Move the metadata retrieval to an api function, add a parser for read…

…able ids, fall back to initial data

- The metadata retrieval is now in an api function so it's easier to test.
- Added a parser for readable ID - skus are generally readable IDs, but if we have a sku that's for a specific run, the learning_resources API won't know what to do with it.
- If the data returned from the API doesn't contain data, or doesn't contain _some_ data, fall back to the data that was in the product already.
- Added tests for API function
- Fixed some names (we're not just pulling image data now)

system_meta/api.py

@@ -0,0 +1,65 @@
+"""API functions for system metadata."""
+
+import logging
+
+import requests
+from django.conf import settings
+
+from system_meta.models import Product
+from unified_ecommerce.utils import parse_readable_id
+
+log = logging.getLogger(__name__)
+
+
+def update_product_metadata(product_id: int) -> None:
+    """Get product metadata from the Learn API."""
+
+    try:
+        product = Product.objects.get(id=product_id)
+        resource, run = parse_readable_id(product.sku)
+        response = requests.get(
+            f"{settings.MITOL_LEARN_API_URL}learning_resources/",
+            params={"platform": product.system.slug, "readable_id": resource},
+            timeout=10,
+        )
+        response.raise_for_status()
+        results_data = response.json()
+
+        if results_data.get("count", 0) == 0:
+            log.warning("No Learn results found for product %s", product)
+            return
+
+        course_data = results_data.get("results")[0]
+
+        image_data = course_data.get("image")
+        product.image_metadata = (
+            {
+                "image_url": image_data.get("url"),
+                "alt_text": image_data.get("alt"),
+                "description": image_data.get("description"),
+            }
+            if image_data
+            else product.image_metadata
+        )
+
+        product.name = course_data.get("title", product.name)
+        product.description = course_data.get("description", product.description)
+
+        # If there are runs, we'll overwrite this with the run's price later.
+        product_prices = course_data.get("prices", [])
+        product_prices.sort()
+        product.price = product_prices.pop() if len(product_prices) else product.price
+
+        runs = course_data.get("runs")
+        if runs:
+            run = next((r for r in runs if r.get("run_id") == product.sku), None)
+            if run:
+                product_prices = run.get("prices", [])
+                product_prices.sort()
+                product.price = (
+                    product_prices.pop() if len(product_prices) else product.price
+                )
+
+        product.save()
+    except requests.RequestException:
+        log.exception("Failed to update metadata for product %s", product.id)

system_meta/api_test.py

@@ -0,0 +1,50 @@
+"""Tests for the system_meta APIs."""
+
+import pytest
+
+from system_meta.api import update_product_metadata
+from system_meta.factories import ProductFactory
+
+pytestmark = pytest.mark.django_db
+
+
+def test_retrieve_product_metadata(mocker):
+    """Test that the retrieve_product_metadata function works."""
+
+    mocked_requests = mocker.patch("requests.get")
+    mocked_requests.return_value.json.return_value = {
+        "results": [
+            {
+                "image": {
+                    "id": 12345,
+                    "url": "https://example.com/image.jpg",
+                    "alt": "Example image",
+                    "description": "Example description",
+                },
+                "title": "Example title",
+                "description": "Example description",
+                "prices": [100, 200],
+                "readable_id": "course-v1:MITx+12.345x",
+                "runs": [
+                    {
+                        "run_id": "123",
+                        "prices": [150, 250],
+                        "readable_id": "course-v1:MITx+12.345x+2T2099",
+                    }
+                ],
+            }
+        ]
+    }
+
+    product = ProductFactory.create(
+        sku="course-v1:MITx+12.345x+2T2099",
+        price=50,
+        description="This is the wrong description.",
+    )
+    update_product_metadata(product.id)
+    product.refresh_from_db()
+    assert product.image_metadata is not None
+    assert product.name == "Example title"
+    assert product.description == "Example description"
+    # This has a run price, so we should have that, and it should be the highest price.
+    assert product.price == 250

system_meta/management/commands/manage_product.py

@@ -195,7 +195,7 @@ def _add_product(
             exception_message = "Product {sku} already exists in system {system_name}."
             raise CommandError(exception_message, returncode=2)
 
-        system = IntegratedSystem.objects.get(name=system_name)
+        system = IntegratedSystem.objects.get(slug=system_name)
         product = Product.objects.create(
             sku=sku,
             name=name,

system_meta/tasks.py

@@ -1,40 +1,33 @@
+"""Tasks for the system_meta app."""
+
 import logging
 from typing import Optional
 
 import requests
 from celery import shared_task
-from django.conf import settings
 
 
 @shared_task
 def update_products(product_id: Optional[int] = None):
     """
-    Update all product's image metadata.  If product_id is provided, only update the
-    product with that ID.  Otherwise, update all products.
+    Update product metadata from the Learn API.
+
+    Updates all products if a product_id is not provided. Pulls the image metadata,
+    name, and description from the Learn API. If the product has a run ID, it also
+    pulls the price from the specific run; otherwise, pulls the price from the
+    resource.
     """
-    from .models import Product
+    from system_meta.api import update_product_metadata
+    from system_meta.models import Product
 
     log = logging.getLogger(__name__)
     if product_id:
         products = Product.objects.filter(id=product_id)
     else:
         products = Product.objects.all()
+
     for product in products:
         try:
-            response = requests.get(
-                f"{settings.MITOL_LEARN_API_URL}learning_resources/",
-                params={"platform": product.system.slug, "readable_id": product.sku},
-                timeout=10,
-            )
-            response.raise_for_status()
-            results_data = response.json()
-            course_data = results_data.get("results")[0]
-            image_data = course_data.get("image")
-            product.image_metadata = {
-                "image_url": image_data.get("url"),
-                "alt_text": image_data.get("alt"),
-                "description": image_data.get("description"),
-            }
-            product.save()
+            update_product_metadata(product.id)
         except requests.RequestException:
-            log.exception("Failed to retrieve image data for product %s", product.id)
+            log.exception("Failed to update metdata for product %s", product.id)

unified_ecommerce/utils.py

@@ -445,3 +445,26 @@ def get_user_from_apisix_headers(request):
         user.refresh_from_db()
 
     return user
+
+
+def parse_readable_id(readable_id: str) -> tuple[str, str]:
+    """
+    Parse a readable ID into a resource ID and a run ID.
+
+    Readable IDs look like "course-v1:MITxT+12.345x" but they may also have a run
+    tacked onto the end ("+1T2024" for instance). If the readable ID isn't for a
+    run of the resource, you'll get a None in the run position.
+
+    Args:
+        readable_id (str): The readable ID to parse
+
+    Returns:
+        tuple[str, str]: The resource ID and the run ID (or None)
+    """
+    if readable_id.count("+") > 1:
+        resource, run = readable_id.rsplit("+", 1)
+    else:
+        resource = readable_id
+        run = None
+
+    return resource, run