Match Kolibri's v2 contentnode and channel filters

Makefile

@@ -41,6 +41,8 @@ migrate:
 deploy-migrate:
 	# studio#5974: remove at cutover.
 	python contentcuration/manage.py backfill_column --model contentcuration.File --source-field file_size --target-field file_size_bigint
+	# studio#6171: remove after release.
+	python contentcuration/manage.py backfill_public_contentnode_modality
 
 contentnodegc:
 	python contentcuration/manage.py garbage_collect

contentcuration/kolibri_public/management/commands/backfill_public_contentnode_modality.py

@@ -0,0 +1,10 @@
+from django.core.management.base import BaseCommand
+from kolibri_public.models import ContentNode
+from kolibri_public.search import annotate_modality
+
+
+class Command(BaseCommand):
+    help = "Set kolibri_public ContentNode.modality from options.modality."
+
+    def handle(self, *args, **options):
+        annotate_modality(ContentNode.objects.all())

contentcuration/kolibri_public/migrations/0011_contentnode_modality.py

@@ -0,0 +1,39 @@
+# Generated by Django 3.2.24 on 2026-09-24 21:28
+from django.contrib.postgres.operations import AddIndexConcurrently
+from django.db import migrations
+from django.db import models
+
+
+class Migration(migrations.Migration):
+
+    atomic = False
+
+    dependencies = [
+        ("kolibri_public", "0010_localfile_file_size_bigint"),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name="contentnode",
+            name="modality",
+            field=models.CharField(
+                blank=True,
+                choices=[
+                    ("COURSE", "Course"),
+                    ("CUSTOM_NAVIGATION", "Custom Navigation"),
+                    ("LESSON", "Lesson"),
+                    ("QUIZ", "Quiz"),
+                    ("SURVEY", "Survey"),
+                    ("UNIT", "Unit"),
+                ],
+                max_length=50,
+                null=True,
+            ),
+        ),
+        AddIndexConcurrently(
+            model_name="contentnode",
+            index=models.Index(
+                fields=["modality"], name="kolibri_pub_modalit_1ae095_idx"
+            ),
+        ),
+    ]

contentcuration/kolibri_public/models.py

@@ -6,6 +6,7 @@
 from kolibri_public.search import contentnode_bitmask_fieldnames
 from kolibri_public.search import contentnode_metadata_bitmasks
 from kolibri_public.search import has_all_labels
+from le_utils.constants import modalities
 from mptt.managers import TreeManager
 from mptt.querysets import TreeQuerySet
 
@@ -53,9 +54,15 @@ class ContentNode(base_models.ContentNode):
     ancestors = JSONField(
         default=[], null=True, blank=True, load_kwargs={"strict": False}
     )
+    modality = models.CharField(
+        max_length=50, blank=True, null=True, choices=modalities.choices
+    )
 
     objects = ContentNodeManager()
 
+    class Meta:
+        indexes = [models.Index(fields=["modality"])]
+
 
 for field_name in contentnode_bitmask_fieldnames:
     field = models.BigIntegerField(default=0, null=True, blank=True)

contentcuration/kolibri_public/search.py

@@ -17,6 +17,7 @@
 from django.db.models import Max
 from django.db.models import Value
 from django.db.models import When
+from le_utils.constants import modalities
 from le_utils.constants.labels.accessibility_categories import (
     ACCESSIBILITYCATEGORIESLIST,
 )
@@ -219,6 +220,21 @@ def annotate_contentnode_label_bitmasks(queryset):
     return annotate_label_bitmasks(queryset, contentnode_bitmask_fieldnames)
 
 
+def annotate_modality(queryset):
+    """Update queryset to annotate `modality` field based on `options.modality`"""
+    queryset = queryset.filter(options__contains='"modality":')
+
+    when_statements = [
+        When(
+            options__contains=f'"modality": "{modality_value}"',
+            then=Value(modality_value),
+        )
+        for modality_value, _ in modalities.choices
+    ]
+
+    queryset.update(modality=Case(*when_statements))
+
+
 def annotate_channelmetadata_label_bitmasks(queryset):
     return annotate_label_bitmasks(queryset, channelmetadata_bitmask_fieldnames)
 

contentcuration/kolibri_public/tests/test_backfill_public_contentnode_modality.py

@@ -0,0 +1,31 @@
+import uuid
+
+from django.core.management import call_command
+from django.test import TestCase
+from kolibri_public import models
+from le_utils.constants import content_kinds
+from le_utils.constants import modalities
+
+
+class BackfillPublicContentNodeModalityTestCase(TestCase):
+    def _create_node(self, options):
+        return models.ContentNode.objects.create(
+            pk=uuid.uuid4().hex,
+            channel_id=uuid.uuid4().hex,
+            content_id=uuid.uuid4().hex,
+            kind=content_kinds.TOPIC,
+            title="node",
+            options=options,
+        )
+
+    def test_sets_modality_from_options(self):
+        quiz = self._create_node({"modality": modalities.QUIZ})
+        course = self._create_node({"modality": modalities.COURSE})
+        plain = self._create_node({})
+
+        call_command("backfill_public_contentnode_modality")
+
+        self.assertEqual(
+            dict(models.ContentNode.objects.values_list("id", "modality")),
+            {quiz.id: modalities.QUIZ, course.id: modalities.COURSE, plain.id: None},
+        )

contentcuration/kolibri_public/tests/test_content_app.py

@@ -13,6 +13,7 @@
 from kolibri_public.tests.base import ChannelBuilder
 from kolibri_public.tests.base import OKAY_TAG
 from le_utils.constants import content_kinds
+from le_utils.constants import modalities
 from rest_framework.test import APITestCase
 
 from contentcuration.models import generate_storage_url
@@ -165,6 +166,7 @@ def _assert_node(self, actual, expected):
                 "lft": expected.lft,
                 "rght": expected.rght,
                 "tree_id": expected.tree_id,
+                "modality": expected.modality,
                 "ancestors": [],
                 "tags": list(
                     expected.tags.all()
@@ -332,6 +334,137 @@ class ContentNodeAPITestCase(ContentNodeAPIBase, APITestCase):
     Testcase for content API methods
     """
 
+    def _list_ids(self, **params):
+        response = self._get(reverse("publiccontentnode-list"), data=params)
+        self.assertEqual(response.status_code, 200)
+        return {node["id"] for node in response.data}
+
+    def _available_ids(self):
+        return set(
+            models.ContentNode.objects.filter(available=True).values_list(
+                "id", flat=True
+            )
+        )
+
+    def _insert_second_tree(self):
+        # Its lft/rght ranges overlap the fixture tree's; only tree_id tells them apart.
+        ChannelBuilder(levels=2, num_children=2).insert_into_default_db()
+        models.ContentNode.objects.all().update(available=True)
+
+    def _set_modality(self, node, modality):
+        models.ContentNode.objects.filter(id=node.id).update(modality=modality)
+
+    def _mark_lesson_and_course(self):
+        lesson, course = (
+            models.ContentNode.objects.filter(kind=content_kinds.TOPIC)
+            .exclude(parent=None)
+            .order_by("lft")[:2]
+        )
+        self._set_modality(lesson, modalities.LESSON)
+        self._set_modality(course, modalities.COURSE)
+        return lesson, course
+
+    def test_contentnode_modality_filter(self):
+        lesson, _ = self._mark_lesson_and_course()
+
+        response = self.client.get(
+            reverse("publiccontentnode-list"), data={"modality": modalities.LESSON}
+        )
+
+        self.assertEqual([node["id"] for node in response.data], [lesson.id])
+        self.assertEqual(response.data[0]["modality"], modalities.LESSON)
+
+    def test_contentnode_exclude_modalities_filter(self):
+        lesson, course = self._mark_lesson_and_course()
+
+        self.assertEqual(
+            self._list_ids(exclude_modalities=modalities.COURSE),
+            self._available_ids() - {course.id},
+        )
+        self.assertEqual(
+            self._list_ids(
+                exclude_modalities="{},{}".format(modalities.COURSE, modalities.LESSON)
+            ),
+            self._available_ids() - {course.id, lesson.id},
+        )
+
+    def test_contentnode_exclude_course_ancestry_filter(self):
+        course = self.root.get_children().first()
+        self._set_modality(course, modalities.COURSE)
+        self._insert_second_tree()
+        descendant_ids = set(course.get_descendants().values_list("id", flat=True))
+
+        self.assertEqual(
+            self._list_ids(exclude_course_ancestry=True),
+            self._available_ids() - descendant_ids,
+        )
+        self.assertEqual(
+            self._list_ids(exclude_course_ancestry=False), self._available_ids()
+        )
+
+    def test_contentnode_contains_quiz_filter(self):
+        quiz = models.ContentNode.objects.exclude(kind=content_kinds.TOPIC).first()
+        self._set_modality(quiz, modalities.QUIZ)
+        self._insert_second_tree()
+
+        self.assertEqual(
+            self._list_ids(contains_quiz=True),
+            set(quiz.get_ancestors(include_self=True).values_list("id", flat=True)),
+        )
+        self.assertEqual(self._list_ids(contains_quiz="false"), self._available_ids())
+
+    def _set_search_fixtures(self):
+        photo, resp = models.ContentNode.objects.exclude(
+            kind=content_kinds.TOPIC
+        ).order_by("lft")[:2]
+        models.ContentNode.objects.filter(id=photo.id).update(
+            title="Photosynthesis basics", description="Chlorophyll absorbs light"
+        )
+        models.ContentNode.objects.filter(id=resp.id).update(
+            title="Energy", description="Cellular respiration"
+        )
+        return photo.id, resp.id
+
+    def test_contentnode_search_terms_match_across_fields(self):
+        photo, _ = self._set_search_fixtures()
+        self.assertEqual(self._list_ids(search="photosynthesis chlorophyll"), {photo})
+
+    def test_contentnode_search_every_term_must_match(self):
+        self._set_search_fixtures()
+        self.assertEqual(self._list_ids(search="photosynthesis respiration"), set())
+
+    def test_contentnode_search_quoted_phrase(self):
+        photo, _ = self._set_search_fixtures()
+        self.assertEqual(self._list_ids(search='"photosynthesis basics"'), {photo})
+        self.assertEqual(self._list_ids(search='"basics photosynthesis"'), set())
+
+    def test_contentnode_search_drops_stopwords(self):
+        _, resp = self._set_search_fixtures()
+        self.assertEqual(self._list_ids(search="which respiration"), {resp})
+
+    def test_contentnode_search_keywords_param(self):
+        _, resp = self._set_search_fixtures()
+        self.assertEqual(self._list_ids(keywords="respiration"), {resp})
+
+    def test_contentnode_search_precedence(self):
+        photo, _ = self._set_search_fixtures()
+        self.assertEqual(
+            self._list_ids(
+                search="photosynthesis",
+                question="respiration",
+                keywords="respiration",
+            ),
+            {photo},
+        )
+        self.assertEqual(
+            self._list_ids(question="photosynthesis", keywords="respiration"),
+            {photo},
+        )
+
+    def test_contentnode_search_punctuation_only_is_unfiltered(self):
+        self.assertEqual(self._list_ids(search="!?,"), self._available_ids())
+        self.assertEqual(self._list_ids(keywords="!?,"), self._available_ids())
+
     def test_prerequisite_for_filter(self):
         response = self.client.get(
             reverse("publiccontentnode-list"),
@@ -466,7 +599,7 @@ def test_channelmetadata_content_available_field_false(self):
         response = self.client.get(reverse("publicchannel-list"))
         self.assertEqual(response.data[0]["available"], False)
 
-    def test_channelmetadata_has_exercises_filter(self):
+    def test_channelmetadata_exercise_filters(self):
         # Has nothing else for that matter...
         no_exercise_channel = models.ContentNode.objects.create(
             pk="6a406ac66b224106aa2e93f73a94333d",
@@ -492,13 +625,12 @@ def test_channelmetadata_has_exercises_filter(self):
         )
         no_filter_response = self.client.get(reverse("publicchannel-list"))
         self.assertEqual(len(no_filter_response.data), 2)
-        with_filter_response = self.client.get(
-            reverse("publicchannel-list"), {"has_exercise": True}
-        )
-        self.assertEqual(len(with_filter_response.data), 1)
-        self.assertEqual(
-            with_filter_response.data[0]["name"], self.channel_data["name"]
-        )
+        for param in ("has_exercise", "contains_exercise"):
+            with self.subTest(param=param):
+                response = self.client.get(reverse("publicchannel-list"), {param: True})
+                self.assertEqual(
+                    [c["name"] for c in response.data], [self.channel_data["name"]]
+                )
 
     def test_channelmetadata_public_filter_default_true(self):
         community_channel = models.ContentNode.objects.create(

contentcuration/kolibri_public/tests/test_mapper.py

@@ -15,6 +15,7 @@
 from kolibri_public.tests.base import OKAY_TAG
 from kolibri_public.utils.mapper import ChannelMapper
 from le_utils.constants import content_kinds
+from le_utils.constants import modalities
 from le_utils.constants.labels.subjects import SUBJECTSLIST
 
 from contentcuration.models import Channel
@@ -290,6 +291,26 @@ def test_categories_bitmask_annotation(self):
             )
             self.assertEqual(mapper.mapped_channel.categories_bitmask_0, 1 | 4 | 16)
 
+    def test_modality_annotation(self):
+        with using_content_database(self.tempdb):
+            source = (
+                kolibri_content_models.ContentNode.objects.filter(
+                    channel_id=self.channel.id, kind=content_kinds.TOPIC
+                )
+                .exclude(parent=None)
+                .first()
+            )
+            source.options = {"modality": modalities.COURSE}
+            source.save()
+
+            ChannelMapper(self.channel).run()
+
+            mapped = kolibri_public_models.ContentNode.objects.all()
+            self.assertEqual(mapped.get(id=source.id).modality, modalities.COURSE)
+            self.assertFalse(
+                mapped.exclude(id=source.id).filter(modality__isnull=False).exists()
+            )
+
     def tearDown(self):
         # Clean up datbase connection after the test
         self._date_patcher.stop()

contentcuration/kolibri_public/utils/mapper.py

@@ -4,6 +4,7 @@
 from kolibri_content.base_models import MAX_TAG_LENGTH
 from kolibri_public import models as kolibri_public_models
 from kolibri_public.search import annotate_contentnode_label_bitmasks
+from kolibri_public.search import annotate_modality
 from kolibri_public.utils.annotation import set_channel_metadata_fields
 from le_utils.constants import content_kinds
 
@@ -69,9 +70,9 @@ def run(self):
             self.mapped_channel.public = self.public
             self.mapped_channel.save_base(raw=True)
 
-            annotate_contentnode_label_bitmasks(
-                self.mapped_root.get_descendants(include_self=True)
-            )
+            mapped_nodes = self.mapped_root.get_descendants(include_self=True)
+            annotate_contentnode_label_bitmasks(mapped_nodes)
+            annotate_modality(mapped_nodes)
             # Rather than set the ancestors fields after mapping, like it is done in Kolibri
             # here we set it during mapping as we are already recursing through the tree.