DEV Community: Dhiraj Bhatt

How I implemented a random word generator in java?

Dhiraj Bhatt — Sat, 25 Jan 2020 13:44:54 +0000

I was building an app to display random youtube videos. For the app, I needed an easy way to get a random english word. Suprisingly, finding an API which fit my requirements was harder than I expected. Finally, I ended up writing my own random word generator using Datamuse API.

Why I couldn't use existing options?

I analysed a bunch of APIs, and none of them suited my requirements

Wordnik - Rich API with lots of documentation, but they were taking more than 7 days to send me an API key unless I donated money.
WordsAPI - Sleak website with straightforward random word API, but the free plan allowed only 2500 requests per day, which is a bit too low in my opinion.
random-word-api - Pretty good except that it generates a "random" word by choosing from a static list of words, which may result in words getting repeated frequently.
DataMuse API - This was most promising with 100,000 free requests per day, though they didn't have a straightforward random word api.

Finally, I decided to write my own random word generator java library using the DataMuse word-search API.

Writing my own random word generator

As explained before, Datamuse API was my best option. Sadly, instead of a random-word API they only had a word-search API. You make a REST API call to them with a topics parameter, and they send you back a list of words on those topics. The random word generator library which is just a wrapper around the Datamuse API. The library allows you to get a random word with a simple static method call as below

// Import the class
import com.github.dhiraj072.randomwordgenerator.RandomWordGenerator;

// A simple static method call to get the random word
String randomWord = RandomWordGenerator.getRandomWord()

Behind the scenes, a few things are going on:

We have
- RandomWordGenerator class which holds a list of words(randomWords) in memory
- Topics class which holds a static list of various topics e.g. acting, chess, etc.
During initialization, RandomWordGenerator makes an HTTP request to DataMuse API with a random Topic and updates randomWords value with the list of words returned by the API
When a user makes a call to RandomWordGenerator.getRandomWord()
- an HTTP request to DataMuse API is initiated to get the next list of words in a separate thread
- a randomly chosen word from the randomWords list is returned
In case the HTTP request made above fails, system defaults to getting a random word from the Topics class itself. Something is better than nothing.

Note that in a separate thread part is pretty important above. Instantiating the HTTP request to DataMuse API in a separate thread allows the RandomWordGenerator.getRandomWord() call to be instantaneous as it doesn't need to wait for the HTTP response to come back.

The above design achieves a few things:

Every call to RandomWordGenerator.getRandomWord() is instantaneous since we return the word from the randomWords list already present in memory. I believe this is better than most/all REST APIs as there is no overhead of an HTTP request.
Randomness of the words is ensured over time as every call to RandomWordGenerator.getRandomWord() has the side-effect of updating the current randomWords list with words for a new random Topic
Randomness of words is ensured for quick multiple calls to RandomWordGenerator.getRandomWord(), because for every call a random word is chosen from the randomWords list
It's robust. If due to any reason the call to DataMuse API fails, then we fall back to the offline Topics class to get a random word.

In case you want to have a look at the source code or use this library in your own project, the link is below. Please feel free to post your suggestions/questions in comments!

Dhiraj072 / random-word-generator

A java library to generate random words

Random word generator

A java library to generate random words. This utilizes the DataMuse API as its word-finding engine.

Usage

Import from Maven Central

Using Maven

<dependency&gt
  <groupId>com.github.dhiraj072</groupId&gt
  <artifactId>random-word-generator</artifactId&gt
  <version>1.1.0</version>
</dependency>

Using Gradle

compile 'com.github.dhiraj072:random-word-generator:1.1.0'

Get a random word

// Import the class
import com.github.dhiraj072.randomwordgenerator.RandomWordGenerator;

// Get a random word with a simple static method call
String randomWord = RandomWordGenerator.getRandomWord();

Get a random word skewed towards various parameters supported by DataMuseRequest

import com.github.dhiraj072.randomwordgenerator.datamuse.WordsRequest;
import com.github.dhiraj072.randomwordgenerator.datamuse.DataMuseRequest;

// Get a random word skewed towards topics "Car" and "Road"
WordsRequest customRequest = new DataMuseRequest().topics("Car", "Road");
String randomWord = RandomWordGenerator.getRandomWord(customRequest);

Built With

Gradle - Build tool
DataMuse API - For getting list of words
JUnit - Testing Framework

License

This project…

View on GitHub

How to deploy a static web app on AWS S3 for free?

Dhiraj Bhatt — Sat, 18 Jan 2020 16:23:28 +0000

So you built a static react web app, and you want the world to see it. This post provides a step-by-step guide on deploying your static web app on Amazon S3 for free.

For this tutorial, you will need

a basic understanding of react web apps, terraform and AWS S3
an activated AWS account. If you don't have one yet, you can follow instructions here to set it up
an AWS IAM user with
- AWS API keys (Access Key ID and Secret Access Key) set up. You can follow instructions here to setup and configure the IAM user.
- Permissions to create/delete S3 buckets, add IAM policies
Terraform CLI to set up our infrastructure on AWS
AWS CLI which has been configured with the IAM user API keys

AWS has a free tier so running the steps below should not cost you anything. In any case, make sure you TERMINATE all your resources after completing this tutorial to avoid any unwanted charges in your AWS account.

Preparing the package

I presume you already have a static react web app which you want to deploy. If not, you can pull a sample web application from my repository. I will be writing this guide using the indecision-app below. It has been build using yarn and webpack, though the steps should be easily reproducible with other package managers (npm, gulp, etc.)

$ git clone https://github.com/Dhiraj072/indecision-app.git

Once you have the app, build a production package.

$ yarn webpack -p

This above command will create a public/ directory inside your project folder, containing everything required to host your web application, including an index.html, bundle.js, etc.

Setting up AWS infrastructure

We will write the terraform configuration file main.tf step-by-step in this section. This file tells terraform which AWS resources to instantiate, and their respective configuration. Full version of the file can be found here.

Create a blank main.tf file in your project dir
Add below to our main.tf to tell terraform we are using AWS for our infrastructure

provider "aws" {
  # region you want your resource to be in
  region = "ap-southeast-1" 
}

We store our terraform states locally for now, so add

terraform {
  backend = "local"
}

To define local variables we will be using later on, add

locals {
  s3_origin_id = "myappS3Origin"
  app_name = "my_app"
}

Finally, add the config to create the S3 bucket we want our app to be hosted in. This also sets the IAM policy to allow public read as we want everyone to be able to access our web app.

resource "aws_s3_bucket" "my_app" {
  bucket = "${local.app_name}" # S3 bucket name
  acl    = "public-read"
  website {
    index_document = "index.html" # our web app's entry point
  }
  policy = <<POLICY
{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Sid": "AllowPublicReadAccess",
      "Effect": "Allow",
      "Principal": "*",
      "Action": [
        "s3:GetObject"
      ],
      "Resource": [
        "arn:aws:s3:::${local.app_name}/*"
      ]
    }
  ]
}
POLICY
}

Save your main.tf file
Set the environment variable for terraform to be able to access AWS

$ export AWS_SECRET_ACCESS_KEY=<YOUR_AWS_SECRET_ACCESS_KEY>
$ export AWS_ACCESS_KEY_ID=<YOUR_AWS_ACCESS_KEY_ID>

Initialize terraform

$ terraform init

Spin up the AWS infra, type y when asked for confirmation

$ terraform apply

Terraform may take 5-10 minutes to spin everything up. Once complete, you can login to you AWS account and you should see you AWS bucket created.

Your AWS infrastructure is now ready to host your web application.

Deploying the package

Finally, you can deploy the package we built earlier to S3 by running the following AWS CLI command in the project dir

$ aws s3 sync public/ s3://my_app

Accessing deployed web app

You should be able to access the application at https://s3-ap-southeast-1.amazonaws.com/my_app/index.html. This is basically the Object URL of the index.html file inside our S3 bucket my_app.

Terminating created resources

AWS has a free tier so running the steps in this tutorial should not cost you anything. In any case, do TERMINATE all your resources to avoid any unwanted charges by running the following command in your project dir

$ terraform destroy