jomendez / udacity-image-processing-api

Udacity fullstack nanodegree

Udacity Image Processing Api

Udacity fullstack nanodegree

Setup project

• Clone the repository and install npm dependencies
• npm install to install dependencies
• npm run test to run the jasmin tests
• npm run lint - script runs eslint with prettier plugin
• npm run build - script compiles typescript files from ./src to js in ./dist
• npm run start-dev - script runs nodemon on src/index.ts file for development
• npm run start - script runs node using ./dist/index.js file

Set up a project structure that promotes scalability

• Source code is in ./src folder, compiled code is in ./dist folder
• Tests are in ./src/tests
• Separated modules getCache and createThumbImage created for image processing and caching

How to use it

after running npm run start:dev to start a local server in your computer, then you can open a browser and navigate to:

• http://localhost:3000/ root of the api
• http://localhost:3000/api/image/?name=palmtunnel.jpg&width=300&height=400
• http://localhost:3000/api/image/?name=fjord.jpg&width=500&height=500

Available Image name options

encenadaport fjord icelandwaterfall palmtunnel santamonica

Using Express endpoints

• '/' route respond with a text to confirm express is running correctly
• '/api/image' route with url query strings for name, width, and height will process the image. Eg. '/api/images?name=fjord.jpg&width=500&height=500'

Middleware

• used multiple custom middleware - validCheck to check for valid filename and query parameters, imgSharp to process image.
• validCheck middleware returns errors if query parameters are missing or the filename is invalid

Automated Test with Jasmine and Supertest

Tests for getCache and createThumbImage helpers + test for API endpoint

Functionality

• User can query endpoint using various params and queries to retrive an image with a specified name, height and width.
• The default height and width is set to 200px.
• The images requested will be saved on the server disk.
• There is a cache layer. If a user requests an image size that has already been requested, then the previously resized image will be returned.