DEVELOPMENT.md

FSA development

This document explains how to integrate the AWS Lambda functions detailed in SPECIFICATION.md in the AWS SDK of your choice.

Add new Lambda function configurations

The FUNCTIONS map in functions.ts maps an arbitrary deployment name to a list of Lambda functions required by Feedback Sentiment Analyzer (FSA). This is the main entry point for defining your Lambda functions that will run with FSA. These functions are deployed when you run cdk deploy as described in the README. The following steps explain how to add new functions.

1. Add a new entry to the FUNCTIONS constant. For example, adding JavaScript functions might look like this:

const FUNCTIONS: Record<string, AppFunctionConfig[]> = {
  examplelang: EXAMPLE_LANG_FUNCTIONS,
  javascript: JAVASCRIPT_FUNCTIONS,
};

2. JAVASCRIPT_FUNCTIONS is undefined. Define a constant named JAVASCRIPT_FUNCTIONS and set it equal to an empty AppFunctionConfig array.

const JAVASCRIPT_FUNCTIONS: AppFunctionConfig[] = [];

3. Define the objects inside the JAVASCRIPT_FUNCTIONS array. Each object is an AppFunctionConfig that extends FunctionProps from the AWS Cloud Development Kit (AWS CDK) library. FSA requires the following named Lambda functions:

• ExtractText
• AnalyzeSentiment
• TranslateText
• SynthesizeAudio

A definition for the ExtractText function might look like the following:

  {
    ...BASE_APP_FUNCTION,
    codeAsset() {
      return Code.fromInline(`
        exports.handler = async (event) => {
          // implementation goes here
        }
    `);
    },
    name: "ExtractText",
  },

The codeAsset function must return a Code object. You can include inline code or create a bundled asset.

Implement Lambda functions

FSA requires a specific set of Lambda functions. The accompanying CDK script orchestrates these functions as AWS Step Functions. For a detailed specification of these functions, visit the following links in SPECIFICATION.md.

• ExtractText
• AnalyzeSentiment
• TranslateText
• SynthesizeAudio

Bundling (language dependent)

The AWS CDK can compile resources during the deployment using Docker. To configure this, use the bundling option of Code.fromAsset. Languages with a well-known Runtime usually have a Runtime.image property that has an appropriate Docker base image. Otherwise, any public Docker image is suitable.

Docker will execute the command for the bundling step. The assets, specified as the first argument to Code.fromAsset, are mounted in the container at /asset-input/ (which is also the working directory). The command should perform any compilation steps necessary, and move any necessary artifacts to /asset-output/. The contents of /asset-output/ are archived in the Lambda's S3 bucket. These contents are the executable that the Runtime uses.