886dcbc82f
* add batch job strategy interface * update plugin loaders * remove comment * make map async * ensure that only one of each strategy is registered * register strategies plural * add identifier and batchType properties * extend batch job strategy identification method * initial test * update loaders to accomodate different ways of accessing batch job strategies * identifier batch type field * redo merge in plugins * update interface and load only js files * use switches instead of elif * remove comments * use static properties for strategy registration * update tests * fix unit tests * update test names * update isBatchJobStrategy method * add check for TransactionBaseService in services for plugins * update interfaces export * update batchjob strategy interface with a prepare script * update loaders * update batchjob strategy interface * remove everything but public interface methods from batchJobStrategy * add default implementation to prepareBathJobForProcessing * remove unused import
Structure
-
Models (
/models) This is where the data layer lives. Define data models here no logic only schema and data access layer. (Default is MongoDB so we have data access layer defined for us already) -
Services (
/services) This is where our business logic lives. Define services that perform calculations, update the data layer, synchronize services, etc. -
Controllers (
/api) This is the interface lives. Define how the user interacts with the service layer. Ensure that the user has permission to do what they intend to, authenticate requests, call service layer. -
Jobs (
/jobs) This is where background and recurring tasks live. Want to send some data somewhere every night, this would be where to do it. Calls service layer methods and should, like controllers, not contain business logic. -
Subscribers (
/subscribers) This is where events live. Want to perform a certain task whenever something else happens, this is where to do it.
Extending the core
The core will look for files in the folders listed above, and inject the custom code.
Checkout flow
To create an order from a cart the customer must have filled in:
- their details (shipping/billing address, email)
- shipping method
- payment method
The steps can be done in any order. The standard path would probably be:
- submit details (PUT /cart/shipping-address, PUT /cart/email)
- select shipping method (PUT /cart/shipping-method)
- enter payment details (PUT /cart/payment-method)
- complete order (POST /order)
Assuming that shipping methods are static within each region we can display all shipping methods at checkout time. If shipping is dynamically calculated the price of the shipping method may change, we will ask the fulfillment provider for new rates.
Payment details can be entered at any point as long as the final amount is known. If the final amount changes after the payment details are entered the payment method may therefore be invalidated.
Within the store UI you could imagine each step being taken care of by a single button click, which calls all endpoints.