Transactions

I implemented a web server that exposes an API used for adding and analysing transactions performed at a bank. Each transction has a timestamp at which it was performed, the two people involved in the transaction and the amount.

Solution

For this application I used Flask, the microframework for Python.

1. First I imported the Flask class.
2. Next I create a connection to MongoDB using MongoClient and I create an instance of the Flask class.
3. The main logic of the application is in the transactions and balance methods. I used route() decorator to tell Flask what URL should trigger our functions. - POST to http://127.0.0.1:5000/transactions/: I created an json object then that contains “sender” - integer, “receiver” - integer, “timestamp” - integer, “sum”. Then I used insert() method to insert the document into a collection named transactionsTable. - GET to http://127.0.0.1:5000/transactions/?user=XXXX&day=YYYY&threshold=ZZZZ: I used request.args.get() to get the parameters value by name and then I used find() to issue a query to retrieve data from a collection in MongoDB. - GET to http://127.0.0.1:5000/balance/?user=XXXX&since=YYYY&until=ZZZZ: similar with the previous GET I used find to retrieve the data from the collection. Then to calculate the balance I iterate the cursor for the query where the user was a receiver first, and then when the user was a sender. The balance is the difference between the total amount received minus the total amount sent.

To create the Docker image and run the container run:

./run.sh

Running the tests local

I created some tests to run the application on your machine also. Firstly, you need to install the MongoDB. After that, you need to insert a few transactions into the database.

./insert_some_data.sh

Then start the server and run test.sh which contains 3 tests, one for each operation.

python web-server.py
./test.sh