ddmad/string-reply

Your Task

Our company has released a beta version of String Reply Service and it has been a huge success. In the current implementation (as part of boilerplate code), the String Reply Service takes in an input string (in the format of [a-z0-9]*) and returns the input in a JSON object.

For example,

GET /reply/kbzw9ru
{
    "data": "kbzw9ru"
}

As the service is widely adopted, there have been increasing feature requests. Our project manager has come back with the following requirements for V2 of the service:

The input string will now be comprised of two components, a rule and a string, separated by a dash (-). Rules always contain two numbers. Each number represents a string operation.

The supported numbers are:

• 1: reverse the string

  E.g. kbzw9ru becomes ur9wzbk

• 2: encode the string via MD5 hash algorithm and display as hex

  E.g. kbzw9ru becomes 0fafeaae780954464c1b29f765861fad

The numbers are applied in sequence, i.e. the output of the first rule will serve as the input of the second rule. The numbers can also be repeated, i.e. a rule of 11 would mean reversing the string twice, resulting in no change to the string.

Giving a few examples,

GET /v2/reply/11-kbzw9ru
{
    "data": "kbzw9ru"
}

GET /v2/reply/12-kbzw9ru
{
    "data": "5a8973b3b1fafaeaadf10e195c6e1dd4"
}

GET /v2/reply/22-kbzw9ru
{
    "data": "e8501e64cf0a9fa45e3c25aa9e77ffd5"
}

What you need to do

Use the boilerplate given and implement the above requirement. Your implementation should also consider:

• Maintain the existing endpoint for backward compatibility.
• Implement V2 endpoint for the above new requirements.
• Additional rules are expected in future releases. The updates in rule set should have minimal code changes and impact to existing functionality.
• Testability for individual rule and the application. Unit tests are highly recommended.
• Endpoints should return correct status code and response message. For invalid request, it should return status code 400 with message "Invalid input", for example:

  GET /v2/reply/13-kbzw9ru
  {
      "message": "Invalid input"
  }
  

Upon completing the task, please feel free to (though not required):

• host your code on Github
• include any readme to explain your setup/environment
• add/implement anything you think would be beneficial

Build project

To build the project, simply run

./gradlew build

Start project

To start the project, simply run

./gradlew bootRun

Run tests

To run the tests, simply run

./gradlew test

Once the service started, the endpoint will be available at localhost:8080, so you can make request to the service endpoint

GET localhost:8080/reply/helloworld

{
    data: "helloword"
}

GET localhost:8080/v2/reply/12-helloworld

{
    data: "49afed3c7cf18693ac7f319cd01ffae4"
}

If the message is invalid, then error message will be returned (the status code will be 400):

GET localhost:8080/v2/reply/13-helloworld

{
    data: "Invalid input",
    status: "BAD_REQUEST"
}

PS: currently we treat empty message as a valid input, but the data in response will always be "Message is empty":

GET localhost:8080/reply

{
    data: "Message is empty"
}

GET localhost:8080/v2/reply

{
    data: "Message is empty"
}