Merge branch 'dev' into feature-ai

src/documents/serialisers.py

@@ -1041,7 +1041,7 @@ class DocumentSerializer(
             request.version if request else settings.REST_FRAMEWORK["DEFAULT_VERSION"],
         )
 
-        if api_version < 9:
+        if api_version < 9 and "created" in self.fields:
             # provide created as a datetime for backwards compatibility
             from django.utils import timezone
 

src/documents/tests/test_api_documents.py

@@ -172,6 +172,35 @@ class TestDocumentApi(DirectoriesMixin, DocumentConsumeDelayMixin, APITestCase):
         results = response.data["results"]
         self.assertEqual(len(results[0]), 0)
 
+    def test_document_fields_api_version_8_respects_created(self):
+        Document.objects.create(
+            title="legacy",
+            checksum="123",
+            mime_type="application/pdf",
+            created=date(2024, 1, 15),
+        )
+
+        response = self.client.get(
+            "/api/documents/?fields=id",
+            headers={"Accept": "application/json; version=8"},
+            format="json",
+        )
+        self.assertEqual(response.status_code, status.HTTP_200_OK)
+        results = response.data["results"]
+        self.assertIn("id", results[0])
+        self.assertNotIn("created", results[0])
+
+        response = self.client.get(
+            "/api/documents/?fields=id,created",
+            headers={"Accept": "application/json; version=8"},
+            format="json",
+        )
+        self.assertEqual(response.status_code, status.HTTP_200_OK)
+        results = response.data["results"]
+        self.assertIn("id", results[0])
+        self.assertIn("created", results[0])
+        self.assertRegex(results[0]["created"], r"^2024-01-15T00:00:00.*$")
+
     def test_document_legacy_created_format(self):
         """
         GIVEN:
@@ -2250,6 +2279,23 @@ class TestDocumentApi(DirectoriesMixin, DocumentConsumeDelayMixin, APITestCase):
         self.assertEqual(response.status_code, status.HTTP_200_OK)
         self.assertListEqual(response.data, ["test", "test2"])
 
+    def test_get_log_with_limit(self):
+        log_data = "test1\ntest2\ntest3\n"
+        with (Path(settings.LOGGING_DIR) / "paperless.log").open("w") as f:
+            f.write(log_data)
+        response = self.client.get("/api/logs/paperless/", {"limit": 2})
+        self.assertEqual(response.status_code, status.HTTP_200_OK)
+        self.assertListEqual(response.data, ["test2", "test3"])
+
+    def test_get_log_with_invalid_limit(self):
+        log_data = "test1\ntest2\n"
+        with (Path(settings.LOGGING_DIR) / "paperless.log").open("w") as f:
+            f.write(log_data)
+        response = self.client.get("/api/logs/paperless/", {"limit": "abc"})
+        self.assertEqual(response.status_code, status.HTTP_400_BAD_REQUEST)
+        response = self.client.get("/api/logs/paperless/", {"limit": -5})
+        self.assertEqual(response.status_code, status.HTTP_400_BAD_REQUEST)
+
     def test_invalid_regex_other_algorithm(self):
         for endpoint in ["correspondents", "tags", "document_types"]:
             response = self.client.post(

src/documents/tests/test_api_remote_version.py

@@ -1,3 +1,4 @@
+from django.core.cache import cache
 from pytest_httpx import HTTPXMock
 from rest_framework import status
 from rest_framework.test import APIClient
@@ -8,6 +9,9 @@ from paperless import version
 class TestApiRemoteVersion:
     ENDPOINT = "/api/remote_version/"
 
+    def setup_method(self):
+        cache.clear()
+
     def test_remote_version_enabled_no_update_prefix(
         self,
         rest_api_client: APIClient,

src/documents/views.py

@@ -7,6 +7,7 @@ import re
 import tempfile
 import zipfile
 from collections import defaultdict
+from collections import deque
 from datetime import datetime
 from pathlib import Path
 from time import mktime
@@ -53,6 +54,7 @@ from django.utils.timezone import make_aware
 from django.utils.translation import get_language
 from django.views import View
 from django.views.decorators.cache import cache_control
+from django.views.decorators.cache import cache_page
 from django.views.decorators.csrf import ensure_csrf_cookie
 from django.views.decorators.http import condition
 from django.views.decorators.http import last_modified
@@ -73,6 +75,7 @@ from rest_framework import parsers
 from rest_framework import serializers
 from rest_framework.decorators import action
 from rest_framework.exceptions import NotFound
+from rest_framework.exceptions import ValidationError
 from rest_framework.filters import OrderingFilter
 from rest_framework.filters import SearchFilter
 from rest_framework.generics import GenericAPIView
@@ -1489,6 +1492,13 @@ class UnifiedSearchViewSet(DocumentViewSet):
                 type=OpenApiTypes.STR,
                 location=OpenApiParameter.PATH,
             ),
+            OpenApiParameter(
+                name="limit",
+                type=OpenApiTypes.INT,
+                location=OpenApiParameter.QUERY,
+                description="Return only the last N entries from the log file",
+                required=False,
+            ),
         ],
         responses={
             (200, "application/json"): serializers.ListSerializer(
@@ -1520,8 +1530,22 @@ class LogViewSet(ViewSet):
         if not log_file.is_file():
             raise Http404
 
+        limit_param = request.query_params.get("limit")
+        if limit_param is not None:
+            try:
+                limit = int(limit_param)
+            except (TypeError, ValueError):
+                raise ValidationError({"limit": "Must be a positive integer"})
+            if limit < 1:
+                raise ValidationError({"limit": "Must be a positive integer"})
+        else:
+            limit = None
+
         with log_file.open() as f:
-            lines = [line.rstrip() for line in f.readlines()]
+            if limit is None:
+                lines = [line.rstrip() for line in f.readlines()]
+            else:
+                lines = [line.rstrip() for line in deque(f, maxlen=limit)]
 
         return Response(lines)
 
@@ -2533,6 +2557,7 @@ class UiSettingsView(GenericAPIView):
         )
 
 
+@method_decorator(cache_page(60 * 15), name="dispatch")
 @extend_schema_view(
     get=extend_schema(
         description="Get the current version of the Paperless-NGX server",

src/paperless_mail/tests/test_parsers_live.py

@@ -53,6 +53,15 @@ class TestUrlCanary:
     Verify certain URLs are still available so testing is valid still
     """
 
+    # Wikimedia rejects requests without a browser-like User-Agent header and returns 403.
+    _WIKIMEDIA_HEADERS = {
+        "User-Agent": (
+            "Mozilla/5.0 (X11; Linux x86_64) "
+            "AppleWebKit/537.36 (KHTML, like Gecko) "
+            "Chrome/123.0.0.0 Safari/537.36"
+        ),
+    }
+
     def test_online_image_exception_on_not_available(self):
         """
         GIVEN:
@@ -70,6 +79,7 @@ class TestUrlCanary:
         with pytest.raises(httpx.HTTPStatusError) as exec_info:
             resp = httpx.get(
                 "https://upload.wikimedia.org/wikipedia/en/f/f7/nonexistent.png",
+                headers=self._WIKIMEDIA_HEADERS,
             )
             resp.raise_for_status()
 
@@ -90,7 +100,10 @@ class TestUrlCanary:
         """
 
         # Now check the URL used in samples/sample.html
-        resp = httpx.get("https://upload.wikimedia.org/wikipedia/en/f/f7/RickRoll.png")
+        resp = httpx.get(
+            "https://upload.wikimedia.org/wikipedia/en/f/f7/RickRoll.png",
+            headers=self._WIKIMEDIA_HEADERS,
+        )
         resp.raise_for_status()