Local preferences

Architect preferences (preferences.arc, or prefs.arc) defines settings for local Architect workflows. This file is intended to be added to .gitignore.

  • @create - Preferences for resource creation with arc init
  • @env - Configure environment variables
  • @sandbox - Define Sandbox preferences
  • @sandbox-start - Hook into Sandbox’s startup

@create

Preferences for resource creation with arc init.

autocreate

By adding the @create pragma to your preferences file and specifying autocreate true, you can enable arc sandbox, arc deploy, and other workflows to automatically run arc init to create boilerplate Lambda handlers and static assets if they do not exist.

@create
autocreate true

templates

Define custom boilerplate Lambda handlers on a per-pragma basis with templates:

@create
templates
  http path/to/template/http.js
  events path/to/template/events.py

In the above example, new @http functions will use your path/to/template/http.js template instead of the Architect default, while creating new @events functions will use the path/to/template/events.py. This will work for either autocreate true or the arc init command.

@env

Configure environment variables for testing with Sandbox and deployed staging and production environments.

Sync environment variables to your project by using the arc env CLI command. If the preferences file does not exist Architect will generate preferences.arc file.

Note: any time you run arc env, your unsynced local environment variables will be overwritten.

Example

@env
testing
  A_TESTING_ENV_VAR something-for-testing
  ANOTHER_VAR only-for-testing

staging
  A_STAGING_ENV_VAR something-for-staging

production
  A_PRODUCTION_ENV_VAR something-for-production

.env file support

Architect Sandbox supports loading environment variables from a .env file. The .env will override your preferences.arc or prefs.arc, and environment variables it defines are only loaded for the testing environment. If you require locally configured env vars for staging or production environments, you must use pref[erence]s.arc.

Note: as a friendly reminder, key / value pairs in .env files are separated by the = symbol.

Example .env file

A_TESTING_ENV_VAR=something-for-testing
ANOTHER_VAR=only-for-testing

@sandbox

Define Sandbox preferences. If you are not using a .env file then any environment variables set using the arc env CLI will be stored in the preferences file. In this scenario it is best not to revision the preferences file in source control.

livereload - Boolean

Enable automatic reload for HTML views when @http Lambda (get or any), @shared, or @views code changes. livereload is helpful when developing view layouts and styling.

@sandbox
livereload true

Note: livereload will execute your @http handler with each change so long as it is a get or any path. Traditionally, these routes don’t create data, but be mindful of how a reload might interact with your app’s data layer before enabling.

ports - List

Designate the local ports used by Sandbox services. Sandbox will scan for and use available ports unless specified. If a specified port is unavailable, Sandbox will fail to boot.

@sandbox
ports
  http 4200
  events 4211
  queues 4222
  tables 4255

env - String

Advanced option: set the ARC_ENV stage to staging or production and use the env vars for that stage (see the @env pragma above); if not specified, defaults to testing. This setting may introduce unexpected side effects, so only use it if you have a specific technical reason.

@sandbox
env staging

useAWS - Boolean

Advanced option that uses live AWS infrastructure where deployed, specifically: @tables / @tables-indexes (DynamoDB), @events (EventBridge), and @queues (SQS). Notes:

  • To use this feature, your local AWS credentials file must have valid keys to use this infrastructure (or calls to AWS will fail)
  • If you do not specify an environment, staging will be set automatically; you can also use production
@sandbox
useAWS true

no-hydrate - Boolean

Disables hydration

@sandbox
no-hydrate true

@sandbox-start

Hook up CLI commands into Sandbox startup. Helpful for repetitive tasks like seeding a database or starting up additional services for local development. Each command should be a separate unindented line under the @sandbox-start pragma.

Example

@sandbox-start
node scripts/seed_db.js
echo 'hello'

Note: the older alias @sandbox-startup still works, but will be deprecated in a future release. Please use @sandbox-start