Merge pull request #2910 from paperless-ngx/feature-improved-statistics-widget

Feature: Improved statistics widget

src/documents/tests/test_api.py

@@ -1039,9 +1039,24 @@ class TestDocumentApi(DirectoriesMixin, APITestCase):
 
     def test_statistics(self):
 
-        doc1 = Document.objects.create(title="none1", checksum="A")
-        doc2 = Document.objects.create(title="none2", checksum="B")
-        doc3 = Document.objects.create(title="none3", checksum="C")
+        doc1 = Document.objects.create(
+            title="none1",
+            checksum="A",
+            mime_type="application/pdf",
+            content="abc",
+        )
+        doc2 = Document.objects.create(
+            title="none2",
+            checksum="B",
+            mime_type="application/pdf",
+            content="123",
+        )
+        doc3 = Document.objects.create(
+            title="none3",
+            checksum="C",
+            mime_type="text/plain",
+            content="hello",
+        )
 
         tag_inbox = Tag.objects.create(name="t1", is_inbox_tag=True)
 
@@ -1051,6 +1066,16 @@ class TestDocumentApi(DirectoriesMixin, APITestCase):
         self.assertEqual(response.status_code, status.HTTP_200_OK)
         self.assertEqual(response.data["documents_total"], 3)
         self.assertEqual(response.data["documents_inbox"], 1)
+        self.assertEqual(response.data["inbox_tag"], tag_inbox.pk)
+        self.assertEqual(
+            response.data["document_file_type_counts"][0]["mime_type_count"],
+            2,
+        )
+        self.assertEqual(
+            response.data["document_file_type_counts"][1]["mime_type_count"],
+            1,
+        )
+        self.assertEqual(response.data["character_count"], 11)
 
     def test_statistics_no_inbox_tag(self):
         Document.objects.create(title="none1", checksum="A")
@@ -1058,6 +1083,7 @@ class TestDocumentApi(DirectoriesMixin, APITestCase):
         response = self.client.get("/api/statistics/")
         self.assertEqual(response.status_code, status.HTTP_200_OK)
         self.assertEqual(response.data["documents_inbox"], None)
+        self.assertEqual(response.data["inbox_tag"], None)
 
     @mock.patch("documents.views.consume_file.delay")
     def test_upload(self, m):

src/documents/views.py

@@ -20,7 +20,9 @@ from django.db.models import Case
 from django.db.models import Count
 from django.db.models import IntegerField
 from django.db.models import Max
+from django.db.models import Sum
 from django.db.models import When
+from django.db.models.functions import Length
 from django.db.models.functions import Lower
 from django.http import Http404
 from django.http import HttpResponse
@@ -186,6 +188,7 @@ class TagViewSet(ModelViewSet, PassUserMixin):
     )
 
     def get_serializer_class(self, *args, **kwargs):
+        print(self.request.version)
         if int(self.request.version) == 1:
             return TagSerializerVersion1
         else:
@@ -795,17 +798,38 @@ class StatisticsView(APIView):
 
     def get(self, request, format=None):
         documents_total = Document.objects.all().count()
-        if Tag.objects.filter(is_inbox_tag=True).exists():
-            documents_inbox = (
-                Document.objects.filter(tags__is_inbox_tag=True).distinct().count()
+
+        inbox_tag = Tag.objects.filter(is_inbox_tag=True)
+
+        documents_inbox = (
+            Document.objects.filter(tags__is_inbox_tag=True).distinct().count()
+            if inbox_tag.exists()
+            else None
+        )
+
+        document_file_type_counts = (
+            Document.objects.values("mime_type")
+            .annotate(mime_type_count=Count("mime_type"))
+            .order_by("-mime_type_count")
+            if documents_total > 0
+            else 0
+        )
+
+        character_count = (
+            Document.objects.annotate(
+                characters=Length("content"),
             )
-        else:
-            documents_inbox = None
+            .aggregate(Sum("characters"))
+            .get("characters__sum")
+        )
 
         return Response(
             {
                 "documents_total": documents_total,
                 "documents_inbox": documents_inbox,
+                "inbox_tag": inbox_tag.first().pk if inbox_tag.exists() else None,
+                "document_file_type_counts": document_file_type_counts,
+                "character_count": character_count,
             },
         )