docs/api/classes/pdftextfield#gettext)), but it **cannot** extract plain text on a page outside of a form field. This is a difficult feature to implement, but it is within the scope of this library and may be added to `pdf-lib` in the future. See
  [#93](https://github.com/Hopding/pdf-lib/issues/93),
  [#137](https://github.com/Hopding/pdf-lib/issues/137),
  [#177](https://github.com/Hopding/pdf-lib/issues/177),
  [#329](https://github.com/Hopding/pdf-lib/issues/329), and
  [#380](https://github.com/Hopding/pdf-lib/issues/380).
- `pdf-lib` **can** remove and edit the content of text fields (see [`PDFTextField.setText`](https://pdf-lib.js.org/docs/api/classes/pdftextfield#settext)), but it does **not** provide APIs for removing or editing text on a page outside of a form field. This is also a difficult feature to implement, but is within the scope of `pdf-lib` and may be added in the future. See
  [#93](https://github.com/Hopding/pdf-lib/issues/93),
  [#137](https://github.com/Hopding/pdf-lib/issues/137),
  [#177](https://github.com/Hopding/pdf-lib/issues/177),
  [#329](https://github.com/Hopding/pdf-lib/issues/329), and
  [#380](https://github.com/Hopding/pdf-lib/issues/380).
- `pdf-lib` does **not** support the use of HTML or CSS when adding content to a PDF. Similarly, `pdf-lib` **cannot** embed HTML/CSS content into PDFs. As convenient as such a feature might be, it would be extremely difficult to implement and is far beyond the scope of this library. If this capability is something you need, consider using [Puppeteer](https://github.com/puppeteer/puppeteer).
- `convertToPDFA()` and `embedFacturX()` add the **structural** PDF/A (and Factur-X XMP / attachment) pieces, but they do **not** rewrite page content for compliance (fonts, transparency, JavaScript, …). `embedFacturX()` also does **not** generate or validate the invoice XML. Always validate archival / e-invoice output with dedicated tools (e.g. veraPDF, Mustang).

## Help and Discussion

[Discussions](https://github.com/Hopding/pdf-lib/discussions) is the best place to chat with us, ask questions, and learn more about pdf-lib!

See also [MAINTAINERSHIP.md#communication](docs/MAINTAINERSHIP.md#communication) and [MAINTAINERSHIP.md#discord](docs/MAINTAINERSHIP.md#discord).

## Encryption Handling

**`pdf-lib` does support encrypted documents.**

If you do not have a password yet, you can load the document without decrypting it to check whether it is encrypted:

```js
// Load without decrypting (encrypted content will not be readable):
const doc = await PDFDocument.load(content, { ignoreEncryption: true });
const isEncrypted = doc.isEncrypted;
// If isEncrypted is true, ask the user for the password.
```

To decrypt and load the document, pass the password. `ignoreEncryption` is not required:

```js
// Load and decrypt an encrypted document:
const doc = await PDFDocument.load(content, { password: 'The password' });
```

An empty password is valid for some PDFs. Pass it explicitly — do not omit the option or treat `''` as "no password":

```js
const doc = await PDFDocument.load(content, { password: '' });
```

If the password is wrong, `PDFDocument.load` throws an error.

To encrypt a document, call `encrypt()` before saving. The default cipher is
**AES-256** (ISO 32000-2 revision 6). The document's PDF version is not used to
pick the algorithm — pass `algorithm` if you must target an older viewer.
RC4 is refused unless you also set `allowWeakCryptography: true`:

```js
const doc = await PDFDocument.create();
doc.addPage().drawText('Hello!');

doc.encrypt({
  userPassword: 'user-pw',
  ownerPassword: 'owner-pw',
  permissions: { printing: 'highResolution' },
  // algorithm: 'AES-128', // optional; default is 'AES-256'
});

const bytes = await doc.save();
```

> **Requirement:** `encrypt()` needs the Web Crypto API
> (`crypto.getRandomValues`) to generate secure random keys, salts, and IVs.
> That is available in modern browsers, **Node >= 18**, Deno, and Bun. On older
> Node versions, or in React Native without a polyfill such as
> [`react-native-get-random-values`](https://www.npmjs.com/package/react-native-get-random-values),
> `encrypt()` throws. No other part of `pdf-lib` requires it — loading,
> decrypting, and saving unencrypted documents all work without it.

## Contributing

We welcome contributions from the open source community! If you are interested in contributing to `pdf-lib`, please take a look at the [CONTRIBUTING.md](docs/CONTRIBUTING.md) file. It contains information to help you get `pdf-lib` setup and running on your machine. (We try to make this as simple and fast as possible! :rocket:)

## Maintainership

Check out [MAINTAINERSHIP.md](docs/MAINTAINERSHIP.md) for details on how this repo is maintained and how we use [issues](docs/MAINTAINERSHIP.md#issues), [PRs](docs/MAINTAINERSHIP.md#pull-requests), and [discussions](docs/MAINTAINERSHIP.md#discussions).

## Tutorials and Cool Stuff

- [labelmake](https://github.com/hand-dot/labelmake) - a library for declarative PDF generation created by @hand-dot
- [Möbius Printing helper](https://shreevatsa.net/mobius-print/) - a tool created by @shreevatsa
- [Extract PDF pages](https://shreevatsa.net/pdf-pages/) - a tool created by @shreevatsa
- [Travel certificate generator](https://github.com/LAB-MI/deplacement-covid-19) - a tool that creates travel certificates for French citizens under quarantine due to COVID-19
- [How to use pdf-lib in AWS Lambdas](https://medium.com/swlh/create-pdf-using-pdf-lib-on-serverless-aws-lambda-e9506246dc88) - a tutorial written by Crespo Wang
- [Working With PDFs in Node.js Using pdf-lib](http://thecodebarbarian.com/working-with-pdfs-in-node-js.html) - a tutorial by Valeri Karpov
- [Electron app for resizing PDFs](https://github.com/vegarringdal/simple-pdf-resizer) - a tool created by @vegarringdal
- [PDF Shelter](https://pdfshelter.com) - online PDF manipulation tools by Lucas Morais

## Prior Art

- [`pdfkit`](https://github.com/devongovett/pdfkit) is a PDF generation library for Node and the Browser. This library was immensely helpful as a reference and existence proof when creating `pdf-lib`. `pdfkit`'s code for [font embedding](src/core/embedders/CustomFontEmbedder.ts#L17-L21), [PNG embedding](src/core/embedders/PngEmbedder.ts#L7-L11), and [JPG embedding](src/core/embedders/JpegEmbedder.ts#L25-L29) was especially useful.
- [`pdf.js`](https://github.com/mozilla/pdf.js) is a PDF rendering library for the Browser. This library was helpful as a reference when writing `pdf-lib`'s parser. Some of the code for stream decoding was [ported directly to TypeScript](src/core/streams) for use in `pdf-lib`.
- [`pdfbox`](https://pdfbox.apache.org/) is a PDF generation and modification library written in Java. This library was an invaluable reference when implementing form creation and filling APIs for `pdf-lib`.
- [`jspdf`](https://github.com/MrRio/jsPDF) is a PDF generation library for the browser.
- [`pdfmake`](https://github.com/bpampuch/pdfmake) is a PDF generation library for the browser.
- [`hummus`](https://github.com/galkahana/HummusJS) is a PDF generation and modification library for Node environments. `hummus` is a Node wrapper around a [C++ library](https://github.com/galkahana/PDF-Writer), so it doesn't work in many JavaScript environments - like the Browser or React Native.
- [`react-native-pdf-lib`](https://github.com/Hopding/react-native-pdf-lib) is a PDF generation and modification library for React Native environments. `react-native-pdf-lib` is a wrapper around [C++](https://github.com/galkahana/PDF-Writer) and [Java](https://github.com/TomRoush/PdfBox-Android) libraries.
- [`pdfassembler`](https://github.com/DevelopingMagic/pdfassembler) is a PDF generation and modification library for Node and the browser. It requires some knowledge about the logical structure of PDF documents to use.

## Git History Rewrite

This repo used to contain a file called `pdf_specification.pdf` in the root directory. This was a copy of the [PDF 1.7 specification](https://www.adobe.com/content/dam/acom/en/devnet/pdf/pdfs/PDF32000_2008.pdf), which is made freely available by Adobe. On 8/30/2021, we received a DMCA complaint requiring us to remove the file from this repo. Simply removing the file via a new commit to `master` was insufficient to satisfy the complaint. The file needed to be completely removed from the repo's git history. Unfortunately, the file was added over two years ago, this meant we had to rewrite the repo's git history and force push to `master` 😔.

### Steps We Took

We removed the file and rewrote the repo's history using [BFG Repo-Cleaner](https://rtyley.github.io/bfg-repo-cleaner/) as outlined [here](https://docs.github.com/en/github/authenticating-to-github/keeping-your-account-and-data-secure/removing-sensitive-data-from-a-repository). For full transparency, here are the exact commands we ran:

```
$ git clone git@github.com:Hopding/pdf-lib.git
$ cd pdf-lib
$ rm pdf_specification.pdf
$ git commit -am 'Remove pdf_specification.pdf'
$ bfg --delete-files pdf_specification.pdf
$ git reflog expire --expire=now --all && git gc --prune=now --aggressive
$ git push --force
```

### Why Should I Care?

If you're a user of `pdf-lib`, you shouldn't care! Just keep on using `pdf-lib` like normal 😃 ✨!

If you are a `pdf-lib` developer (meaning you've forked `pdf-lib` and/or have an open PR) then this does impact you. If you forked or cloned the repo prior to 8/30/2021 then your fork's git history is out of sync with this repo's `master` branch. Unfortunately, this will likely be a headache for you to deal with. Sorry! We didn't want to rewrite the history, but there really was no alternative.

It's important to note that pdf-lib's _source code_ has not changed at all. It's exactly the same as it was before the git history rewrite. The repo still has the exact same number of commits (and even the same commit contents, except for the commit that added `pdf_specification.pdf`). What has changed are the SHAs of those commits.

The simplest way to deal with this fact is to:

1. Reclone pdf-lib
2. Manually copy any changes you've made from your old clone to the new one
3. Use your new clone going forward
4. Reopen your unmerged PRs using your new clone

See this [StackOverflow answer](https://stackoverflow.com/a/48268766) for a great, in depth explanation of what a git history rewrite entails.

## License

[MIT](LICENSE.md)

---
_Source: https://npm.io/package/@cantoo/pdf-lib · Machine-readable twin of the npm.io package page. Health data is recomputed on every publish._