Spree Pusher

Connect your SpreeCommerce Storefront to an API endpoint

Installation

Add spree_pusher to your Gemfile:

gem 'spree_pusher', github: 'MisinformedDNA/spree_pusher', branch: 'master'

Bundle your dependencies and run the installation generator:

bundle
bundle exec rails g spree_pusher:install

Configuration

All the configuration is done inside the initializer here: config/initializers/pusher.rb all the default settings can be found there as well. Below we will explain all of them

credentials

Sets the credentials for the API calls.

To set the credentials for all the requests

# Wombat
Spree::Pusher::Config.configure do |config|
  config.credentials = {
    :connection_token => "YOUR TOKEN"
    :connection_id => "YOUR CONNECTION ID"
  }
end

# Azure Logic Apps
Spree::Pusher::Config.configure do |config|
  config.credentials = { 
    :access_key, "ACCESS_KEY"
  }
end

To set the credentials for a specific object

config.payload_builder = {
  "Spree::Order"  => {
    :serializer => "Spree::Pusher::OrderSerializer", 
    :credentials = {
      :access_key => "ACCESS_KEY"
    }
  }
}

pusher_client

By default, all objects are pushed to Wombat. To push them elsewhere, such as Azure Logic Apps, you can set pusher_client

config.pusher_client = "Spree::Pusher::AzureLogicAppClient"

You can also specify :pusher_client in the payload_builder

The difference between pusher_client and push_url is that a pusher_client can perform API specific tasks like additional formatting or customize the HTTP request, where as push_url is simply the URL the request is made against.

push_objects

The push_objects is an array of model names that are selected to push to the API endpoints. We use these as keys in other places as well to configure how the payload is serialized and to keep track of the last time we pushed the objects.

config.push_objects = ["Spree::Order", "Spree::Product"]

By default we only push Spree::Order and Spree::Product models.

payload_builder

To push the data, we need to configure the way on how to construct the JSON payload.

config.payload_builder = {
  "Spree::Order"  => {:serializer => "Spree::Pusher::OrderSerializer", :root => "orders"},
  "Spree::Product" => {:serializer => "Spree::Pusher::ProductSerializer", :root => "products"},
}

The payload builder is a hash, the key is the model name we also use in the push_objects config.

Each model has a serializer and a root field that defines the serializer we use to serialize to JSON and the root defines the root node for that JSON.

We have defined serializers for the default objects, you can find them here

To push other objects to an API endpoint, you only need to add an entry in the push_objects and the payload_builder configurations.

last_pushed_timestamps

For every model we push to an API endpoint, we keep track of when we pushed the objects.

Do not add this in config/initializers/pusher.rb otherwise it will reset the data on each restart.

Instead, if you need to reset data or want to update a timestamp for an object you can do so in the console

timestamps = Spree::Pusher::Config[:last_pushed_timestamps]
timestamps["Spree::Order"] = 2.days.ago
Spree::Pusher::Config[:last_pushed_timestamps] = timestamps

This will update the preference in the database and will use your updated timestamp for, in this case, 'Spree::Order'

Push to API endpoints

To push objects to API endpoints we provide you with the following rake task:

bundle exec rake pusher:push_it

This task will collect all the objects from push_objects that are not yet pushed (defined in last_pushed_timestamps) and will push those objects in batches of 10 to the specified API endpoint.

You could also add a background task to make that happen, all you need there are these lines:

Spree::Pusher::Config[:push_objects].each do |object|
  Spree::Pusher::Client.push_batches(object)
end

If you want to push Spree::Orders manually for example, you can call this:

Spree::Pusher::Client.push_batches("Spree::Order")

push_url

The default is https://push.wombat.co. You can override the default url to push your data to.

config.push_url = "http://mycustomurl"

You can also specify :push_url in the payload_builder

Create custom serializers

bundle exec rails g spree_pusher:serializer Spree::Order MyOrderSerializer

This will generate a serializer for the provided model name, when the model is already configured in the payload_builder we use that serializer name as super class to inherit from. With active_model_serializer you also inherit the attributes so you can keep the existing configuration and only change that what's needed.

The generator will also automatically set the correct configuration in config/initializers/pusher.rb

Testing

First bundle your dependencies, then run rake. rake will default to building the dummy app if it does not exist, then it will run specs. The dummy app can be regenerated by using rake test_app.

bundle
bundle exec rake

dcheung / spree_pusher