⚡️ Serverless Export Env Plugin

About

This Serverless plugin exports the environment variables defined in serverless.yml into a distinct .env file. This allows you to access these environment variables from local scripts such as for integration tests. You will find the .env file in the root folder of your project.

It will collect the global environment variables of the poject as well as all environment variables of the functions. It will also add API_ENDPOINT and IS_OFFLINE to your environment if you run the plugin via serverless offline.

Environment variables referencing CloudFormation resources (e.g. Ref: MyDynamoDbTable), or import values (e.g. Fn::ImportValue: MyExportedValue) are automatically resolved to their respective values. This, however, requires the stack to be deployed before the plugin can access any of these variables.

This plugin is based on the serverless-dotenv Plugin by Jimdo but largely rewritten to fit our needs.

Why another plugin?

There're plenty of environment and dotenv plugins available for Serverless. However, some are already obsolete, others are very limited in use case. We needed a possibility to access Serverless environment variables from command line during integration testing of our code. As some of these environment variables are referencing CloudFormation resources, none of the existing plugins was able to solve this.

Referencing CloudFormation resources

Serverless offers a very powerful feature: You are able to reference AWS resources anywhere from within your .yaml and it will automatically resolve them to their respective values during deployment. A common example is to bind a DynamoDB table name to an environment variable, so you can access it in your Lambda function implementation later:

provider:
  environment:
    TABLE_NAME:
      Ref: MyDynamoDbTable
# ...
resources:
  Resources:
    MyDynamoDbTable:
      Type: AWS::DynamoDB::Table
      DeletionPolicy: Retain
      Properties:
        # ...

Later in your code you can simply access process.env.TABLE_NAME to get the proper DynamoDB table name without having to hardcode anything.

require("dotenv").config({ path: "../.env" /* path to your project root folder */ });

const AWS = require('aws-sdk');
const docClient = new AWS.DynamoDB.DocumentClient({ /* ... */ });
docClient.get({
	TableName: process.env.TABLE_NAME,
	Key: { foo: "bar" }
}, result => {
	console.log(result);
});

The Serverless Export Env Plugin supports references to resources created within the serverless.yml, to resources imported from another stack via Fn::ImportValue, pseudo parameters such as AWS::Region and AWS::AccountId as well as the commonly used Fn::Join intrinsic function.

The plugin allows you to make use of these references (and all other environment variables of course) from the command line by exporting them into a .env file in your project folder. Then use a library such as dotenv to read them during runtime.

Usage

Add the npm package to your project:

# Via yarn
$ yarn add arabold/serverless-export-env --dev

# Via npm
$ npm install arabold/serverless-export-env --save-dev

Add the plugin to your serverless.yml:

plugins:
  - serverless-export-env

That's it! You can now call serverless export-env in your terminal to generate the .env file based on your Serverless configuration. Alternative you can just start serverless invoke local -f FUNCTION or serverless offline to generate it.

Provided lifecycle events

• export-env:collect - Collect environment variables from Serverless
• export-env:resolve - Resolve CloudFormation references and import variables
• export-env:apply - Set environment variables when testing Lambda functions locally
• export-env:write - Write environment variables to file

Example

serverless export-env

This example will export all environment variables into a .env file in your project root folder.

Releases

1.2.0

• Use operating system-specific end-of-line when creating .env file

1.1.3

• Fixed an issue with AWS::AccountId being resolved as [Object Promise] instead of the actual value.

1.1.2

• Fixed an issue with CloudFormation resources not being resolved properly if the stack has more than 100 resources or exports.

1.1.1

• Fix issue with multiple environment variables for function (thanks to @Nevon).

1.1.0

• Support Fn::Join operation (contribution by @jonasho)
• Support pseudo parameters AWS::Region, AWS::AccountId, AWS::StackId and AWS::StackName.

1.0.2

• The plugin now properly resolves and sets the environment variables if a Lambda function is invoked locally (serverless invoke local -f FUNCTION). This allows seamless as if the function would be deployed on AWS.

1.0.1

• Corrected plugin naming
• Improved documentation

1.0.0

• This is the initial release with all basic functionality

To-Dos

• Add support for more intrinsic functions such as Fn::GetAtt, Fn::Sub, etc. (see http://docs.aws.amazon.com/AWSCloudFormation/latest/UserGuide/intrinsic-function-reference.html)
• Write some tests!