This project aims to compile the statistical language R (https://www.r-project.org/) into WebAssembly for use with a browser, via Emscripten (https://emscripten.org/). The repo includes patches to R's source code so that it can work in the browser environment provided by Emscripten and also includes a web-based IDE through the use of xterm.js and CodeMirror.
A demo of the webR IDE can be found at https://webr.r-wasm.org/latest/. Please be patient as the Wasm runtime downloads and executes. R will display a banner message when it is ready to use.
Documentation showing how to use webR in your own projects can be found at https://docs.r-wasm.org/webr/latest/
The webR JavaScript package is available for download through npm and on CDN.
Complete release packages, including R WebAssembly binaries, are available to download for self hosting in the GitHub Releases section.
Docker images containing a pre-built webR development environment can be found in the GitHub Packages section.
R's source code and supporting libraries are written in both C/C++ and Fortran. Source files can be compiled with either a custom development version of LLVM flang (the default) or with gfortran and Dragonegg (using the --with-dragonegg configure option).
If you are compiling webR using the default toolchain, ensure that you first install the following required prerequisites in your build environment:
- Emscripten SDK (>=3.1.35)
- cmake
- gperf
- liblzma
- libpcre2
- node (>=17.0.0)
- quilt
- wget
Clone the repo into a new directory, cd into the directory, then run ./configure && make. You can configure make variables in a ~/.webr-config.mk file.
A dist directory is created which when finished contains the R Wasm files and an index.html file ready to serve the included webR IDE.
WebR relies on additional libraries compiled for Wasm for both Cairo graphics support and for building R packages that depend on certain system libraries. By default, only a minimal set of libraries are built for use with webR.
If you'd prefer to build all of the available system libraries for Wasm, cd into the libs directory and run make all to build the additional libraries, then finally cd .. and run make clean-webr && make to rebuild webR. R will automatically detect the additional Wasm libraries and integrate Cairo graphics support as part of the build.
Included in the source repository is a Dockerfile which can be used to setup everything that's required for the webR build environment, including LLVM flang and all supported WebAssembly system libraries.
Pre-built docker images can be found in the GitHub Packages section. To build the docker image and webR from source, cd into the webR source directory then run docker build .
The resulting docker image contains a version of R configured to be built for WebAssembly, and so the image can also be used to build custom R packages for webR.
If you are using Nix, you can start a development environment by running nix develop. Then you can build as usual, with ./configure && make.
Note that this requires that your Nix installation has support for flakes enabled. The easiest way to do this is to install using the Nix installer from Determinate Systems.
WebR requires compiler and runtime support for WebAssembly.Exception, used internally for R error handling. This requires a version of Emscripten >= 3.1.35 and Node >= 17.0.0, which may be newer than the versions provided by your system package manager. An easy way to install and manage multiple versions of Node and Emscripten is by using nvm and emsdk.
The version of Node currently bundled by emsdk is 16.0.0. When building webR with this version of Node the process will fail with configure logs containing the error
WebAssembly.Tag is not a constructor
If this occurs, a newer version of Node should be installed and the following environment variable set before building webR, instructing Emscripten to use the newer version of Node:
export EM_NODE_JS=$(HOME)/.nvm/versions/node/v20.1.0/bin/node
If you are unsure of the correct path to Node the command which node should print the path in full.
At the time of writing the version of R used as the base for webR does not build cleanly using the macOS Ventura development SDK. If you are not using the included Dockerfile to build webR, the following extra setup must be done before starting the build process,
- Install the GNU version of the patch program: e.g.
brew install gpatch
If you intend to build webR using Dragonegg to handle Fortran sources, older versions of the gcc/gfortran toolchain are required than what is provided by modern operating system repositories. The docker file tools/dragonegg/Dockerfile can be used to setup the required C/C++/Fortran toolchain and development environment for compiling webR with Dragonegg.