Introduction

This application shows how to read a file and generate metrics based on the contents of that file.

Architecture

The code showcases two different ways to read a file

• Buffered Approach - the whole file is buffered into memory in a single chunk and processed in one step
• Streaming Approach - the file is streamed into memory in multiple chunks and processed in multiple steps

Buffered Approach

Pros

• The code is simpler to understand as the logic is sequential
• The code will run faster as everything it needs is in memory

Cons

• The program will crash with a file too large to fit in memory
• The program can potentially fill up memory affecting system performance

Streamed Approach

Pros

• The code can process files larger than that which can fit into memory
• The code can tune chunk size to ensure not too much memory is used up

Cons

• The code is harder to understand as the logic is concurrent
• The code will run slower as the file is loaded in multiple chunks to be processed

Install Steps

• Install Node 7.8.0

Build Steps

Open up a command line

$

Install the project dependencies

$ npm install

Run Steps

Run the application to produce outputs from running both approaches on same text file

$ npm start

Confirm the results of running both approaches produce the same output (For example on Mac OS X)

$ diff output/buffered.json output/streamed.json

Take a sneak peek to check word count and primality (For example on Mac OS X)

$ head -n 30 output/buffered.json
$ head -n 30 output/streamed.json

Test Steps

Run the tests

$ npm test

View the test results in your browser (For example on Mac OS X)

$ open coverage/lcov-report/index.html

Technology Stack

• Node - used to build and run the application