Skip to main content

Step 3: Split Nested Structures

The "Store -> Split -> Restore" pattern you learned in Step 1 works for any JSON structure, no matter how complex. This step demonstrates how to apply that same pattern to a more realistic, nested e-commerce order.

The Goal​

You will transform a single, complex order object into multiple line_item messages, each containing the relevant context from the parent order.

Input: E-commerce Order
{
  "order_id": "order-12345",
  "customer": {
    "customer_id": "customer-001",
    "email": "[email protected]"
  },
  "items": [
    {
      "sku": "WIDGET-A",
      "quantity": 5,
      "unit_price": 19.99
    },
    {
      "sku": "GADGET-B",
      "quantity": 2,
      "unit_price": 49.99
    }
  ]
}

Implementation​

The pattern is exactly the same as before, you just need to store more context from the nested structure.

  1. Start with the Foundation: Copy the order-processing-foundation.yaml file to a new file named order-splitter.yaml.

    cp examples/data-routing/order-processing-foundation.yaml order-splitter.yaml

  2. Add the Splitting Logic: Open order-splitter.yaml and replace the entire pipeline section with the three-processor block below.

pipeline: processors:

1. STORE context from the nested structure

  • mapping: | meta order_id = this.order_id meta customer_id = this.customer.customer_id meta customer_email = this.customer.email root = this

2. SPLIT the 'items' array

  • unarchive: format: json_array field: items

3. RESTORE the context into each new line item message

  • mapping: | root = this root.order_id = meta("order_id") root.customer_id = meta("customer_id") root.customer_email = meta("customer_email")

3.  **Deploy and Test:**
```bash
# Send the test order
curl -X POST http://localhost:8080/orders \
  -H "Content-Type: application/json" \
  -d '{ 
    "order_id": "order-12345",
    "customer": { "customer_id": "customer-001", "email": "[email protected]" },
    "items": [
      { "sku": "WIDGET-A", "quantity": 5, "unit_price": 19.99 },
      { "sku": "GADGET-B", "quantity": 2, "unit_price": 49.99 }
    ]
  }'
  1. Verify: Check your output. You will see two separate line_item messages, each containing the order_id and customer details, demonstrating that you have successfully preserved context from a nested structure.

This scalable pattern is the key to handling complex, hierarchical data.