VS Code Dev Containers allows you to run your code and environment in a container. This is especially useful for Swift when developing on macOS and deploying to Linux. You can ensure there are no compatibility issues in Foundation when running your code. The extension also works with GitHub Codespaces to allow you to write your code on the web.
As well as installing the Swift extension, you must install Docker on your machine to run the dev container in. See the Visual Studio Code documentation for more details.
Next, install the Remote Development extension pack that contains extensions for working in remote environments in VS Code. If you only want to work with dev containers (and not use the SSH or WSL containers), you may want to only install the Dev Containers extension instead.
VS Code requires a .devcontainer
directory which defines the settings in devcontainer.json
and optionally a Dockerfile
defining the container to run in.
First create the directory. Next, create devcontainer.json
and insert the following:
{
"name": "Swift 5.5",
"image": "swift:5.5",
"extensions": [
"sswg.swift-lang"
],
"settings": {
"lldb.library": "/usr/lib/liblldb.so"
},
"forwardPorts": [8080]
}
This defines the minimum settings required to run a Swift package in a dev container. Here's what each thing does:
name
: Used to specify the name of the dev container.image
: The Docker container image to run. You can choose whichever version of Swift you like, including nightlies.extensions
: Extensions to install in your dev environment. You do not need to specify extensions' dependencies, such as LLDB.settings
: Override any settings for extensions. The above example sets the LLDB path to stop the Swift extension from attempting to set it up.forwardPorts
: Ports to enable forwarding for. You may want to include this if building a Swift server application for example.
That's all you need to get a dev container working!
You may want to use a custom Docker container that's version controlled to make it easy to set up a development environment for you team. In devcontainer.json
replace image
with the following:
{
"build": { "dockerfile": "Dockerfile" },
// ...
}
This will use the Dockerfile
provided in .devcontainer
. Create that file and insert your custom Dockerfile. For example:
FROM swift:5.5
For more complex development environments you may need to use Docker Compose. The devcontainer.json
file has three settings you need to include if you want to use Docker Compose:
dockerComposeFile
your docker compose fileservice
the service you want to runworkspaceFolder
the root folder for your project
Your devcontainer.json
should look something like this
{
"name": "MyService: 5.6-focal",
"dockerComposeFile": "docker-compose.yml",
"service": "app",
"workspaceFolder": "/workspace",
"extensions": [
"sswg.swift-lang",
],
"settings": {
"lldb.library": "/usr/lib/liblldb.so"
},
}
Below is an example of a docker-compose.yml
file that brings up a redis server
version: "3.3"
services:
app:
image: swift:5.6-focal
volumes:
- ..:/workspace
depends_on:
- redis
environment:
- REDIS_HOST=redis
command: sleep infinity
redis:
image: redis
ports:
- "6379:6379"
Note the service
and workspace
variables from the devcontainer.json
reference the service app
and workspace volume in the docker-compose.yml
. The app
service is required to run all the time you are using the devcontainer. We do this here by using the command sleep infinity
. It is dependent on the redis
service which can be referenced in code using the service name redis
. I have also included a REDIS_HOST
environment variable which can be used in your code.
VS Code allows you to automatically configure your project with a dev container. In the command palette (F1
) choose Dev Containers: Add Development Container Configuration Files... and choose Swift.
Once you've set up your .devcontainer
, in the command palette run Dev Containers: Reopen in Container. VS Code will relaunch running in your dev container!
For more details about running your project in a dev container, and the available configuration options, see the Visual Studio Code documentation.