mirror of
https://github.com/paperless-ngx/paperless-ngx.git
synced 2025-04-02 13:45:10 -05:00
Feature: Cache metadata and suggestions in Redis (#5638)
This commit is contained in:
parent
45e2b7f814
commit
25542c56b9
197
src/documents/caching.py
Normal file
197
src/documents/caching.py
Normal file
@ -0,0 +1,197 @@
|
||||
import logging
|
||||
from binascii import hexlify
|
||||
from dataclasses import dataclass
|
||||
from typing import TYPE_CHECKING
|
||||
from typing import Final
|
||||
from typing import Optional
|
||||
|
||||
from django.core.cache import cache
|
||||
|
||||
from documents.models import Document
|
||||
|
||||
if TYPE_CHECKING:
|
||||
from documents.classifier import DocumentClassifier
|
||||
|
||||
logger = logging.getLogger("paperless.caching")
|
||||
|
||||
|
||||
@dataclass(frozen=True)
|
||||
class MetadataCacheData:
|
||||
original_checksum: str
|
||||
original_metadata: list
|
||||
archive_checksum: Optional[str]
|
||||
archive_metadata: Optional[list]
|
||||
|
||||
|
||||
@dataclass(frozen=True)
|
||||
class SuggestionCacheData:
|
||||
classifier_version: int
|
||||
classifier_hash: str
|
||||
suggestions: dict
|
||||
|
||||
|
||||
CLASSIFIER_VERSION_KEY: Final[str] = "classifier_version"
|
||||
CLASSIFIER_HASH_KEY: Final[str] = "classifier_hash"
|
||||
CLASSIFIER_MODIFIED_KEY: Final[str] = "classifier_modified"
|
||||
|
||||
CACHE_1_MINUTE: Final[int] = 60
|
||||
CACHE_5_MINUTES: Final[int] = 5 * CACHE_1_MINUTE
|
||||
CACHE_50_MINUTES: Final[int] = 50 * CACHE_1_MINUTE
|
||||
|
||||
|
||||
def get_suggestion_cache_key(document_id: int) -> str:
|
||||
"""
|
||||
Returns the basic key for a document's suggestions
|
||||
"""
|
||||
return f"doc_{document_id}_suggest"
|
||||
|
||||
|
||||
def get_suggestion_cache(document_id: int) -> Optional[SuggestionCacheData]:
|
||||
"""
|
||||
If possible, return the cached suggestions for the given document ID.
|
||||
The classifier needs to be matching in format and hash and the suggestions need to
|
||||
have been cached once.
|
||||
"""
|
||||
from documents.classifier import DocumentClassifier
|
||||
|
||||
doc_key = get_suggestion_cache_key(document_id)
|
||||
cache_hits = cache.get_many([CLASSIFIER_VERSION_KEY, CLASSIFIER_HASH_KEY, doc_key])
|
||||
# The document suggestions are in the cache
|
||||
if doc_key in cache_hits:
|
||||
doc_suggestions: SuggestionCacheData = cache_hits[doc_key]
|
||||
# The classifier format is the same
|
||||
# The classifier hash is the same
|
||||
# Then the suggestions can be used
|
||||
if (
|
||||
CLASSIFIER_VERSION_KEY in cache_hits
|
||||
and cache_hits[CLASSIFIER_VERSION_KEY] == DocumentClassifier.FORMAT_VERSION
|
||||
and cache_hits[CLASSIFIER_VERSION_KEY] == doc_suggestions.classifier_version
|
||||
) and (
|
||||
CLASSIFIER_HASH_KEY in cache_hits
|
||||
and cache_hits[CLASSIFIER_HASH_KEY] == doc_suggestions.classifier_hash
|
||||
):
|
||||
return doc_suggestions
|
||||
else: # pragma: no cover
|
||||
# Remove the key because something didn't match
|
||||
cache.delete(doc_key)
|
||||
return None
|
||||
|
||||
|
||||
def set_suggestions_cache(
|
||||
document_id: int,
|
||||
suggestions: dict,
|
||||
classifier: Optional["DocumentClassifier"],
|
||||
*,
|
||||
timeout=CACHE_50_MINUTES,
|
||||
) -> None:
|
||||
"""
|
||||
Caches the given suggestions, which were generated by the given classifier. If there is no classifier,
|
||||
this function is a no-op (there won't be suggestions then anyway)
|
||||
"""
|
||||
if classifier is not None:
|
||||
doc_key = get_suggestion_cache_key(document_id)
|
||||
print(classifier.last_auto_type_hash)
|
||||
cache.set(
|
||||
doc_key,
|
||||
SuggestionCacheData(
|
||||
classifier.FORMAT_VERSION,
|
||||
hexlify(classifier.last_auto_type_hash).decode(),
|
||||
suggestions,
|
||||
),
|
||||
timeout,
|
||||
)
|
||||
|
||||
|
||||
def refresh_suggestions_cache(
|
||||
document_id: int,
|
||||
*,
|
||||
timeout: int = CACHE_50_MINUTES,
|
||||
) -> None:
|
||||
"""
|
||||
Refreshes the expiration of the suggestions for the given document ID
|
||||
to the given timeout
|
||||
"""
|
||||
doc_key = get_suggestion_cache_key(document_id)
|
||||
cache.touch(doc_key, timeout)
|
||||
|
||||
|
||||
def get_metadata_cache_key(document_id: int) -> str:
|
||||
"""
|
||||
Returns the basic key for a document's metadata
|
||||
"""
|
||||
return f"doc_{document_id}_metadata"
|
||||
|
||||
|
||||
def get_metadata_cache(document_id: int) -> Optional[MetadataCacheData]:
|
||||
"""
|
||||
Returns the cached document metadata for the given document ID, as long as the metadata
|
||||
was cached once and the checksums have not changed
|
||||
"""
|
||||
doc_key = get_metadata_cache_key(document_id)
|
||||
doc_metadata: Optional[MetadataCacheData] = cache.get(doc_key)
|
||||
# The metadata exists in the cache
|
||||
if doc_metadata is not None:
|
||||
try:
|
||||
doc = Document.objects.get(pk=document_id)
|
||||
# The original checksums match
|
||||
# If it has one, the archive checksums match
|
||||
# Then, we can use the metadata
|
||||
if (
|
||||
doc_metadata.original_checksum == doc.checksum
|
||||
and doc.has_archive_version
|
||||
and doc_metadata.archive_checksum is not None
|
||||
and doc_metadata.archive_checksum == doc.archive_checksum
|
||||
):
|
||||
# Refresh cache
|
||||
cache.touch(doc_key, CACHE_50_MINUTES)
|
||||
return doc_metadata
|
||||
else: # pragma: no cover
|
||||
# Something didn't match, delete the key
|
||||
cache.delete(doc_key)
|
||||
except Document.DoesNotExist: # pragma: no cover
|
||||
# Basically impossible, but the key existed, but the Document didn't
|
||||
cache.delete(doc_key)
|
||||
return None
|
||||
|
||||
|
||||
def set_metadata_cache(
|
||||
document: Document,
|
||||
original_metadata: list,
|
||||
archive_metadata: Optional[list],
|
||||
*,
|
||||
timeout=CACHE_50_MINUTES,
|
||||
) -> None:
|
||||
"""
|
||||
Sets the metadata into cache for the given Document
|
||||
"""
|
||||
doc_key = get_metadata_cache_key(document.pk)
|
||||
cache.set(
|
||||
doc_key,
|
||||
MetadataCacheData(
|
||||
document.checksum,
|
||||
original_metadata,
|
||||
document.archive_checksum,
|
||||
archive_metadata,
|
||||
),
|
||||
timeout,
|
||||
)
|
||||
|
||||
|
||||
def refresh_metadata_cache(
|
||||
document_id: int,
|
||||
*,
|
||||
timeout: int = CACHE_50_MINUTES,
|
||||
) -> None:
|
||||
"""
|
||||
Refreshes the expiration of the metadata for the given document ID
|
||||
to the given timeout
|
||||
"""
|
||||
doc_key = get_metadata_cache_key(document_id)
|
||||
cache.touch(doc_key, timeout)
|
||||
|
||||
|
||||
def get_thumbnail_modified_key(document_id: int) -> str:
|
||||
"""
|
||||
Builds the key to store a thumbnail's timestamp
|
||||
"""
|
||||
return f"doc_{document_id}_thumbnail_modified"
|
@ -10,8 +10,13 @@ from pathlib import Path
|
||||
from typing import Optional
|
||||
|
||||
from django.conf import settings
|
||||
from django.core.cache import cache
|
||||
from sklearn.exceptions import InconsistentVersionWarning
|
||||
|
||||
from documents.caching import CACHE_50_MINUTES
|
||||
from documents.caching import CLASSIFIER_HASH_KEY
|
||||
from documents.caching import CLASSIFIER_MODIFIED_KEY
|
||||
from documents.caching import CLASSIFIER_VERSION_KEY
|
||||
from documents.models import Document
|
||||
from documents.models import MatchingModel
|
||||
|
||||
@ -208,6 +213,15 @@ class DocumentClassifier:
|
||||
and self.last_doc_change_time >= latest_doc_change
|
||||
) and self.last_auto_type_hash == hasher.digest():
|
||||
logger.info("No updates since last training")
|
||||
# Set the classifier information into the cache
|
||||
# Caching for 50 minutes, so slightly less than the normal retrain time
|
||||
cache.set(
|
||||
CLASSIFIER_MODIFIED_KEY,
|
||||
self.last_doc_change_time,
|
||||
CACHE_50_MINUTES,
|
||||
)
|
||||
cache.set(CLASSIFIER_HASH_KEY, hasher.hexdigest(), CACHE_50_MINUTES)
|
||||
cache.set(CLASSIFIER_VERSION_KEY, self.FORMAT_VERSION, CACHE_50_MINUTES)
|
||||
return False
|
||||
|
||||
# subtract 1 since -1 (null) is also part of the classes.
|
||||
@ -322,6 +336,12 @@ class DocumentClassifier:
|
||||
self.last_doc_change_time = latest_doc_change
|
||||
self.last_auto_type_hash = hasher.digest()
|
||||
|
||||
# Set the classifier information into the cache
|
||||
# Caching for 50 minutes, so slightly less than the normal retrain time
|
||||
cache.set(CLASSIFIER_MODIFIED_KEY, self.last_doc_change_time, CACHE_50_MINUTES)
|
||||
cache.set(CLASSIFIER_HASH_KEY, hasher.hexdigest(), CACHE_50_MINUTES)
|
||||
cache.set(CLASSIFIER_VERSION_KEY, self.FORMAT_VERSION, CACHE_50_MINUTES)
|
||||
|
||||
return True
|
||||
|
||||
def preprocess_content(self, content: str) -> str: # pragma: no cover
|
||||
|
@ -1,9 +1,16 @@
|
||||
import pickle
|
||||
from datetime import datetime
|
||||
from datetime import timezone
|
||||
from typing import Optional
|
||||
|
||||
from django.conf import settings
|
||||
from django.core.cache import cache
|
||||
|
||||
from documents.caching import CACHE_5_MINUTES
|
||||
from documents.caching import CACHE_50_MINUTES
|
||||
from documents.caching import CLASSIFIER_HASH_KEY
|
||||
from documents.caching import CLASSIFIER_MODIFIED_KEY
|
||||
from documents.caching import CLASSIFIER_VERSION_KEY
|
||||
from documents.caching import get_thumbnail_modified_key
|
||||
from documents.classifier import DocumentClassifier
|
||||
from documents.models import Document
|
||||
|
||||
@ -14,18 +21,25 @@ def suggestions_etag(request, pk: int) -> Optional[str]:
|
||||
suggestions if the classifier has not been changed and the suggested dates
|
||||
setting is also unchanged
|
||||
|
||||
TODO: It would be nice to not duplicate the partial loading and the loading
|
||||
between here and the actual classifier
|
||||
"""
|
||||
# If no model file, no etag at all
|
||||
if not settings.MODEL_FILE.exists():
|
||||
return None
|
||||
with open(settings.MODEL_FILE, "rb") as f:
|
||||
schema_version = pickle.load(f)
|
||||
if schema_version != DocumentClassifier.FORMAT_VERSION:
|
||||
return None
|
||||
_ = pickle.load(f)
|
||||
last_auto_type_hash: bytes = pickle.load(f)
|
||||
return f"{last_auto_type_hash}:{settings.NUMBER_OF_SUGGESTED_DATES}"
|
||||
# Check cache information
|
||||
cache_hits = cache.get_many(
|
||||
[CLASSIFIER_VERSION_KEY, CLASSIFIER_HASH_KEY],
|
||||
)
|
||||
# If the version differs somehow, no etag
|
||||
if (
|
||||
CLASSIFIER_VERSION_KEY in cache_hits
|
||||
and cache_hits[CLASSIFIER_VERSION_KEY] != DocumentClassifier.FORMAT_VERSION
|
||||
):
|
||||
return None
|
||||
elif CLASSIFIER_HASH_KEY in cache_hits:
|
||||
# Refresh the cache and return the hash digest and the dates setting
|
||||
cache.touch(CLASSIFIER_HASH_KEY, CACHE_5_MINUTES)
|
||||
return f"{cache_hits[CLASSIFIER_HASH_KEY]}:{settings.NUMBER_OF_SUGGESTED_DATES}"
|
||||
return None
|
||||
|
||||
|
||||
def suggestions_last_modified(request, pk: int) -> Optional[datetime]:
|
||||
@ -34,14 +48,23 @@ def suggestions_last_modified(request, pk: int) -> Optional[datetime]:
|
||||
as there is not way to track the suggested date setting modification, but it seems
|
||||
unlikely that changes too often
|
||||
"""
|
||||
# No file, no last modified
|
||||
if not settings.MODEL_FILE.exists():
|
||||
return None
|
||||
with open(settings.MODEL_FILE, "rb") as f:
|
||||
schema_version = pickle.load(f)
|
||||
if schema_version != DocumentClassifier.FORMAT_VERSION:
|
||||
return None
|
||||
last_doc_change_time = pickle.load(f)
|
||||
return last_doc_change_time
|
||||
cache_hits = cache.get_many(
|
||||
[CLASSIFIER_VERSION_KEY, CLASSIFIER_MODIFIED_KEY],
|
||||
)
|
||||
# If the version differs somehow, no last modified
|
||||
if (
|
||||
CLASSIFIER_VERSION_KEY in cache_hits
|
||||
and cache_hits[CLASSIFIER_VERSION_KEY] != DocumentClassifier.FORMAT_VERSION
|
||||
):
|
||||
return None
|
||||
elif CLASSIFIER_MODIFIED_KEY in cache_hits:
|
||||
# Refresh the cache and return the last modified
|
||||
cache.touch(CLASSIFIER_MODIFIED_KEY, CACHE_5_MINUTES)
|
||||
return cache_hits[CLASSIFIER_MODIFIED_KEY]
|
||||
return None
|
||||
|
||||
|
||||
def metadata_etag(request, pk: int) -> Optional[str]:
|
||||
@ -52,7 +75,7 @@ def metadata_etag(request, pk: int) -> Optional[str]:
|
||||
try:
|
||||
doc = Document.objects.get(pk=pk)
|
||||
return doc.checksum
|
||||
except Document.DoesNotExist:
|
||||
except Document.DoesNotExist: # pragma: no cover
|
||||
return None
|
||||
return None
|
||||
|
||||
@ -66,7 +89,7 @@ def metadata_last_modified(request, pk: int) -> Optional[datetime]:
|
||||
try:
|
||||
doc = Document.objects.get(pk=pk)
|
||||
return doc.modified
|
||||
except Document.DoesNotExist:
|
||||
except Document.DoesNotExist: # pragma: no cover
|
||||
return None
|
||||
return None
|
||||
|
||||
@ -82,6 +105,46 @@ def preview_etag(request, pk: int) -> Optional[str]:
|
||||
and request.query_params["original"] == "true"
|
||||
)
|
||||
return doc.checksum if use_original else doc.archive_checksum
|
||||
except Document.DoesNotExist:
|
||||
except Document.DoesNotExist: # pragma: no cover
|
||||
return None
|
||||
return None
|
||||
|
||||
|
||||
def preview_last_modified(request, pk: int) -> Optional[datetime]:
|
||||
"""
|
||||
Uses the documents modified time to set the Last-Modified header. Not strictly
|
||||
speaking correct, but close enough and quick
|
||||
"""
|
||||
try:
|
||||
doc = Document.objects.get(pk=pk)
|
||||
return doc.modified
|
||||
except Document.DoesNotExist: # pragma: no cover
|
||||
return None
|
||||
return None
|
||||
|
||||
|
||||
def thumbnail_last_modified(request, pk: int) -> Optional[datetime]:
|
||||
"""
|
||||
Returns the filesystem last modified either from cache or from filesystem.
|
||||
Cache should be (slightly?) faster than filesystem
|
||||
"""
|
||||
try:
|
||||
doc = Document.objects.get(pk=pk)
|
||||
if not doc.thumbnail_path.exists():
|
||||
return None
|
||||
doc_key = get_thumbnail_modified_key(pk)
|
||||
|
||||
cache_hit = cache.get(doc_key)
|
||||
if cache_hit is not None:
|
||||
cache.touch(doc_key, CACHE_50_MINUTES)
|
||||
return cache_hit
|
||||
|
||||
# No cache, get the timestamp and cache the datetime
|
||||
last_modified = datetime.fromtimestamp(
|
||||
doc.thumbnail_path.stat().st_mtime,
|
||||
tz=timezone.utc,
|
||||
)
|
||||
cache.set(doc_key, last_modified, CACHE_50_MINUTES)
|
||||
return last_modified
|
||||
except Document.DoesNotExist: # pragma: no cover
|
||||
return None
|
||||
|
@ -4,6 +4,7 @@ import shutil
|
||||
import tempfile
|
||||
import uuid
|
||||
import zoneinfo
|
||||
from binascii import hexlify
|
||||
from datetime import timedelta
|
||||
from pathlib import Path
|
||||
from unittest import mock
|
||||
@ -13,12 +14,17 @@ from dateutil import parser
|
||||
from django.conf import settings
|
||||
from django.contrib.auth.models import Permission
|
||||
from django.contrib.auth.models import User
|
||||
from django.core.cache import cache
|
||||
from django.test import override_settings
|
||||
from django.utils import timezone
|
||||
from guardian.shortcuts import assign_perm
|
||||
from rest_framework import status
|
||||
from rest_framework.test import APITestCase
|
||||
|
||||
from documents.caching import CACHE_50_MINUTES
|
||||
from documents.caching import CLASSIFIER_HASH_KEY
|
||||
from documents.caching import CLASSIFIER_MODIFIED_KEY
|
||||
from documents.caching import CLASSIFIER_VERSION_KEY
|
||||
from documents.models import Correspondent
|
||||
from documents.models import CustomField
|
||||
from documents.models import CustomFieldInstance
|
||||
@ -40,6 +46,7 @@ class TestDocumentApi(DirectoriesMixin, DocumentConsumeDelayMixin, APITestCase):
|
||||
|
||||
self.user = User.objects.create_superuser(username="temp_admin")
|
||||
self.client.force_authenticate(user=self.user)
|
||||
cache.clear()
|
||||
|
||||
def testDocuments(self):
|
||||
response = self.client.get("/api/documents/").data
|
||||
@ -1162,6 +1169,9 @@ class TestDocumentApi(DirectoriesMixin, DocumentConsumeDelayMixin, APITestCase):
|
||||
self.assertEqual(meta["original_size"], os.stat(source_file).st_size)
|
||||
self.assertEqual(meta["archive_size"], os.stat(archive_file).st_size)
|
||||
|
||||
response = self.client.get(f"/api/documents/{doc.pk}/metadata/")
|
||||
self.assertEqual(response.status_code, status.HTTP_200_OK)
|
||||
|
||||
def test_get_metadata_invalid_doc(self):
|
||||
response = self.client.get("/api/documents/34576/metadata/")
|
||||
self.assertEqual(response.status_code, status.HTTP_404_NOT_FOUND)
|
||||
@ -1266,7 +1276,7 @@ class TestDocumentApi(DirectoriesMixin, DocumentConsumeDelayMixin, APITestCase):
|
||||
},
|
||||
)
|
||||
|
||||
@mock.patch("documents.conditionals.pickle.load")
|
||||
@mock.patch("documents.views.load_classifier")
|
||||
@mock.patch("documents.views.match_storage_paths")
|
||||
@mock.patch("documents.views.match_document_types")
|
||||
@mock.patch("documents.views.match_tags")
|
||||
@ -1278,7 +1288,7 @@ class TestDocumentApi(DirectoriesMixin, DocumentConsumeDelayMixin, APITestCase):
|
||||
match_tags,
|
||||
match_document_types,
|
||||
match_storage_paths,
|
||||
mocked_pickle_load,
|
||||
mocked_load,
|
||||
):
|
||||
"""
|
||||
GIVEN:
|
||||
@ -1287,23 +1297,43 @@ class TestDocumentApi(DirectoriesMixin, DocumentConsumeDelayMixin, APITestCase):
|
||||
- Classifier has not been modified
|
||||
THEN:
|
||||
- Subsequent requests are returned alright
|
||||
- ETag and last modified are called
|
||||
- ETag and last modified headers are set
|
||||
"""
|
||||
settings.MODEL_FILE.touch()
|
||||
|
||||
# setup the cache how the classifier does it
|
||||
from documents.classifier import DocumentClassifier
|
||||
|
||||
last_modified = timezone.now()
|
||||
settings.MODEL_FILE.touch()
|
||||
|
||||
# ETag first, then modified
|
||||
mock_effect = [
|
||||
DocumentClassifier.FORMAT_VERSION,
|
||||
"dont care",
|
||||
b"thisisachecksum",
|
||||
DocumentClassifier.FORMAT_VERSION,
|
||||
last_modified,
|
||||
classifier_checksum_bytes = b"thisisachecksum"
|
||||
classifier_checksum_hex = hexlify(classifier_checksum_bytes).decode()
|
||||
|
||||
# Two loads, so two side effects
|
||||
mocked_load.side_effect = [
|
||||
mock.Mock(
|
||||
last_auto_type_hash=classifier_checksum_bytes,
|
||||
FORMAT_VERSION=DocumentClassifier.FORMAT_VERSION,
|
||||
),
|
||||
mock.Mock(
|
||||
last_auto_type_hash=classifier_checksum_bytes,
|
||||
FORMAT_VERSION=DocumentClassifier.FORMAT_VERSION,
|
||||
),
|
||||
]
|
||||
mocked_pickle_load.side_effect = mock_effect
|
||||
|
||||
last_modified = timezone.now()
|
||||
cache.set(CLASSIFIER_MODIFIED_KEY, last_modified, CACHE_50_MINUTES)
|
||||
cache.set(CLASSIFIER_HASH_KEY, classifier_checksum_hex, CACHE_50_MINUTES)
|
||||
cache.set(
|
||||
CLASSIFIER_VERSION_KEY,
|
||||
DocumentClassifier.FORMAT_VERSION,
|
||||
CACHE_50_MINUTES,
|
||||
)
|
||||
|
||||
# Mock the matching
|
||||
match_correspondents.return_value = [Correspondent(id=88), Correspondent(id=2)]
|
||||
match_tags.return_value = [Tag(id=56), Tag(id=123)]
|
||||
match_document_types.return_value = [DocumentType(id=23)]
|
||||
match_storage_paths.return_value = [StoragePath(id=99), StoragePath(id=77)]
|
||||
|
||||
doc = Document.objects.create(
|
||||
title="test",
|
||||
@ -1311,12 +1341,8 @@ class TestDocumentApi(DirectoriesMixin, DocumentConsumeDelayMixin, APITestCase):
|
||||
content="this is an invoice from 12.04.2022!",
|
||||
)
|
||||
|
||||
match_correspondents.return_value = [Correspondent(id=88), Correspondent(id=2)]
|
||||
match_tags.return_value = [Tag(id=56), Tag(id=123)]
|
||||
match_document_types.return_value = [DocumentType(id=23)]
|
||||
match_storage_paths.return_value = [StoragePath(id=99), StoragePath(id=77)]
|
||||
|
||||
response = self.client.get(f"/api/documents/{doc.pk}/suggestions/")
|
||||
self.assertEqual(response.status_code, status.HTTP_200_OK)
|
||||
self.assertEqual(
|
||||
response.data,
|
||||
{
|
||||
@ -1327,7 +1353,6 @@ class TestDocumentApi(DirectoriesMixin, DocumentConsumeDelayMixin, APITestCase):
|
||||
"dates": ["2022-04-12"],
|
||||
},
|
||||
)
|
||||
mocked_pickle_load.assert_called()
|
||||
self.assertIn("Last-Modified", response.headers)
|
||||
self.assertEqual(
|
||||
response.headers["Last-Modified"],
|
||||
@ -1336,15 +1361,11 @@ class TestDocumentApi(DirectoriesMixin, DocumentConsumeDelayMixin, APITestCase):
|
||||
self.assertIn("ETag", response.headers)
|
||||
self.assertEqual(
|
||||
response.headers["ETag"],
|
||||
f"\"b'thisisachecksum':{settings.NUMBER_OF_SUGGESTED_DATES}\"",
|
||||
f'"{classifier_checksum_hex}:{settings.NUMBER_OF_SUGGESTED_DATES}"',
|
||||
)
|
||||
|
||||
mocked_pickle_load.rest_mock()
|
||||
mocked_pickle_load.side_effect = mock_effect
|
||||
|
||||
response = self.client.get(f"/api/documents/{doc.pk}/suggestions/")
|
||||
self.assertEqual(response.status_code, status.HTTP_200_OK)
|
||||
mocked_pickle_load.assert_called()
|
||||
|
||||
@mock.patch("documents.parsers.parse_date_generator")
|
||||
@override_settings(NUMBER_OF_SUGGESTED_DATES=0)
|
||||
|
@ -35,6 +35,7 @@ from django.utils.translation import get_language
|
||||
from django.views import View
|
||||
from django.views.decorators.cache import cache_control
|
||||
from django.views.decorators.http import condition
|
||||
from django.views.decorators.http import last_modified
|
||||
from django.views.generic import TemplateView
|
||||
from django_filters.rest_framework import DjangoFilterBackend
|
||||
from langdetect import detect
|
||||
@ -62,12 +63,21 @@ from documents import bulk_edit
|
||||
from documents.bulk_download import ArchiveOnlyStrategy
|
||||
from documents.bulk_download import OriginalAndArchiveStrategy
|
||||
from documents.bulk_download import OriginalsOnlyStrategy
|
||||
from documents.caching import CACHE_50_MINUTES
|
||||
from documents.caching import get_metadata_cache
|
||||
from documents.caching import get_suggestion_cache
|
||||
from documents.caching import refresh_metadata_cache
|
||||
from documents.caching import refresh_suggestions_cache
|
||||
from documents.caching import set_metadata_cache
|
||||
from documents.caching import set_suggestions_cache
|
||||
from documents.classifier import load_classifier
|
||||
from documents.conditionals import metadata_etag
|
||||
from documents.conditionals import metadata_last_modified
|
||||
from documents.conditionals import preview_etag
|
||||
from documents.conditionals import preview_last_modified
|
||||
from documents.conditionals import suggestions_etag
|
||||
from documents.conditionals import suggestions_last_modified
|
||||
from documents.conditionals import thumbnail_last_modified
|
||||
from documents.data_models import ConsumableDocument
|
||||
from documents.data_models import DocumentMetadataOverrides
|
||||
from documents.data_models import DocumentSource
|
||||
@ -379,10 +389,12 @@ class DocumentViewSet(
|
||||
|
||||
try:
|
||||
return parser.extract_metadata(file, mime_type)
|
||||
except Exception:
|
||||
except Exception: # pragma: no cover
|
||||
logger.exception(f"Issue getting metadata for {file}")
|
||||
# TODO: cover GPG errors, remove later.
|
||||
return []
|
||||
else:
|
||||
else: # pragma: no cover
|
||||
logger.warning(f"No parser for {mime_type}")
|
||||
return []
|
||||
|
||||
def get_filesize(self, filename):
|
||||
@ -407,16 +419,37 @@ class DocumentViewSet(
|
||||
except Document.DoesNotExist:
|
||||
raise Http404
|
||||
|
||||
document_cached_metadata = get_metadata_cache(doc.pk)
|
||||
|
||||
archive_metadata = None
|
||||
archive_filesize = None
|
||||
if document_cached_metadata is not None:
|
||||
original_metadata = document_cached_metadata.original_metadata
|
||||
archive_metadata = document_cached_metadata.archive_metadata
|
||||
refresh_metadata_cache(doc.pk)
|
||||
else:
|
||||
original_metadata = self.get_metadata(doc.source_path, doc.mime_type)
|
||||
|
||||
if doc.has_archive_version:
|
||||
archive_filesize = self.get_filesize(doc.archive_path)
|
||||
archive_metadata = self.get_metadata(
|
||||
doc.archive_path,
|
||||
"application/pdf",
|
||||
)
|
||||
set_metadata_cache(doc, original_metadata, archive_metadata)
|
||||
|
||||
meta = {
|
||||
"original_checksum": doc.checksum,
|
||||
"original_size": self.get_filesize(doc.source_path),
|
||||
"original_mime_type": doc.mime_type,
|
||||
"media_filename": doc.filename,
|
||||
"has_archive_version": doc.has_archive_version,
|
||||
"original_metadata": self.get_metadata(doc.source_path, doc.mime_type),
|
||||
"original_metadata": original_metadata,
|
||||
"archive_checksum": doc.archive_checksum,
|
||||
"archive_media_filename": doc.archive_filename,
|
||||
"original_filename": doc.original_filename,
|
||||
"archive_size": archive_filesize,
|
||||
"archive_metadata": archive_metadata,
|
||||
}
|
||||
|
||||
lang = "en"
|
||||
@ -426,16 +459,6 @@ class DocumentViewSet(
|
||||
pass
|
||||
meta["lang"] = lang
|
||||
|
||||
if doc.has_archive_version:
|
||||
meta["archive_size"] = self.get_filesize(doc.archive_path)
|
||||
meta["archive_metadata"] = self.get_metadata(
|
||||
doc.archive_path,
|
||||
"application/pdf",
|
||||
)
|
||||
else:
|
||||
meta["archive_size"] = None
|
||||
meta["archive_metadata"] = None
|
||||
|
||||
return Response(meta)
|
||||
|
||||
@action(methods=["get"], detail=True)
|
||||
@ -454,6 +477,12 @@ class DocumentViewSet(
|
||||
):
|
||||
return HttpResponseForbidden("Insufficient permissions")
|
||||
|
||||
document_suggestions = get_suggestion_cache(doc.pk)
|
||||
|
||||
if document_suggestions is not None:
|
||||
refresh_suggestions_cache(doc.pk)
|
||||
return Response(document_suggestions.suggestions)
|
||||
|
||||
classifier = load_classifier()
|
||||
|
||||
dates = []
|
||||
@ -463,27 +492,30 @@ class DocumentViewSet(
|
||||
{i for i in itertools.islice(gen, settings.NUMBER_OF_SUGGESTED_DATES)},
|
||||
)
|
||||
|
||||
return Response(
|
||||
{
|
||||
"correspondents": [
|
||||
c.id for c in match_correspondents(doc, classifier, request.user)
|
||||
],
|
||||
"tags": [t.id for t in match_tags(doc, classifier, request.user)],
|
||||
"document_types": [
|
||||
dt.id for dt in match_document_types(doc, classifier, request.user)
|
||||
],
|
||||
"storage_paths": [
|
||||
dt.id for dt in match_storage_paths(doc, classifier, request.user)
|
||||
],
|
||||
"dates": [
|
||||
date.strftime("%Y-%m-%d") for date in dates if date is not None
|
||||
],
|
||||
},
|
||||
)
|
||||
resp_data = {
|
||||
"correspondents": [
|
||||
c.id for c in match_correspondents(doc, classifier, request.user)
|
||||
],
|
||||
"tags": [t.id for t in match_tags(doc, classifier, request.user)],
|
||||
"document_types": [
|
||||
dt.id for dt in match_document_types(doc, classifier, request.user)
|
||||
],
|
||||
"storage_paths": [
|
||||
dt.id for dt in match_storage_paths(doc, classifier, request.user)
|
||||
],
|
||||
"dates": [date.strftime("%Y-%m-%d") for date in dates if date is not None],
|
||||
}
|
||||
|
||||
# Cache the suggestions and the classifier hash for later
|
||||
set_suggestions_cache(doc.pk, resp_data, classifier)
|
||||
|
||||
return Response(resp_data)
|
||||
|
||||
@action(methods=["get"], detail=True)
|
||||
@method_decorator(cache_control(public=False, max_age=5 * 60))
|
||||
@method_decorator(condition(etag_func=preview_etag))
|
||||
@method_decorator(
|
||||
condition(etag_func=preview_etag, last_modified_func=preview_last_modified),
|
||||
)
|
||||
def preview(self, request, pk=None):
|
||||
try:
|
||||
response = self.file_response(pk, request, "inline")
|
||||
@ -492,7 +524,8 @@ class DocumentViewSet(
|
||||
raise Http404
|
||||
|
||||
@action(methods=["get"], detail=True)
|
||||
@method_decorator(cache_control(public=False, max_age=315360000))
|
||||
@method_decorator(cache_control(public=False, max_age=CACHE_50_MINUTES))
|
||||
@method_decorator(last_modified(thumbnail_last_modified))
|
||||
def thumb(self, request, pk=None):
|
||||
try:
|
||||
doc = Document.objects.get(id=pk)
|
||||
@ -506,8 +539,6 @@ class DocumentViewSet(
|
||||
handle = GnuPG.decrypted(doc.thumbnail_file)
|
||||
else:
|
||||
handle = doc.thumbnail_file
|
||||
# TODO: Send ETag information and use that to send new thumbnails
|
||||
# if available
|
||||
|
||||
return HttpResponse(handle, content_type="image/webp")
|
||||
except (FileNotFoundError, Document.DoesNotExist):
|
||||
|
@ -762,8 +762,12 @@ CELERY_BEAT_SCHEDULE_FILENAME = os.path.join(DATA_DIR, "celerybeat-schedule.db")
|
||||
# django setting.
|
||||
CACHES = {
|
||||
"default": {
|
||||
"BACKEND": "django.core.cache.backends.redis.RedisCache",
|
||||
"BACKEND": os.environ.get(
|
||||
"PAPERLESS_CACHE_BACKEND",
|
||||
"django.core.cache.backends.redis.RedisCache",
|
||||
),
|
||||
"LOCATION": _CHANNELS_REDIS_URL,
|
||||
"KEY_PREFIX": os.getenv("PAPERLESS_REDIS_PREFIX", ""),
|
||||
},
|
||||
}
|
||||
|
||||
|
@ -3,6 +3,7 @@ DJANGO_SETTINGS_MODULE = paperless.settings
|
||||
addopts = --pythonwarnings=all --cov --cov-report=html --cov-report=xml --numprocesses auto --maxprocesses=16 --quiet --durations=50
|
||||
env =
|
||||
PAPERLESS_DISABLE_DBHANDLER=true
|
||||
PAPERLESS_CACHE_BACKEND=django.core.cache.backends.locmem.LocMemCache
|
||||
|
||||
[coverage:run]
|
||||
source =
|
||||
|
Loading…
x
Reference in New Issue
Block a user