This repository contains helper/wrapper scripts to make building Electron easier.

A handful of prerequisites, such as git, python, and npm, are required for building Electron itself; these can be found in Platform Prerequisites. npm can be used with build-tools itself as well, but we've configured it to run with yarn, so we also recommend you install it to your system.

From here, you'll need a command-line prompt. On Mac and Linux, this will be a terminal with a shell, e.g. bash or zsh. You can also use these on Windows if you install them, or use built-in tools like Windows' Command Prompt.

Please note that build-tools (due to nested dependencies) might not work properly in powershell, please use cmd on Windows for optimum results.

# Install build-tools package globally:
npm i -g @electron/build-tools

You can run a new Electron build with this command:

# The 'Hello, World!' of build-tools: get and build `main`
# Choose the directory where Electron's source and build files will reside.
# You can specify any path you like; this command defaults to `$PWD/electron`.
# If you're going to use multiple branches, you may want something like:
# `--root=~/electron/branch` (e.g. `~/electron-gn/main`)
e init --root=~/electron --bootstrap testing

That command's going to run for awhile. While you're waiting, grab a cup of hot caffeine and read about what your computer is doing:

Electron's build-tools command is named e. Like nvm and git, you'll invoke e with commands and subcommands. See e --help or e help <cmd> for many more details.

e also borrows another inspiration from nvm: having multiple configurations that you can switch between so that one is the current, active configuration. Many choices go into an Electron build:

• Which Electron branch is used (e.g. main, 13-x-y)
• Which .gn config file is imported (e.g. testing or release)
• Any compile-time options (e.g. Clang's asan or tsan)

e holds all these variables together in a build configuration. You can have multiple build configurations and manage them in a way similar to nvm:

Getting the source code is a lot more than cloning electron/electron. Electron is built on top of Chromium (with Electron patches) and Node (with more Electron patches). A source tree needs to have all of the above and for their versions to be in sync with each other. Electron uses Chromium's Depot Tools and GN for wrangling and building the code. e wraps these tools:

e init initializes a new local development environment for Electron.

To see all potential options for this command, run:

$ e init --help

New build configs are created with e init. It has several command-line options to specify the build configuration, e.g. the path to the source code, compile-time options, and so on. See e init --help for in-depth details.

Each build config has a name, chosen by you to use as a mnemonic when switching between build configs with e use <name>. This is the name's only purpose, so choose whatever you find easiest to work with — whether it's electron, 6-1-x--testing, or chocolate-onion-popsicle.

Each build also needs a root directory. All the source code and built files will be stored somewhere beneath it. e init uses $PWD/electron by default, but you can choose your own with --root=/some/path. If you want to make multiple build types of the same branch, you can reuse an existing root to share it between build configs.

As an example, let's say you're starting from scratch and want both testing and release builds of the main branch in electron/electron. You might do this:

# making 'release' and 'testing' builds from main

$ e init main-testing -i testing --root=~/src/electron
Creating '~/src/electron'
New build config 'main-testing' created
Now using config 'main-testing'
$ e show current
main-testing

$ e init main-release -i release --root=~/src/electron
INFO Root '~/src/electron' already exists.
INFO (OK if you are sharing $root between multiple build configs)
New build config 'main-release' created
Now using config 'main-release'

$ e show configs
* main-release
  main-testing

$ e show current
main-release
$ e show root
~/src/electron

$ e use main-testing
Now using config 'main-testing'
$ e show current
main-testing
$ e show root
~/src/electron

As a convenience, e init --bootstrap will run e sync and e build after creating the build config. Let's see what those do:

To see all potential options for this command, run:

$ e sync --help

e sync is a wrapper around gclient sync from Depot Tools. If you're starting from scratch, this will (slowly) fetch all the source code. It's also useful after switching Electron branches to synchronize the rest of the sources to the versions needed by the new Electron branch.

e sync is usually all you need. Any extra args are passed along to gclient itself.

$ e show current
main-testing

$ e show root
~/src/electron

$ e sync
Running "gclient sync --with_branch_heads --with_tags" in '~/src/electron/src'
[sync output omitted]

To make your output more verbose, you can add an increasing number of -vs. For example,

# basic verbosity
$ e sync -v
Running "gclient sync --with_branch_heads --with_tags -v" in '~/src/electron/src'
[sync output omitted]

# significant verbosity
$ e sync -vvvv
Running "gclient sync --with_branch_heads --with_tags -vvvv" in '~/src/electron/src'
[sync output omitted]

e build builds an Electron executable.

To see all potential options for this command, run:

$ e build --help

Once you have the source, the next step is to build it with e build [target]. Some of the build targets you may want to build include:

To build a specific ninja target, run e build --target [target]:

$ e build --target electron:node_headers

Running e build with no target will build electron by default.

Any extra args are passed along to ninja, so for example e build -v runs a verbose build of electron.

To see an exhaustive list of all possible build targets, you can run e d gn ls out/[TYPE], where [TYPE] is e.g. Testing or Release depending on your build type. This will log a long list of targets to the console and also allow you to build some of Chromium's targets.

For example, running e d gn ls out/Testing | grep "//ui/views/" produces something like:

//ui/views/controls/webview:test_support
//ui/views/controls/webview:webview
//ui/views/debug:views_debug
//ui/views/examples:copy_content_resources
//ui/views/examples:views_examples
//ui/views/examples:views_examples_lib
//ui/views/examples:views_examples_proc
//ui/views/examples:views_examples_resources_grd
//ui/views/examples:views_examples_resources_grd_grit
//ui/views/examples:views_examples_resources_pak
//ui/views/examples:views_examples_unittests
//ui/views/examples:views_examples_unittests__runner
//ui/views/examples:views_examples_with_content
//ui/views/examples:views_examples_with_content_lib
//ui/views/resources:resources
//ui/views/resources:resources_grd
//ui/views/resources:resources_grd_grit
//ui/views/window/vector_icons:vector_icons
//ui/views/window/vector_icons:window_control_vector_icons

You could then run e build --target ui/views/examples:views_examples_with_content to produce Chrome's //ui/views example executable and run it with ./out/Testing/views_examples_with_content.

After you've built Electron, it's time to use it!

As usual, any extra args are passed along to the executable. For example, e node --version will print out Electron's node version.

e debug runs your local Electron build inside of lldb or gdb.

$ uname
Linux
$ e debug
Reading symbols from /home/yourname/electron/gn/main/src/out/Testing/electron...
(gdb)

$ uname
Darwin

$ e debug
target create "/Users/yourname/electron-gn/src/out/Testing/Electron.app/Contents/MacOS/Electron"
(lldb)

e d runs Depot Tools commands. Some useful commands:

# run gclient sync directly
e d gclient sync

# login to reclient
e d rbe login

# check reclient status
e d rbe status

# get the list of all build args 
e d gn args --list -C out/Testing

e test starts the local Electron build's test runner. Any extra args are passed along to the runner.

To see all potential options for this command, run:

$ e test --help

Example:

# run all tests
e test

# run main process tests
e test --runners=main

Possible extra arguments to pass:

• --node - Run Node.js' own tests with Electron in RUN_AS_NODE mode.
• --runners=<main|native> - The set of tests to run, can be either main or native.

e show shows information about the current build config.

To see all potential options for this command, run:

$ e show --help

Example usage:

$ uname
Darwin

$ e show exe
/Users/username/electron-gn-root/src/out/Testing/Electron.app/Contents/MacOS/Electron

$ uname
Linux

$ e show exe
/home/username/electron-gn-root/src/out/Testing/electron

$ e show out
Testing

$ e show src
/home/username/electron-gn-root/src/electron

$ cd `e show src base` && pwd
/home/username/electron-gn-root/src/base

$ ripgrep --t h TakeHeapSnapshot `e show src`

e remove|rm <name> removes a build config from the list.

e open opens the GitHub page for the specified commit, pull request, or issue.

To see all potential options for this command, run:

$ e open --help

For example, e open 0920d01 will find the commit with an abbreviated sha1 of 0920d01, see that it's associated with pull request #23450, and open electron/electron#23450 in your browser. Since you can pass in a pull request or issue number as well, e open 23450 would have the same effect.

e patches exports patches to the specified patch directory in Electron source tree.

To see all potential options for this command, run:

$ e patches --help

Valid patch directories can include:

• node
• v8
• boringssl
• chromium
• perfetto
• icu

e sanitize-config updates and/or overwrites an existing config to conform to latest build-tools updates.

To see all potential options for this command, run:

$ e sanitize-config --help

Sometimes build-tools will make updates to its config requirements. In these events warnings will be output to console to inform you that build-tools has temporarily handled the issues. You can make these warnings go away either by manually updating your config files or by running this command to automatically overwrite the existing configs to update formatting.

e shell launches a shell environment populated with build-tools' environment variables and context.

Developers may for example find themselves wishing to access gn and ninja directly - instead of using e d gn to access that this command can be used. This also enables bash completion and the ability to copy-paste commands from Chromium docs without modification.

$ e shell
# Launching build-tools shell with "/bin/zsh"
# Running "/bin/zsh"

e backport <PR> assists with manual backport processes on the specified pull request.

To see all potential options for this command, run:

$ e backport --help

Example:

# select the PR to backport
e backport 1234

# select branch you want to backport PR to
30-x-y

# resolve any merge conflicts
git cherry-pick --continue

# push changes
git push

# create pull request
e pr --backport 1234

e init checks out the HEAD of the main branch. To build against a specific version of Electron, checkout that version with these commands:

# Change working directory to the Electron source directory
cd `e show src`

# Checkout the desired Electron version (in this case, 11.0.0)
git checkout tags/v11.0.0 -b v11.0.0

# Sync dependencies with the current branch
e sync

# Build Electron
e build

If you want your shell sessions to each have different active configs, try this in your ~/.profile or ~/.zshrc or ~/.bashrc:

export EVM_CURRENT_FILE="$(mktemp --tmpdir evm-current.XXXXXXXX.txt)"

This will create per-shell temporary files in which the active config file can be changed with e use.

With the default configuration, build-tools will automatically check for updates every 4 hours.

You can enable and disable these automatic updates with the following commands:

$ e auto-update enable
$ e auto-update disable

Regardless of whether automatic updates are enabled, you can manually call the following command to immediately trigger an update.

$ e auto-update check