Report Bot

⚠️ Recently, I merged the bot folder into the server folder so we can make use of the @slack/bolt app and Events API instead of RTM. Heroku can't direct requests at both dynos, so the server will now handle both. This means the documentation below is outdated. ⚠️

1 Intro

This project consists of three applications contained in the following folders:

• bot,
• client,
• server.

Bot is a Slack bot which is in charge of scraping data from specified workspace and channels and saving them to the database. It can be run standalone.

Applications Client and Server are tightly coupled and have to be run simultaneously:

• Server is responsible for fetching data from the database and sending them to the client on demand.
• Client is an UI which allows the user to view scraped data categorized by special tags contained in the messages.

2 Database

• The database is needed by Bot and also Server. We will need an instance of Postgres database.

2.1 How to run locally

• The easiest and fastest way how to set up the database is using Docker. You can install and run Postgres database as a Docker container with the following command:

$ docker run --name report-bot-db -p 127.0.0.1:5432:5432 -e POSTGRES_DB=report_bot -e POSTGRES_PASSWORD=postgres -d postgres

• In the root folder, copy .env.template into .env.

• When the database is running you need to run migrations:

$ yarn knex migrate:latest

2.2 Developer notes

• How to check the status of the Docker container with the database:

$ docker ps -a

• How to run the Docker container with the database (e.g. after computer restart):

$ docker start report-bot-db

• How to stop and remove the docker container with the database:

$ docker stop report-bot-db && docker rm report-bot-db

• How to create a new database migration, e.g. for creation of report table:

$ yarn knex migrate:make create_report_table

• How to rollback the last batch of migrations:

$ yarn knex migrate:rollback

3 General info

Project uses yarn workspaces structure. So you should run every yarn command from the root folder.

To install dependencies for each part of the project, just run:

$ yarn

4 Bot application

4.1 Set-up and running

• Create a bot for your workspace.

• Under the section "OAuth & Permissions" in settings of the application, you can find tokens needed to run Bot:

  • "Bot User OAuth Access Token" starts with xoxb and will be called as botToken in our codebase.
  • "Signing secret" will be called as signingSecret in our codebase.

$ yarn build && yarn start

4.2 How to use the Bot

• Invite the Bot into some public channel or private group you want to scrape its content and store them in a database:

/invite @ReportBot 

• If you want to stop scraping the content of some channel you can run the command:

/remove @ReportBot 

• If you want to (un)archive some projects you can run the command:

(un)archive [tag1] [tag2] ... [tagN]

• If you want to change project's name to red, if there was no activity from set period (default is 16 days), you can run the command:

frequency [tag1] [frequency1] [tag2] [frequency2] ... [tagN] [frequency2]

• To categorize the message under some tag you can use special tags in the format :__some-tag:.

5 Server application

5.1 Set-up and running

• In the server folder, copy .env.template into .env and fill missing values.

• Go into the folder server and run the application:

$ yarn build && yarn start

• Then you can run the Client application (see the following section).

6 Client application

6.1 Set-up and running

• Go into the folder client and run the application:

$ yarn build

7 Deployment

We use Heroku for hosting this application. When something is merged into master branch, it is automatically deployed. You can also deploy any branch manually from Heroku Dashboard.

Important parts regards to Heroku to note:

• in package.json:
  • engines.node - which Node version to use
  • scripts.heroku-postbuild - what to run in build phase (instead of scripts.build)
• in Procfile:
  • web - how to start web part of the application
  • bot - how to start bot (non-web) part of the application