Plate Up

This repository contains the source code for Group 5 - Pyromaniacs' variant of Chef's Co-Pilot. This application allows users to browse new recipes and be provided with the necessary information to be able to recreate these recipes. Our Wiki is available for more information for end-users. We welcome you to check it out!

The application follows a client-server model. The client is written in React Native as a mobile application and is found in ./PlateUp-reactnative. The server is written as a Python Flask web application. The Python Flask web application is separated as a GitHub submodule found in ./PlateUp-flask. The reason for maintaining a separate submodule for the flask app is for easier deployment to Heroku for staging, as well as better client-server seperation. The group initially had both modules together, hence the issues/tasks are tracked here, but decided to split the module just before beta release. The group will hence forth develop under this layout.

Distribution

The client application is distributed through Expo and is availiable as a mobile application at https://expo.io/@pyromaniacs/projects/plate-up-uoft-ece444. The server powering the client is availiable at https://sheltered-thicket-73220.herokuapp.com/. See PlateUp-flask submodule for more details on the server application.

Privacy Policy

There is no guaranteed privacy for your data in this beta application. We will do our best to safeguard your data and follow security practices (we never store/get access to your password, all passwords are salted and hashed), but consider any information you input public information. We will also have the right to send you a welcome email after you sign up!

Development

Respository Structure

This repo has two main components: (1) Client Application (2) Server Application.

Client Application - written in React Native using Argon React Native Template, source code in PlateUp-reactnative.

Server Application - written in Python using Flask, source code in PlateUp-flask submodule.

Local Development

To build and start the system, first clone the repository
git clone --recursive https://github.com/ECE444-2020Fall/project1-plateup-group5-pyromaniacs && cd project1-plateup-group5-pyromaniacs

Development Requirements:

1. Docker - Find installation instructions at https://docs.docker.com/get-started/
   
2. Mobile Development - Android/iOS Emulator required locally

Server Application

Running the server application is fairly simple with Docker.

cd PlateUp-flask && docker-compose up

Client Application

Navigate to ./PlateUp-reactnative.

Install all package dependencies using npm install.

Without Backend Server Running

There is an option to run the application using stubbed data. We use MirageJS to make this possible. See mock-http.js to see what HTTP requests are stubbed out. You may need to add additional ones. To use stubbed data, see App.js and set the boolean mockData to true.

With Backend Server Running

For the react-native application to be able to talk to the backend, you must configure the environment variable SERVER_URL. See the next section for configuring environment variables.

Client Environment Variables

Environmental variables used in the project must be defined within an env.js file in the root directory for the React Native portion of the application. This file is not included in the repository, as developers are expected to have their own local version. The format is as follows:

const env = {
    SERVER_URL: "insert_url_here";
}

export default env;

• SERVER_URL: This is the url to the Flask server that the frontend will use to send requests to. If running the application over LAN, use the public IPv4 address of the computer running the server, followed by :5000 for the port. E.g. SERVER_URL: "http://192.168.0.18:5000";

Starting React Native Application

Run npm start and connect to your specific emulator through expo localhost interface.

Linting and Testing

The client application is linted using ESLint. There are no current linting issues. For any contributions to this repo, ensure there are no new linting issues. The client application is tested using Jest. At this moment, we have test coverage for complex reducer logic and certain functions. Components are not currently being tested, this is an area of improvement in the future. The end goal is also to add end to end device integration tests. This was not possible due to time constraints.

Future Improvements

1. Step-By-Step Instructions - Populate real cooking time data and add a numeric timer
2. Shopping List / Grocery Inventory - Support automatic unit conversions
3. Currently, the grocery inventory is mostly manually updated (other than the shopping list flush). The API already supports an endpoint that allows users to "check" whether they meet a recipe's ingredients requirements. If the user meets the ingredients, ingredients will automatically be deducted from their grocery inventory. If the user does not meet the ingredients, missing ingredients will be added to their shopping list. The mobile application does not support this feature at the moment and this can be an improvement for the next release.
4. Search bar improvements - Support multiple ingredients search, currently it is limited by one ingredient at a time
5. Browse recipes - Add pagination support so that a user can browse more than 20 recipes at a time