If you want to work with Paper.js, simply download the latest "stable" version from http://paperjs.org/download/
- Website: http://paperjs.org/
- Questions: https://stackoverflow.com/questions/tagged/paperjs
- Discussion forum: https://groups.google.com/group/paperjs
- Mainline source code: https://github.com/paperjs/paper.js
- Twitter: @paperjs
- Latest releases: http://paperjs.org/download/
- Pre-built development versions:
prebuilt/moduleandprebuilt/distbranches.
The recommended way to install and maintain Paper.js as a dependency in your project is through the Node.js Package Manager (NPM) for browsers, Node.js or Electron.
If NPM is already installed, simply type one of these commands in your project folder:
npm install paperUpon execution, you will find a paper folder inside the project's
node_modules folder.
For more information on how to install Node.js and NPM, read the chapter Installing Node.js and NPM.
With a bundler such as webpack, Rollup, esbuild or Vite, import Paper.js and set it up with a canvas element:
import paper from 'paper'
// Or, with CommonJS: const paper = require('paper')
paper.setup(document.getElementById('canvas'))
const { Path } = paper
const circle = new Path.Circle({
center: paper.view.center,
radius: 30,
fillColor: 'red'
})
paper.view.onFrame = event => {
circle.rotate(1)
}The module exports the global paper scope as its default export, with all
classes as its properties. Use paper.view and paper.project after calling
paper.setup(), rather than importing them, as they only exist from then on.
To use the core version without PaperScript, import 'paper/dist/paper-core'
instead. With TypeScript, the default import requires esModuleInterop.
The various distributions come with two different pre-build versions of Paper.js, in minified and normal variants:
paper-full.js– The full version for the browser, including PaperScript support and Acorn.jspaper-core.js– The core version for the browser, without PaperScript support nor Acorn.js. You can use this to shave off some bytes and compilation time when working with JavaScript directly.
Node.js comes with the Node Package Manager (NPM). There are many tutorials explaining the different ways to install Node.js on different platforms. It is generally not recommended to install Node.js through OS-supplied package managers, as the its development cycles move fast and these versions are often out-of-date.
On macOS, Homebrew is a good option if one version of
Node.js that is kept up to date with brew upgrade is enough:
https://treehouse.github.io/installation-guides/mac/node-mac.html
NVM can be used instead to install and
maintain multiple versions of Node.js on the same platform, as often required by
different projects:
https://nodesource.com/blog/installing-node-js-tutorial-using-nvm-on-mac-os-x-and-ubuntu/
Homebrew is recommended on macOS also if you intend to install Paper.js with rendering to the Canvas on Node.js, as described in the next paragraph.
For Linux, see https://nodejs.org/download/ to locate 32-bit and 64-bit Node.js binaries as well as sources, or use NVM, as described in the paragraph above.
Paper.js comes in three different versions on NPM: paper, paper-jsdom and
paper-jsdom-canvas. Depending on your use case, you need to required a
different one:
paperis the main library, and can be used directly in a browser context, e.g. a web browser or worker.paper-jsdomis a shim module for Node.js, offering headless use with SVG importing and exporting through jsdom.paper-jsdom-canvasis a shim module for Node.js, offering canvas rendering through Node-Canvas as well as SVG importing and exporting through jsdom.
In order to install paper-jsdom-canvas, you need the Cairo Graphics
library installed in your system:
Paper.js relies on Node-Canvas for rendering, which in turn relies on the native libraries Cairo and Pango.
Paper.js relies on Node-Canvas for rendering, which in turn relies on Cairo and Pango. The easiest way to install Cairo is through Homebrew, by issuing the command:
brew install cairo pango
Note that currently there is an issue on macOS with Cairo. If the above causes errors, the following will most likely fix it:
PKG_CONFIG_PATH=/opt/X11/lib/pkgconfig/ npm install paper
Also, whenever you would like to update the modules, you will need to execute:
PKG_CONFIG_PATH=/opt/X11/lib/pkgconfig/ npm update
If you keep forgetting about this requirement, or would like to be able to type
simple and clean commands, add this to your .bash_profile file:
# PKG Config for Pango / Cairo
export PKG_CONFIG_PATH=/usr/local/lib/pkgconfig:/opt/X11/lib/pkgconfig
After adding this line, your commands should work in the expected way:
npm install paper
npm update
sudo apt-get install pkg-config libcairo2-dev libpango1.0-dev libssl-dev libjpeg62-dev libgif-dev
You might also need to install the build-essential package if you don't usually build from c++ sources:
sudo apt-get install build-essential
In order to build Node-Canvas for use of paper-jsdom-canvas in Electron, which
is likely to use a different version of V8 than the Node binary installed in
your system, you need to manually specify the location of Electron’s headers.
Follow these steps to do so:
Electron — Using Native Node Modules
You should now be able to install the Paper.js module with jsdom and Canvas rendering from NPM:
npm install paper-jsdom-canvas
The main Paper.js source tree is hosted on
GitHub. git is required to create a
clone of the repository, and can be easily installed through your preferred
package manager on your platform.
git clone --recursive git://github.com/paperjs/paper.js.git
cd paper.js
To refresh your clone and fetch changes from origin, run:
git fetch origin
To update the jsdoc-toolkit submodule, used to generate the documentation,
run:
git submodule update --init --recursive
Paper.js uses Gulp.js for building, and has a couple of dependencies as NPM modules. Read the chapter Installing Node.js and NPM if you still need to install these.
Development uses pnpm, which needs to be installed too. See Installing pnpm for details.
In order to be able to build Paper.js, after checking out the repository, paper has dependencies that need to be installed. Install them by issuing the following commands from the Paper.js directory:
pnpm install
The Paper.js sources are distributed across many separate files, organised in
subfolders inside the src folder. To compile them all into distributable
files, you can run the build task:
pnpm build
You will then find the built library files inside the dist folder, named
paper-full.js and paper-core.js, along with their minified versions. Read
more about this in Which Version to Use?.
As a handy alternative to building the library after each change to try it out
in your scripts, there is the load task, that replaces the built libraries
with symbolic links to the scrc/load.js script. This script then load the
library directly from all the separate source files in the src folder, through
the Prepro.js JavaScript preprocessing
library.
This means you can switch between loading from sources and loading a built library simply by running.
pnpm load
And to go back to a built library
pnpm build
Note that your PaperScripts examples do not need to change, they can keep
loading dist/paper-full.js, which will always do the right thing. Note also
that src/load.js handles both browsers and Node.js, as supported by Prepro.js.
Create a final zipped distribution file inside the dist folder:
pnpm dist
Since the release of version 0.9.22, Paper.js has adopted aspects of the Git-
Flow workflow. All development is taking place in the
develop branch, which is
only merged into main
when a new release occurs.
As of version 0.9.26, the dist folder is excluded on all branches, and the
building is now part of the pnpm publish process by way of the prepublish
script.
We also offer prebuilt versions of the latest state of the develop branch on
prebuilt/module
and prebuilt/dist.
Similarly to building the library, you can run the docs task to build the
documentation:
pnpm docs
Your docs will then be located at dist/docs.
Paper.js was developed and tested from day 1 using proper unit testing through
jQuery's Qunit. To run the tests after any
change to the library's source, simply open index.html inside the test
folder in your web browser. There should be a green bar at the top, meaning all
tests have passed. If the bar is red, some tests have not passed. These will be
highlighted and become visible when scrolling down.
If you are testing on Chrome, some of the tests will fail due to the browser's CORS restrictions. In order to run the browser based tests on Chrome, you need to run a local web-server through Gulp.js. The following command will handle it for you, and will also open the browser at the right address straight away:
pnpm test:browser
To run them in headless Chrome instead, as CI does, and report the result in the terminal:
pnpm test:chrome
To test the Node.js version of Paper.js, use this command:
pnpm test
The main Paper.js source tree is hosted on GitHub, thus you should create a fork of the repository in which you perform development. See https://help.github.com/articles/fork-a-repo/.
We prefer that you send a pull request on GitHub which will then be merged into the official main line repository.
Also, in your first contribution, add yourself to the end of AUTHORS.md (which
of course is optional).
In addition to contributing code you can also triage issues which may include reproducing bug reports or asking for vital information, such as version numbers or reproduction instructions. If you would like to start triaging issues, one easy way to get started is to subscribe to paper.js on CodeTriage.
Get the source (for contributing):
If you want to contribute to the project you will have to make a fork. Then do this:
git clone --recursive git@github.com:yourusername/paper.js.git
cd paper.js
git remote add upstream git@github.com:paperjs/paper.js.git
To then fetch changes from upstream, run
git fetch upstream
As mentioned above, we prefer that you send a pull request on GitHub:
-
Create a fork of the upstream repository by visiting https://github.com/paperjs/paper.js/fork. If you feel insecure, here's a great guide: https://help.github.com/articles/fork-a-repo/
-
Clone of your repository:
git clone https://yourusername@github.com/yourusername/paper.js.git -
This is important: Create a so-called topic branch based on the
developbranch:git checkout -tb name-of-my-patch developwherename-of-my-patchis a short but descriptive name of the patch you're about to create. Don't worry about the perfect name though -- you can change this name at any time later on. -
Hack! Make your changes, additions, etc., commit them then push them to your GitHub fork:
git push origin name-of-my-patch -
Send a pull request to the upstream repository's owner by visiting your repository's site at GitHub (i.e. https://github.com/yourusername/paper.js) and press the "Pull Request" button. Make sure you are creating the pull request to the
developbranch, not themainbranch. Here's a good guide on pull requests: https://help.github.com/articles/about-pull-requests/
Don't mix different kinds of patches in the same branch. Instead, merge them all
together into your develop branch (or develop everything in your develop
branch and then cherry-pick-and-merge into the different topic branches). Git
provides for an extremely flexible workflow, which in many ways causes more
confusion than it helps you when new to collaborative software development. The
guides provided by GitHub at https://help.github.com/ are a really good
starting point and reference. If you are fixing an issue, a convenient way to
name the branch is to use the issue number as a prefix, like this: git checkout -tb issue-937-feature-add-text-styling.
For a list of authors and contributors, please see AUTHORS.
Distributed under the MIT license. See LICENSE for details.