Skip to content
This repository has been archived by the owner on Aug 14, 2024. It is now read-only.

Latest commit

 

History

History
490 lines (354 loc) · 16.2 KB

README.md

File metadata and controls

490 lines (354 loc) · 16.2 KB
Raw

Unemployment Insurance Claimant Intake

⚠️ Disclaimer

The code in this repository represents the development efforts towards building a state-specific claimant intake form. This public repository exists purely for the purposes of hosting a snapshot of code to serve as a reference for similar efforts, and is therefore not actively maintained. Readers of this message are looking at outdated, incomplete, and un-runnable code, without specific credentials. It is recommended to use this as a resource or reference rather than attempting to expand on or assimilate large sections of code.
Entities wishing to use this code will need to remove unusable integrations locked behind state secured infrastructure, and should do standard due diligence and make sure to use up-to-date dependencies.

Test and Lint Deploy Dev Image Scans

Modern Unemployment Insurance (UI) intake application for the state of New Jersey Department of Labor.

Set up

This project uses several tools which are either recommended or required for development.

Recommended tools

This project uses Make as a convenience for environment set up and other development commands.

ℹ️ Note:

  • On Windows, Winget or Chocolaty can be used to install Make.
  • It is possible to interact with the project without make. To do so, simply run the commands associated with the corresponding make commands from the Makefile. Setup instructions in this project will assume the installation of Make.

Required tools

  • Python (with PIP)
  • Zscaler - Installed as part of onboarding when you get a .dol.nj.gov email.
    • This project integrates with services (e.g. Address Verification) that is only accessible when on the NJ state network. When running the app locally, Zscaler must be running and authorized. If not, you will see a 500 error.

Client and Cypress (e2e)

  • Node
    • This project's required Node version can be found in .tool-versions
    • Since Node versions enter and exit active LTS status regularly, use a Node version manager. On Mac use asdf for version management by following these steps:
      • Install asdf if not already installed (https://asdf-vm.com/guide/getting-started.html)
      • Install GnuPG (a prerequisite for asdf controlling Nodejs version) 
        brew install gpg gawk
      • Install the asdf NodeJS plugin 
        asdf plugin add nodejs https://github.com/asdf-vm/asdf-nodejs.git
      • Check that the nodejs plugin installed correctly with the below command 
        asdf plugin list
        NOTE: you should see 'nodejs' listed in the terminal as a response
      • Install the correct NodeJS version specified in .tool-versions 
        asdf install nodejs <version number>
      • Set the version used above to be applied globally 
        asdf global nodejs <version number>
      • Double check NodeJs is using the correct version 
        node -v
      • If you have an issue run the following command and then close and re-open your terminal to be sure changes have taken affect 
        asdf reshim nodejs
      • If you are still running into an issue due to previous installations of node, do the following
        • Add the node version to your path file by opening your startup file (ex. ~/.zshrc or ~/.bash-profile) and adding the following to the bottom: 
          export PATH=“/Users/<YourUsername>/.asdf/installs/nodejs/<DesiredVersion>//bin:$PATH”
        • rerun the asdf reshim command to ensure asdf forces a recheck on your setting 
          asdf reshim nodejs
        • Close and reopen your terminal and check that node is now on the correct version: 
          node -v
    • for Windows environments, nvm-windows is highly encouraged.
  • Yarn

Server

  • Java
    • The above link is directed to the Eclipse Temurin SDK. The Eclipse foundation has taken over AdoptOpenJDK as of 2021 which is still the best choice for free OpenJDK builds. You can either download and manually install the binaries from Adoptium, or use an SDK manager, as outlined below.
    • This project uses the current LTS version of Java, specified in .tool-versions.
    • Since Java versions now enter and exit active LTS on a much faster cadence than in its early days, using a version manager like asdf is highly encouraged. While asdf is only supported on UNIX-like platforms. WSL is the recommended way to use asdf on Windows.
    • With asdf installed, proceed with installing Java:
      • Add the asdf java plugin 
        asdf plugin-add java
      • List all available version 
        asdf list-all java
      • Select and install the java version you would like to use 
        asdf install java {java-release-candidate}
        Where {java-release-candidate} is the name of the release you would like to use. The temurin- prefixed candidates are from the AdoptOpenJDK/Adoptium vendor. Make sure not to accidentally select the candidates prefixed with {vendor}-jre-, as that is just the JRE. Developers will want the JDK.
      • Set the installed version as your global default 
        asdf global java {java-release-candidate}
      • Set JAVA_HOME by adding the relevant initialization script to your shell's initialization file, e.g. for zsh shell, add the following to your ~/.zshrc 
        . ~/.asdf/plugins/java/set-java-home.zsh
      • Verify the installation of Java 
        java --version
      • Verify JAVA_HOME is set correctly 
        echo $JAVA_HOME

Dependencies

The following should install everything you need on macOS:

Sets up pre-commits:

make dev-deps

Sets up client app in /client:

make client-deps

Sets up server app in /server:

make server-deps

Sets up Cypress for end-to-end testing in /e2e:

make e2e-deps

OR run all of the above dependencies:

make all-deps

Configuring your environment

Locally, the client and the server make use of .env files to store environment secrets that should not be committed to version control. You will not be able to run the applications locally without a properly configured environment.

To set up .env files locally:

Make a copy of client/.env-template and name it .env:

cp client/.env-template client/.env

Make a copy of server/.env-template and name it .env:

cp server/.env-template server/.env

Make a copy of e2e/cypress.env.json-template and name it cypress.env.json

cp e2e/cypress.env.json-template e2e/cypress.env.json

In your copies of the env files, you should remove comments/references about not adding secrets which would have been carried over from the .env-template files, since you will be adding secrets there.

Add values for the environment variables. Note some variables in the .env-template include instructions for what values to use. For variables including specific secrets, the Truss team uses a 1Password vault to securely store/retrieve .env secrets. Reach out to another engineer to determine the best way to get the correct values for each environment variable.

Reminder: These values are sensitive and should not be committed to version control, therefore, .env files are .gitignore'd. Take care not to accidentally expose these variables via other means.

Web client

To run the client server locally:

make client-dev

Troubleshooting Local Web Client

When running the above commands, you may encounter one of the following:

500 server error from /partial-claim

Run make docker-clean

Failed network requests for Address Verification or Occupations API

These downstream services require running on the NJ network. Make sure Zscaler is running and your session still is active.

Web server

To run the api server locally:

make server-build
make server-bootRun

Troubleshooting Local Web Server

When running the above commands, you may encounter one of the following:

Hikari Pool Definition Exception:

org.postgresql.util.PSQLException: FATAL: role "localdev-postgres" does not exist

This error occurs when another instance or postgres is running in the background. Use brew services list or your Activity Monitor to find any existing postgres instances and end them

After closing any other instances of Postgres, re-run make server-bootRun and the port/postgres instance should be free to host this application's DB

Script make server-bootRun remains at "80% EXECUTING"

The client is dependent upon locally hosted backend services. Running make server-bootRun starts an open task that runs continuously. There is no need to wait for the status to reach 100%

The `make client-dev' script for the frontend should be executed in a different terminal window

API documentation

The web server exposes several API endpoints, and uses swagger as a self-documenting API tool.

To use the swagger user interface /intake-api/swagger-ui.html

To view the JSON-formatted api documentation, navigate to /intake-api/v3/api-docs

Makefile

Run make or make help to see the available make commands we use with this repository.

Docker setup

To run the application locally in docker, use:

make docker-up

Or, if you prefer to run the application(s) outside of docker (in dev-mode for example) alongside dockerized services, use:

make docker-services-up

Then, run the relevant application(s) individually.

To shut the docker services down, use:

make docker-down

LocalStack

The application integrates with AWS services in the deployed environments, such as Amazon Simple Storage Service (S3), and we use LocalStack to emulate those AWS services locally.

Dependency: AWS command-line interface (CLI)

You can use the AWS CLI to interact with the LocalStack environment.

To install the AWS CLI:

# MacOS
brew install awscli

# Windows
choco install awscli

See Setting up local region and credentials to run LocalStack for instructions to configure the AWS CLI to work with LocalStack.

Usage

LocalStack runs as a docker container and will be started along with the other application services by running the following command:

make docker-up

or

make docker-services-up

To confirm the LocalStack container is running, run the following command:

docker compose ps

# The output should include the following container:
NAME                COMMAND                  SERVICE             STATUS              PORTS
dol-ui-localstack   "docker-entrypoint.sh"   localstack          running (healthy)   0.0.0.0:4566->4566/tcp

Once the container is running, you can use the AWS CLI to interact with it by including the --endpoint-url=<localstack-url> flag:

# to list all buckets in the LocalStack environment
aws --endpoint-url=http://localhost:4566 s3 ls

# to list the objects within a particular bucket, for example:
aws s3api list-objects-v2 --endpoint-url http://localhost:4566 --bucket dol-ui-claims

# to download an object from a particular bucket, for example:
aws s3api get-object --endpoint-url http://localhost:4566 --bucket dol-ui-claims --key "YOUR_OBJECT_KEY_HERE" "NAME_OF_OUTPUT_FILE"

# to sync the entire bucket to a local directory, for example:
aws s3 sync --endpoint-url http://localhost:4566 s3://dol-ui-claims tmp

Nginx

In the deployed environments, the frontend and backend applications sit behind an Application Load Balancer. We can replicate some of this configuration locally by using Nginx.

To use Nginx locally, first run the following command to generate a certificate and key for the Nginx server:

make nginx-cert

Note: Do not commit the created certificate or key to the repository. We've configured the .gitignore file to ignore them and we use a pre-commit hook that checks for and prevents private keys from being committed.

Next, add this entry to your /etc/hosts (or Windows equivalent) file:

127.0.0.1  sandbox-claimant-intake

Then run the following command to start everything in docker:

make docker-up

You can now visit the application at https://sandbox-claimant-intake:8443.

Wiremock

Locally, some API endpoints are not available, or it is not desired to make external requests for certain endpoints (e.g. Wage record). For those instances, Wiremock is available for stubbing external API endpoints.

This project runs Wiremock in a docker container when invoking make docker-services-up.

API requests made by the application running locally are intercepted by Wiremock, and (provided a matching JSON stub), specific responses can be returned.

See wiremock/mappings for the JSON mappings stored and shared in version control.

CI/CD

See CI/CD Configurations for an overview of the application's approach to CI/CD.

Testing

Client tests

To run the Jest unit tests:

make client-test

To run Jest unit tests that will run with every change:

make client-test-watch

Server tests

To run the JUnit tests:

make server-test

To generate the test code coverage report:

make server-test-coverage-report

To view the report, open server/build/reports/jacoco/test/html/index.html in your browser.

End-to-end tests (Cypress)

To run Cypress (e2e) tests locally there are a few considerations:

  • You must run the server app with the 'e2e' profile active for mocked authentication to work

Option A: Run client, server in dev mode

To run Cypress tests while running the client and server locally in dev mode:

# Run the supporting services (db, localstack) in docker
make docker-services-up

# Terminal #1: Run the client
make client-dev

# Terminal #2: Run the server
# Note: server-bootRun-e2e is only for running cypress tests, which requires
# mocked authentication for the cypress tests to work. When not running cypress
# tests, use server-bootRun to enable real login functionality.
make server-bootRun-e2e

# Terminal #3: Run cypress (two options)
make e2e-test-gui-local      # Runs Cypress tests in browser
make e2e-test-headless-local # Runs Cypress tests on the command line
make e2e-test-gui-local-fast # Runs Cypress test in browser without lighthouse

Option B: Run client, server in docker

To run Cypress tests while running the application in docker:

# Run the application and supporting services in docker
# Note: docker-e2e-up is only for running cypress tests, which requires
# mocked authentication for the cypress tests to work. If you want to run
# everything in docker with real login functionality, use docker-up.
make docker-e2e-up

# Run cypress (two options)
make e2e-test-gui-docker      # Runs Cypress tests in browser
make e2e-test-headless-docker # Runs Cypress tests on the command line