Fix: cache per-request effective-document resolution for thumb/metadata/preview (#13166)

Django's condition() decorator invokes etag_func and last_modified_func
separately, and the view itself may resolve again -- each call to
resolve_effective_document_by_pk() was redoing the same root/version
DB lookups. Memoize the resolution on the request object so a single
thumb/metadata/preview request resolves the effective document once
instead of up to three times.

Related to paperless-ngx/paperless-ngx#13161.

Co-authored-by: Claude Sonnet 5 <noreply@anthropic.com>
This commit is contained in:
Trenton H
2026-07-19 18:04:16 -07:00
committed by GitHub
co-authored by Claude Sonnet 5
parent dcba44ad78
commit 4e52dd5710
3 changed files with 69 additions and 26 deletions
+3 -3
View File
@@ -1,9 +1,9 @@
from datetime import UTC
from datetime import datetime
from typing import Any
from django.conf import settings
from django.core.cache import cache
from rest_framework.request import Request
from documents.caching import CACHE_5_MINUTES
from documents.caching import CACHE_50_MINUTES
@@ -117,7 +117,7 @@ def preview_last_modified(request, pk: int) -> datetime | None:
return doc.modified
def thumbnail_etag(request: Any, pk: int) -> str | None:
def thumbnail_etag(request: Request, pk: int) -> str | None:
"""
Thumbnails are version-dependent, so use the effective document checksum as
the ETag to invalidate cache when the latest version changes.
@@ -128,7 +128,7 @@ def thumbnail_etag(request: Any, pk: int) -> str | None:
return doc.checksum
def thumbnail_last_modified(request: Any, pk: int) -> datetime | None:
def thumbnail_last_modified(request: Request, pk: int) -> datetime | None:
"""
Returns the filesystem last modified either from cache or from filesystem.
Cache should be (slightly?) faster than filesystem
@@ -1,8 +1,12 @@
from datetime import timedelta
from types import SimpleNamespace
from typing import TYPE_CHECKING
from typing import cast
from unittest import mock
from django.db import connection
from django.test import TestCase
from django.test.utils import CaptureQueriesContext
from django.utils import timezone
from documents.conditionals import metadata_etag
@@ -13,6 +17,9 @@ from documents.models import Document
from documents.tests.utils import DirectoriesMixin
from documents.versioning import resolve_effective_document_by_pk
if TYPE_CHECKING:
from rest_framework.request import Request
class TestConditionals(DirectoriesMixin, TestCase):
def test_metadata_etag_uses_latest_version_for_root_request(self) -> None:
@@ -29,14 +36,18 @@ class TestConditionals(DirectoriesMixin, TestCase):
mime_type="application/pdf",
root_document=root,
)
request = SimpleNamespace(query_params={})
request = cast("Request", SimpleNamespace(query_params={}))
self.assertEqual(
metadata_etag(request, root.id),
f"{latest.checksum}:{latest.modified.isoformat()}",
)
self.assertEqual(preview_etag(request, root.id), latest.archive_checksum)
self.assertEqual(thumbnail_etag(request, root.id), latest.checksum)
# preview_etag/thumbnail_etag resolve the same (pk, request) and should
# reuse metadata_etag's cached resolution instead of re-querying.
with CaptureQueriesContext(connection) as ctx:
self.assertEqual(preview_etag(request, root.id), latest.archive_checksum)
self.assertEqual(thumbnail_etag(request, root.id), latest.checksum)
self.assertEqual(len(ctx.captured_queries), 0)
def test_metadata_etag_changes_when_document_modified_changes(self) -> None:
doc = Document.objects.create(
@@ -44,15 +55,20 @@ class TestConditionals(DirectoriesMixin, TestCase):
checksum="same-checksum",
mime_type="application/pdf",
)
request = SimpleNamespace(query_params={})
original_etag = metadata_etag(request, doc.id)
# Each call simulates a separate incoming HTTP request, so it gets its
# own request object -- the per-request resolution cache must not leak
# a stale result across genuinely different requests.
original_etag = metadata_etag(
cast("Request", SimpleNamespace(query_params={})),
doc.id,
)
new_modified = timezone.now() + timedelta(seconds=5)
Document.objects.filter(id=doc.id).update(modified=new_modified)
self.assertNotEqual(metadata_etag(request, doc.id), original_etag)
second_request = cast("Request", SimpleNamespace(query_params={}))
self.assertNotEqual(metadata_etag(second_request, doc.id), original_etag)
self.assertEqual(
metadata_etag(request, doc.id),
metadata_etag(second_request, doc.id),
f"{doc.checksum}:{new_modified.isoformat()}",
)
@@ -76,9 +92,13 @@ class TestConditionals(DirectoriesMixin, TestCase):
root_document=other_root,
)
invalid_request = SimpleNamespace(query_params={"version": "not-a-number"})
unrelated_request = SimpleNamespace(
query_params={"version": str(other_version.id)},
invalid_request = cast(
"Request",
SimpleNamespace(query_params={"version": "not-a-number"}),
)
unrelated_request = cast(
"Request",
SimpleNamespace(query_params={"version": str(other_version.id)}),
)
self.assertIsNone(
@@ -105,7 +125,7 @@ class TestConditionals(DirectoriesMixin, TestCase):
latest.thumbnail_path.parent.mkdir(parents=True, exist_ok=True)
latest.thumbnail_path.write_bytes(b"thumb")
request = SimpleNamespace(query_params={})
request = cast("Request", SimpleNamespace(query_params={}))
with mock.patch(
"documents.conditionals.get_thumbnail_modified_key",
return_value="thumb-modified-key",
+34 -11
View File
@@ -8,7 +8,7 @@ from typing import Any
from documents.models import Document
if TYPE_CHECKING:
from django.http import HttpRequest
from rest_framework.request import Request
class VersionResolutionError(StrEnum):
@@ -26,7 +26,7 @@ def _document_manager(*, include_deleted: bool) -> Any:
return Document.global_objects if include_deleted else Document.objects
def get_request_version_param(request: HttpRequest) -> str | None:
def get_request_version_param(request: Request) -> str | None:
if hasattr(request, "query_params"):
return request.query_params.get("version")
return None
@@ -57,7 +57,7 @@ def get_latest_version_for_root(
def resolve_requested_version_for_root(
root_doc: Document,
request: Any,
request: Request,
*,
include_deleted: bool = False,
) -> VersionResolution:
@@ -86,7 +86,7 @@ def resolve_requested_version_for_root(
def resolve_effective_document(
request_doc: Document,
request: Any,
request: Request,
*,
include_deleted: bool = False,
) -> VersionResolution:
@@ -107,18 +107,41 @@ def resolve_effective_document(
return VersionResolution(document=request_doc)
_EFFECTIVE_DOCUMENT_CACHE_ATTR = "_effective_document_resolution_cache"
def resolve_effective_document_by_pk(
pk: int,
request: Any,
request: Request,
*,
include_deleted: bool = False,
) -> VersionResolution:
# Django's `condition()` decorator (used for ETag/Last-Modified) invokes the
# etag_func and last_modified_func separately, and the view itself may resolve
# again -- all against the same request. Cache per-request so a single thumb/
# metadata/preview request doesn't redo this resolution multiple times.
cache = getattr(request, _EFFECTIVE_DOCUMENT_CACHE_ATTR, None)
if cache is None:
cache = {}
setattr(request, _EFFECTIVE_DOCUMENT_CACHE_ATTR, cache)
key = (pk, include_deleted)
if key in cache:
return cache[key]
manager = _document_manager(include_deleted=include_deleted)
request_doc = manager.only("id", "root_document_id").filter(pk=pk).first()
if request_doc is None:
return VersionResolution(document=None, error=VersionResolutionError.NOT_FOUND)
return resolve_effective_document(
request_doc,
request,
include_deleted=include_deleted,
)
resolution = VersionResolution(
document=None,
error=VersionResolutionError.NOT_FOUND,
)
else:
resolution = resolve_effective_document(
request_doc,
request,
include_deleted=include_deleted,
)
cache[key] = resolution
return resolution