feat(filters): add "only_newest" filter for files

This helps selecting files when we're searching (or otherwise looking
for files) and only want the newest version

alexandria/core/filters.py

@@ -4,7 +4,7 @@
 
 from django.conf import settings
 from django.contrib.postgres.search import SearchHeadline, SearchQuery, SearchRank
-from django.db.models import F, FloatField, Q, TextField, Value
+from django.db.models import Exists, F, FloatField, OuterRef, Q, TextField, Value
 from django.db.models.fields.json import KeyTextTransform
 from django.db.models.functions import Cast
 from django_filters import (
@@ -169,17 +169,43 @@ class FileFilterSet(FilterSet):
     document_metainfo = JSONValueFilter(field_name="document__metainfo")
     active_group = ActiveGroupFilter()
     files = BaseCSVFilter(field_name="pk", lookup_expr="in")
+    only_newest = BooleanFilter(method="filter_only_newest")
+
+    def filter_only_newest(self, qs, name, value):
+        if value:
+            return qs.exclude(
+                Exists(
+                    # Find all newer versions of a file. A newer version is a file:
+                    #  - of the same variant
+                    #  - with a different ID (obviously)
+                    #  - a newer created_at timestamp
+                    #  - AND the same document
+                    models.File.objects.all()
+                    .filter(
+                        document=OuterRef("document_id"),
+                        created_at__gt=OuterRef("created_at"),
+                        variant=OuterRef("variant"),
+                    )
+                    .exclude(pk=OuterRef("pk")),
+                )
+            )
+        return qs
 
     class Meta:
         model = models.File
-        fields = ["original", "renderings", "variant", "metainfo", "files"]
+        fields = [
+            "original",
+            "renderings",
+            "variant",
+            "metainfo",
+            "files",
+            "only_newest",
+        ]
 
 
 class SearchFilterSet(FileFilterSet):
     query = CharFilter(method="search_files")
 
-    only_newest = BooleanFilter(method="filter_only_newest")
-
     # Cut-off value, we don't want all the garbage matches
     min_rank = NumberFilter(field_name="search_rank", lookup_expr="gte")
 

alexandria/core/tests/test_filters.py

@@ -1,4 +1,5 @@
 import json
+import time
 from itertools import combinations
 from typing import Optional
 
@@ -344,3 +345,40 @@ def test_document_category_filters(
     data = response.json()["data"]
     assert len(data) == expected_count
     assert sorted([doc["attributes"]["title"] for doc in data]) == snapshot
+
+
+@pytest.mark.parametrize(
+    "filter_val, expect_v1, expect_v2",
+    [
+        (True, False, True),
+        (False, True, True),
+        (None, True, True),
+    ],
+)
+def test_only_newest_filter(
+    db, admin_client, document_factory, file_factory, filter_val, expect_v1, expect_v2
+):
+    doc = document_factory()
+    version_1 = file_factory(document=doc, variant="original")
+    time.sleep(0.5)
+    version_2 = file_factory(document=doc, variant="original")
+
+    # original and thumbnail (?)
+    assert doc.files.all().count() == 4
+
+    filters = {
+        "filter[variant]": "original",
+    }
+    if filter_val is not None:
+        filters["filter[only_newest]"] = filter_val
+
+    response = admin_client.get(reverse("file-list"), filters)
+
+    assert response.status_code == HTTP_200_OK, response.json()
+    received_file_ids = [f["id"] for f in response.json()["data"]]
+
+    has_v1 = str(version_1.pk) in received_file_ids
+    has_v2 = str(version_2.pk) in received_file_ids
+
+    assert has_v1 == expect_v1
+    assert has_v2 == expect_v2