User API

This is a central user API for storing and retrieving user details for HashtagSell.

Documentation

Once the API is started, documentation can be viewed on the server: https://localhost:4523/v1/docs

Note: Please swap out localhost and port above to match configuration

Object Model

Below is an example user from the user-api.

var user = {
      userId : {
        index: true,
        required: true,
        type: String
      },
      firstName : {
        required : false,
        type : String
      },
      lastName : {
        required : false,
        type : String
      },
      email : {
        index : true,
        required : true,
        type : String
      },
      password : {
        required : true,
        type: String
      },
      phone: {
        required: false,
        type: String
      },
      isAdmin : {
        required : true,
        type : Boolean
      }
};

Docker

Environment Variables

• DATA_MONGO_URL - defaults to mongodb://localhost:27017/hashtagsell-user-v1
• ENVIRONMENT - defaults to develop
• LOGGING_LEVEL - defaults to info

An example command to run locally:

docker run -d -p 8880:8880 \
  -e LOGGING_LEVEL=trace \
  HashtagSell/user-api:latest

Development

Prerequisites

• Node v0.10+
• MongoDB v3.0.2

A simple way to install and run Mongo is to use Homebrew (http://brew.sh/):

brew install mongo

Getting Started

To clone the repository and install dependencies, please perform the following:

git clone [email protected]:hashtagsell/user-api.git
cd user-api
npm install

For convenience you may find it useful to install gulp globally as well:

sudo npm install -g gulp

Local Configuration

The server utilizes a mechanism for allowing configuration overrides by environment. To make use of this for a local environment, create a local.json configuration file in a new folder named config at the root of the application. The .gitignore has an entry to ignore this folder so that the local config won't trash other environment configurations.

mkdir ./config
touch ./config/local.json

Now put the following into the local.json configuration file:

{
  "data": {
    "mongo": {
      "documentCountLimit": 1000,
      "url": "mongodb://localhost:27017/hashtagsell-user-v1"
    }
  },
  "logging": {
    "level": "trace"
  },
  "server": {
    "port": 4043,
    "secure": false
  }
}

Run It

Mongo DB

After installing all modules, gulp can assist you in starting Mongo:

gulp mongo-start

Please Note: All mongo data for HashtagSell is placed into /usr/local/var/mongodb/hashtagsell by default

Data Migrations

Once Mongo is started, run the following migrations to get a new database up to the right version:

npm run migrations

API Server

When starting the API server, an environment configuration should be specified (see above for details on creating a local.json environment config). To specify the environment, use the NODE_ENV environment variable in the console to begin the process. The npm start script uses supervisor and pipes the output to Bunyan for logging:

NODE_ENV=local npm start

General Development Notes

As changes are saved to .js files in the project, supervisor will automatically restart the server. It may be useful to periodically check the terminal to make sure a runtime unhandled exception isn't stopping the server. Additionally, using jshint may help to uncover potential problems before running code. The npm test script is connected to gulp jshint to help with this.

Pushing Branches / Pull Requests

Prior to pushing your branch to the remote to create a pull request, please ensure you've run the tests and verified and fixed any JSHint issues or failed unit tests:

npm test

After running npm test code coverage reports are created that can be viewed to determine if model changes have adequate code coverage prior to submission:

open ./reports/lcov-report/index.html

Application Structure

• init - x509 certs, Nginx config and Upstart config scripts are here
• lib - all application code resides in this folder
  • config - default configuration for the application - this should have a master set of all supported configuration keys
  • data - handles mapping of models to the underlying data storage systems - abstracts the database entirely away from all layers above it
  • models - relies on the data layer to store and retrieve model data - contains all logic related to the model including validation and sanitation
  • routes - relies on the model to support various resources exposed via the API
• test - all test code resides in this folder