Bokio.md

Bokio

Test the project

• Source code: https://bitbucket.org/MQuy/bokio
• Demo version: http://bokio-web.studynow.vn
• Or watch online at youtube: https://www.youtube.com/watch?v=9fYaz41mmKo

How to setup the project

• Go to this url: http://railsinstaller.org/en
• Download Ruby 2.3
• After finishing installation, git clone https://bitbucket.org/MQuy/bokio Bokio & cd Bokio
• run gem install bundler and bundle install
• run rails db:create && rails db:migrate (if you have any problem, please check this link https://stackoverflow.com/questions/35545361/rails-the-system-cannot-find-the-path-specified)
• run rails s -p 4000 and keep this console

Open a new console

• run cd frontend
• run bower install && npm install
• run gulp serve and it will open the new tab

Run tests

• run rspec spec

API Format

• Server: JSON Format
  • Import: POST /user_transactions/bulk_import with { data: xxxx }
  • All: GET /user_transactions

How about the project

Structure

├── app                               # Application source code
│   ├── controllers                   # The place to receive parameters which are sent from client
│   └── domains                       # Divide and handle business model
│       └── transaction               # Transaction is responsible for features like importing ....
│           └── page_impport          # Page Report contains three components: check data, parse data, store transactions.
│               ├── parser            # Receive data which is sent from client and parse it. After that returing results which are all the transaction rows
│               └── store             # Store transactions into database after receiving parser's output
├── frontend                          # Folder for frontend
├── spec                              # Test files

Domain Flow

Walkthrough

• The flow: I chose Railway Oriented Programming
• The stucture: I applied DDD to map from businesss requirement/model to the code

The reason I chose railway oriented programming is that it not only helps me to focus on "the happy path" but also provide a clean way to handle error.

There are 3 subdomains inside Transaction/PageImport

1. Check the data:

• Their role is checking whether the input data is correct or not.
• If input data is correct, the data will be passed to Parser subdomain

2. Parser: the current implementation is without knowledge about colum names which can be specified from user.

• It will read line by line from the data above and get only the transaction rows with conditions
  • Row has to contain one column with date format and one column with number format
• After filter all tranasction rows, it will build "plain old objects" for those transactions and pass it to Store subdomain

3. Store:

• It will persist those transactions into database, if there is any error, store will give error message for the Output

I hope with these, it will help you to have a big picture about my code.

Best regards