Serverless Kotlin application which screen scrapes and parses lunch info from lounaat.info and sends wanted data to Slack channel. This is also my first touch to Kotlin/Kotless.
Lunch whisperer uses Kotless - Kotlin serverless framework. Kotless generates Terraform code from your application code and by its own words, gives you one magic button to deploy your Web application as a serverless application on AWS!
Slack is free to use and you can create Slack application very quickly and without any programming.
You have to have AWS account and which can be used with Kotless framework. Kotless uses Terraform under the hood to create the needed AWS resources and configurations. Hadr Hariri has written an instruction how to set up your AWS account for Kotless.
There are few constant variables which you have to fill in to Main.kt. Currently, there is no possibility to use environment variables with Kotless like in Serverless framework. If there are some secrets that you don't want to keep in VCS you can use e.g. AWS secrets manager.
| Variable | Description |
|---|---|
| LOUNAAT_INFO_URL | API key that enables Grafana API usage |
| SLACK_OAUTH_TOKEN | Created automatically by Slack app, enables request verification |
| SLACK_CHANNEL_ID | Slack channel id |
| MENUS_TO_SENT | Restaurant names, which menus you want to send to Slack |
| FAVOURITE_DISHES | List you favourite dishes and if you favourite is available somewhere, menu is sent to Slack |
For deployment purposes fill AWS specific parameters to build.gradle.kts.
You can run application locally using kotless local Gradle task.
Deploying is done using kotless deploy Gradle task.
The only way I managed to debug the application locally was using remote JVM debugging with good old
VM options -agentlib:jdwp=transport=dt_socket,server=y,suspend=n,address=5005 and attaching
to process.