Facebook Bot

Overview

This project tracks an eadbox (https://eadbox.com) school looking for new courses. Every new course is promoted on Facebook and you can choose how it should look, it is possible to include support for multiple schools tracking with minor changes.

Architecture

The project relies on interfaces and depends on a MongoDB instance.

• Interfaces

  IFetcher: Network resources. Define how network resources should be acquired.
  IState: State management. Define how the project should handle state.
  IPublisher: Post logics. It is possible to add support for other social networks.
  IAction: Job actions. The scheduler will dispatch jobs and jobs will dispatch actions.

• MongoDB

  The current implementation uses a MongoDB instance to store/retrieve school with courses. If you wish to use another mechanism to manage state you will need to implement the IState interface.

• Scheduler

  The project is using sundial scheduler (https://github.com/knowm/Sundial), based on a cron rule it will dispatch a job to verify new courses; when new courses were found another job will be scheduled to handle Facebook promotion.

Configuration

You need to configure the project through a .env file that must be placed in the root folder of the project.

FACEBOOK SETTINGS

• FBBOT_FB_PAGE_ID
• FBBOT_FB_PAGE_TOKEN
• FBBOT_FB_PUBLICATION_TYPE (TEXT_ONLY, IMAGE_ONLY, TEXT_AND_IMAGE) defaults to TEXT_ONLY
  Note that if your course doesn't contains an image, the robot will throw an exception and your post will be skipped.
• FBBOT_FB_TEXT_TEMPLATE (defaults to Confira nosso novo curso: %NEW_COURSE_TITLE%)
  Note that some words will be replaced for you %NUMBER_OF_COURSES%, %NEW_COURSE_TITLE%, %SCHOOL_NAME%

ROBOT SETTINGS

• FBBOT_ROBOT_MODE (defaults to POOLING)
  You dont need to define this as, currently, only POOLING is supported.

DATABASE SETTINGS

• FBBOT_DATABASE_URI (defaults to mongodb://database:27017)

SCHEDULER

• FBBOT_SCHEDULER_SCHOOL
• FBBOT_SCHEDULER_INTERVAL (default is 0/60 * * * * ?)

Usage

For a quick start, you can use this initial configuration. Just copy and paste into your .env file.

FBBOT_FB_PAGE_ID=
FBBOT_FB_PAGE_TOKEN=
FBBOT_FB_PUBLICATION_TYPE=TEXT_AND_IMAGE
FBBOT_FB_TEXT_TEMPLATE=Confira nosso mais recente curso de %NEW_COURSE_TITLE%.
FBBOT_ROBOT_MODE=POOLING
FBBOT_SCHEDULER_SCHOOL=arcifius
FBBOT_SCHEDULER_INTERVAL=0/60 * * * * ?
FBBOT_DATABASE_URI=mongodb://database:27017

Note that you still need to provide a page id and page token, both can be obtained from Facebook.
With your .env file ready, start the project with:
docker-compose up --build

Testing

This project implements unit tests. You can run all the tests with mvn clean test

About modes

The project was designed to support another modes. You can add the specific behavior for each new mode on Main class under bootstrap package.
Example: To add support for webhooks you'll need to use some library for the http service, add the new mode into models/RobotMode.java, change your .env file to use the new mode and don't forget to properly start the http server on Main class.