The idea is to have a 2-dimensional array filled with white and black pixels. And be able to find the closest distance to a white pixel for each black pixel. You can find more details about the assignment here (however, do not look on the solution there. It is elegant but quite slow as it appears to be).

I've made a video with explanations of why this solution is chosen, how it works and why is it faster than what can be found over the web.

• 00:00 — Intro
• 00:56 — The challenge explained
• 03:50 — Solutions found around the web
• 08:25 — Benchmarking my solution vs the one found
• 11:00 — Validation
• 13:15 — Input parsing and structure
• 15:15 — Actual BFS implementation and explanation
• 18:20 — Testing and linting

1. Make sure you have Node, npm and typescript installed globally. Preferred is Node v10.
2. Download or clone this repo
3. Run npm install to get all the dependencies

Right after you've cloned the repo and installed dependencies, it is wise to run tests and see how it works. Run npm run test to test once or npm run autotest to keep Mocha rerunning tests after every change.

The tests are made with Mocha and Chai frameworks.

If you want to build the app without running, run npm run build

To build and run, type npm run start < <YOUR_TEST_DATA>. It will clean up the old build, compile the code and process the file.

Use testData.txt to try the program (like npm run start < testData.txt) and if your test file fails, compare this file to yours to find the difference. Even though, the app has a bit of validation build in and will try to help you finding the error.

This repo is using eslint (with typescript addons) and prettier to automatically lint the code.