Download the Crafter Commerce source code
Building and testing Crafter Commerce

Introduction

Crafter Commerce will be a new extension to Crafter 2. It will be an online commerce engine that facilitates guided shopping experiences within an extensive product catalog with rich metadata. It will also interface with billing engines, payment gateways, CRM and analytics platforms. Crafter Commerce will be build around four main pillars, as shown below:

Features

Here is an overview of features that Crafter Commerce will support.

Account managent

Customer profiles
Opt-in/opt-out
Address cleansing & validation

Pricing management

Promotion codes
Royalty/referer codes
Quantity discount calculations
Promotion models (i.e. Buy one, get one free)

Order management

Quotes
- Create (taking into consideration shipping and tax)
- Confirm
- Update

Orders
- History
- Status
- Cancelations
- Returns

List Management

Payment types
Credit card types
Person titles
Shipping methods
States

Security

Authentication
Authorization

Implementation

Crafter Commerce is implemented as a Server and a Client component, which communicate with each other over HTTP through JSON or XML messages. The Server component is build as a Java web application with Spring, Mongo, Hibernate Validator, SolrJ as dependencies. The Client component is build as a Java library that can be embedded in any Java (web) application. It uses Spring REST support to interact with the Server over HTTP and perform marshalling and unmarshalling.

Entity Relationships

Mongo Schema Design

collections:

accounts
products
1. promotions
2. product-prices
orders
1. payments
2. order-line-items
3. order-delivery
payment-gateways
shipping-methods
locations
addresses
attributes

atomic operations:

In Mongo, atomic operations can be performed only on a single BSON document.

Needed when:

Adding a product to a shopping cart (if we manage inventory, so updating a product's available quantity should be atomic).
...

indexes:

Identify the fields to index / query on

id (all collections)
key-name(s) (coll-name)
key-name(s) (coll-name)

sharding:

Not applicable at this time

Promotions

Percentage off per item
Percentage off all
Fixed amount off per item
Fixed amount off for all
Buy X, get one or more items at a discount
Free gift with purchase
Free shipping
Discounted shipping for an order using a selected ship mode
Discounted shipping for all items using a selected ship mode
Discounted shipping per item using a selected ship mode

Open Questions

What persistence store to use?
How to maintain a Crafter Commerce transaction?
What do do with abandoned shopping carts?
Analytics integration?
What user information can we store without having to be PCI compliant?

Download and build it

Prerequisites

The following is a list of prerequisites that Crafter Commerce depends on in order to build and run:

Java 1.5
Maven 2
An SVN client
Apache Solr 3.6.0 (not needed for build)
Mongo DB 2
A decent servlet container such as Jetty, Tomcat or JBoss

Technically Mongo and Solr are not required for building, but Mongo is required for testing as an extension to the build. We're using the Maven Failsafe plugin to run integration tests, for which at least Mongo is assumed to be running at the time of building. For integration tests specifically, Mongo is expected to be run on default port 27017 with no authentication required.

Prerequisites: Mongo and Solr

Mongo

Mongo can be downloaded here: http://www.mongodb.org/downloads. At the time of writing, the latest version was 2.0.5. There are versions for OS X, Linux, Windows and Solaris.

Once you have Mongo installed, start it as follows:

mongod --dbpath=/path/to/mongo/data/db

While it's starting, it'll log to the console the default port on which it will be listening:

[initandlisten] waiting for connections on port 27017

Solr

Apache Solr can be downloaded here: http://lucene.apache.org/solr/downloads.html. The latest version at this moment is 3.6, which is the exact dependency for Crafter Commerce, although older versions may work as well.

Download the binary distribution from one of the Apache mirrors, for example: {{ apache-solr-3.6.0.tgz}}. Then, unpack it and copy the reference schema from the Crafter Commerce source into the Solr example directory:

cd /your/path/to/apache-solr-3.5.0/example/conf
mv schema.xml to schema.xml.ORIG
cp /your/path/to/crafter-commerce/crafter-commerce-itest/solr/conf/schema.xml .

Lastly, startup Solr with embedded Jetty as follows:

cd /your/path/to/apache-solr-3.5.0/example
java -jar start.jar

Jetty will report to the console on which port it is listening for HTTP requests:

INFO::Started SocketConnector@0.0.0.0:8983

To double-check that Solr is indeed running on that port, fire up a browser, and try the following URL:

http://localhost:8983/solr/

Download the Crafter Commerce source code

Use your SVN client to checkout a copy of the source code. From the command line it would look like this:

svn co [https://svn.rivetlogic.com/repos/crafter-community/commerce/trunk crafter-commerce]

Make sure that Solr isn't started on that port while running the integration tests.

Crafter Commerce