Merge pull request #224 from American-Institutes-for-Research/HEA-872/Add-add-cache_memoize-and-etag-decorators

Hea 872/add add cache memoize and etag decorators

Author: girum-air

apps/baseline/tests/test_viewsets.py

@@ -2,13 +2,17 @@
 import json
 import logging
 import warnings
+from datetime import timedelta
 from io import StringIO
 
 import pandas as pd
 from bs4 import BeautifulSoup
 from django.contrib.auth.models import User
+from django.core.cache import cache
 from django.db.models import F
 from django.urls import reverse
+from django.utils.http import http_date
+from django.utils.timezone import now
 from rest_framework.test import APITestCase
 
 from baseline.models import (
@@ -677,6 +681,43 @@ def test_as_of_date_filter_returns_valid_baselines(self):
         self.assertIn(baseline_no_to.id, baseline_ids)
         self.assertIn(baseline_no_dates.id, baseline_ids)
 
+    def test_conditional_request_headers(self):
+        cache.clear()  # Clear cache to ensure clean state
+
+        # Test that 200 response includes ETag, Last-Modified, Cache-Control, and Expires headers
+        response = self.client.get(self.url)
+        self.assertEqual(response.status_code, 200)
+        self.assertIn("ETag", response.headers)
+        self.assertIn("Last-Modified", response.headers)
+        self.assertIn("Cache-Control", response.headers)
+        self.assertIn("Expires", response.headers)
+        self.assertTrue(response.headers["ETag"].startswith('W/"'))  # Weak ETag format
+
+        # Test If-None-Match returns 304 when not modified
+        etag = response.headers["ETag"]
+        cache.clear()
+        response = self.client.get(self.url, HTTP_IF_NONE_MATCH=etag)
+        self.assertEqual(response.status_code, 304)
+        self.assertIn("Cache-Control", response.headers)
+        self.assertIn("Expires", response.headers)
+
+        # Test If-None-Match returns 200 when data is modified
+        cache.clear()  # Clear cache before testing modified data
+        baseline = self.data[0]
+        baseline.population_source = "Updated source"
+        baseline.save()
+        response = self.client.get(self.url, HTTP_IF_NONE_MATCH=etag)
+        self.assertEqual(response.status_code, 200)
+        self.assertNotEqual(response.headers["ETag"], etag)
+
+        # Test If-Modified-Since with future date returns 304
+        cache.clear()
+        future_date = http_date((now() + timedelta(days=1)).timestamp())
+        response = self.client.get(self.url, HTTP_IF_MODIFIED_SINCE=future_date)
+        self.assertEqual(response.status_code, 304)
+        self.assertIn("Cache-Control", response.headers)
+        self.assertIn("Expires", response.headers)
+
 
 class LivelihoodZoneBaselineFacetedSearchViewTestCase(APITestCase):
     def setUp(self):

apps/baseline/viewsets.py

@@ -4,7 +4,10 @@
 from django.db.models import Expression, F, Q, Subquery, TextField, Value
 from django.db.models.functions import Coalesce, NullIf
 from django.utils import translation
+from django.utils.decorators import method_decorator
 from django.utils.translation import override
+from django.views.decorators.cache import cache_page
+from django.views.decorators.http import condition
 from django_filters import rest_framework as filters
 from django_filters.filters import CharFilter
 from rest_framework.permissions import AllowAny
@@ -14,6 +17,7 @@
 
 from common.fields import translation_fields
 from common.filters import DefaultingDateFilter, MultiFieldFilter, UpperCaseFilter
+from common.utils import make_condition_funcs
 from common.viewsets import AggregatingViewSet, BaseModelViewSet
 from metadata.models import WealthGroupCategory
 
@@ -100,6 +104,9 @@
     WildFoodGatheringSerializer,
 )
 
+# Create condition functions for LivelihoodZoneBaseline endpoint caching
+get_baseline_etag, get_baseline_last_modified = make_condition_funcs(LivelihoodZoneBaseline)
+
 
 class SourceOrganizationFilterSet(filters.FilterSet):
     class Meta:
@@ -275,6 +282,11 @@ def get_serializer_class(self):
             return LivelihoodZoneBaselineGeoSerializer  # Use GeoFeatureModelSerializer for GeoJSON
         return LivelihoodZoneBaselineSerializer
 
+    @method_decorator(cache_page(60 * 60 * 24))  # Cache on server for 24 hours - must be above condition per RFC 9110
+    @method_decorator(condition(etag_func=get_baseline_etag, last_modified_func=get_baseline_last_modified))
+    def list(self, request, *args, **kwargs):
+        return super().list(request, *args, **kwargs)
+
 
 class LivelihoodProductCategoryFilterSet(filters.FilterSet):
     class Meta:

apps/common/utils.py

@@ -1,5 +1,6 @@
 import contextlib
 import csv
+import hashlib
 import importlib
 import logging
 import re
@@ -12,6 +13,7 @@
 from dateutil.relativedelta import relativedelta
 from django.apps import apps
 from django.db.migrations.operations.base import Operation
+from django.db.models import Max
 from django.forms.models import modelform_factory
 from openpyxl.utils import get_column_letter
 from treebeard.mp_tree import MP_Node
@@ -368,3 +370,29 @@ def get_all_subclasses(cls):
 # Minimal normalization, doesn't attempt to coerce characters such as æ, which are locale-dependent.
 # Example usage: "café".translate(normalize)
 normalize = str.maketrans(normal_map)
+
+
+def make_condition_funcs(model):
+    """
+    Create etag and last_modfied functions for use with Django's condition decorator.
+    All models inheriting from common.Model have a 'modified' timestamp field from TimeStampedModel.
+    """
+
+    def get_last_modified(request, *args, **kwargs):
+        # Return the most recent modification timestamp for the model.
+        result = model.objects.aggregate(last_modified=Max("modified"))
+        return result["last_modified"]
+
+    def get_etag(request, *args, **kwargs):
+        """
+        Return a weak ETag based on the most recent modification timestamp.
+
+        Uses weak ETag (W/ prefix) because we're comparing semantic equivalence of the data,
+        not byte-for-byte equality. See https://developer.mozilla.org/en-US/docs/Web/HTTP/Reference/Headers/ETag
+        """
+        last_modified = get_last_modified(request, *args, **kwargs)
+        if last_modified is None:
+            return None
+        return f'W/"{hashlib.md5(last_modified.isoformat().encode()).hexdigest()}"'
+
+    return get_etag, get_last_modified