Normalize and casefold input in TitleContentFilter

src/documents/consumer.py

@@ -46,6 +46,7 @@ from documents.signals.handlers import run_workflows
 from documents.templating.workflows import parse_w_workflow_placeholders
 from documents.utils import copy_basic_file_stats
 from documents.utils import copy_file_with_basic_stats
+from documents.utils import normalize_nfc
 from documents.utils import run_subprocess
 from paperless_mail.parsers import MailDocumentParser
 
@@ -111,7 +112,12 @@ class ConsumerPluginMixin:
 
         self.renew_logging_group()
 
-        self.filename = self.metadata.filename or self.input_doc.original_file.name
+        self.metadata.filename = normalize_nfc(self.metadata.filename)
+        self.metadata.title = normalize_nfc(self.metadata.title)
+
+        self.filename = normalize_nfc(
+            self.metadata.filename or self.input_doc.original_file.name,
+        )
 
     def _send_progress(
         self,
@@ -652,6 +658,8 @@ class ConsumerPlugin(
                     f"Error occurred parsing title override '{self.metadata.title}', falling back to original. Exception: {e}",
                 )
 
+        title = normalize_nfc(title)
+
         file_for_checksum = (
             self.unmodified_original
             if self.unmodified_original is not None

src/documents/file_handling.py

@@ -6,6 +6,7 @@ from django.conf import settings
 from documents.models import Document
 from documents.templating.filepath import validate_filepath_template_and_render
 from documents.templating.utils import convert_format_str_to_template_format
+from documents.utils import normalize_nfc
 
 
 def create_source_path_directory(source_path: Path) -> None:
@@ -55,11 +56,11 @@ def generate_unique_filename(doc, *, archive_filename=False) -> Path:
     """
     if archive_filename:
         old_filename: Path | None = (
-            Path(doc.archive_filename) if doc.archive_filename else None
+            Path(normalize_nfc(doc.archive_filename)) if doc.archive_filename else None
         )
         root = settings.ARCHIVE_DIR
     else:
-        old_filename = Path(doc.filename) if doc.filename else None
+        old_filename = Path(normalize_nfc(doc.filename)) if doc.filename else None
         root = settings.ORIGINALS_DIR
 
     # If generating archive filenames, try to make a name that is similar to
@@ -91,7 +92,7 @@ def generate_unique_filename(doc, *, archive_filename=False) -> Path:
         )
         if new_filename == old_filename:
             # still the same as before.
-            return new_filename
+            return Path(normalize_nfc(str(new_filename)))
 
         if (root / new_filename).exists():
             counter += 1
@@ -119,7 +120,7 @@ def format_filename(document: Document, template_str: str) -> str | None:
         "none",
     )  # backward compatibility
 
-    return rendered_filename
+    return normalize_nfc(rendered_filename)
 
 
 def generate_filename(
@@ -174,4 +175,4 @@ def generate_filename(
     if append_gpg and doc.storage_type == doc.STORAGE_TYPE_GPG:
         full_path = full_path.with_suffix(full_path.suffix + ".gpg")
 
-    return full_path
+    return Path(normalize_nfc(str(full_path)))

src/documents/filters.py

@@ -41,6 +41,7 @@ from documents.models import PaperlessTask
 from documents.models import ShareLink
 from documents.models import StoragePath
 from documents.models import Tag
+from documents.utils import normalize_nfc
 
 if TYPE_CHECKING:
     from collections.abc import Callable
@@ -162,7 +163,11 @@ class TitleContentFilter(Filter):
     def filter(self, qs, value):
         value = value.strip() if isinstance(value, str) else value
         if value:
-            return qs.filter(Q(title__icontains=value) | Q(content__icontains=value))
+            normalized = normalize_nfc(value) or ""
+            folded = normalized.casefold()
+            return qs.filter(
+                Q(title__icontains=folded) | Q(content__icontains=folded),
+            )
         else:
             return qs
 

src/documents/index.py

@@ -3,6 +3,7 @@ from __future__ import annotations
 import logging
 import math
 import re
+import unicodedata
 from collections import Counter
 from contextlib import contextmanager
 from datetime import datetime
@@ -58,6 +59,14 @@ if TYPE_CHECKING:
 logger = logging.getLogger("paperless.index")
 
 
+def _normalize_for_index(value: str | None) -> str | None:
+    """Normalize text to NFC for consistent search/index matching."""
+
+    if value is None:
+        return None
+    return unicodedata.normalize("NFC", value)
+
+
 def get_schema() -> Schema:
     return Schema(
         id=NUMERIC(stored=True, unique=True),
@@ -163,37 +172,41 @@ def update_document(writer: AsyncWriter, doc: Document) -> None:
     viewer_ids: str = ",".join([str(u.id) for u in users_with_perms])
     writer.update_document(
         id=doc.pk,
-        title=doc.title,
-        content=doc.content,
-        correspondent=doc.correspondent.name if doc.correspondent else None,
+        title=_normalize_for_index(doc.title),
+        content=_normalize_for_index(doc.content),
+        correspondent=_normalize_for_index(
+            doc.correspondent.name if doc.correspondent else None,
+        ),
         correspondent_id=doc.correspondent.id if doc.correspondent else None,
         has_correspondent=doc.correspondent is not None,
-        tag=tags if tags else None,
+        tag=_normalize_for_index(tags) if tags else None,
         tag_id=tags_ids if tags_ids else None,
         has_tag=len(tags) > 0,
-        type=doc.document_type.name if doc.document_type else None,
+        type=_normalize_for_index(
+            doc.document_type.name if doc.document_type else None,
+        ),
         type_id=doc.document_type.id if doc.document_type else None,
         has_type=doc.document_type is not None,
         created=datetime.combine(doc.created, time.min),
         added=doc.added,
         asn=asn,
         modified=doc.modified,
-        path=doc.storage_path.name if doc.storage_path else None,
+        path=_normalize_for_index(doc.storage_path.name if doc.storage_path else None),
         path_id=doc.storage_path.id if doc.storage_path else None,
         has_path=doc.storage_path is not None,
-        notes=notes,
+        notes=_normalize_for_index(notes),
         num_notes=len(notes),
-        custom_fields=custom_fields,
+        custom_fields=_normalize_for_index(custom_fields),
         custom_field_count=len(doc.custom_fields.all()),
         has_custom_fields=len(custom_fields) > 0,
         custom_fields_id=custom_fields_ids if custom_fields_ids else None,
-        owner=doc.owner.username if doc.owner else None,
+        owner=_normalize_for_index(doc.owner.username if doc.owner else None),
         owner_id=doc.owner.id if doc.owner else None,
         has_owner=doc.owner is not None,
         viewer_id=viewer_ids if viewer_ids else None,
         checksum=doc.checksum,
         page_count=doc.page_count,
-        original_filename=doc.original_filename,
+        original_filename=_normalize_for_index(doc.original_filename),
         is_shared=len(viewer_ids) > 0,
     )
     logger.debug(f"Index updated for document {doc.pk}.")
@@ -421,7 +434,7 @@ class LocalDateParser(English):
 
 class DelayedFullTextQuery(DelayedQuery):
     def _get_query(self) -> tuple:
-        q_str = self.query_params["query"]
+        q_str = _normalize_for_index(self.query_params["query"]) or ""
         q_str = rewrite_natural_date_keywords(q_str)
         qp = MultifieldParser(
             [
@@ -460,7 +473,12 @@ class DelayedFullTextQuery(DelayedQuery):
 class DelayedMoreLikeThisQuery(DelayedQuery):
     def _get_query(self) -> tuple:
         more_like_doc_id = int(self.query_params["more_like_id"])
-        content = Document.objects.get(id=more_like_doc_id).content
+        content = (
+            _normalize_for_index(
+                Document.objects.get(id=more_like_doc_id).content,
+            )
+            or ""
+        )
 
         docnum = self.searcher.document_number(id=more_like_doc_id)
         kts = self.searcher.key_terms_from_text(
@@ -488,6 +506,7 @@ def autocomplete(
     Mimics whoosh.reading.IndexReader.most_distinctive_terms with permissions
     and without scoring
     """
+    term = _normalize_for_index(term) or ""
     terms = []
 
     with ix.searcher(weighting=TF_IDF()) as s:

src/documents/matching.py

@@ -2,10 +2,12 @@ from __future__ import annotations
 
 import logging
 import re
+import unicodedata
 from fnmatch import fnmatch
 from fnmatch import translate as fnmatch_translate
 from typing import TYPE_CHECKING
 
+from django.db.models import Q
 from rest_framework import serializers
 
 from documents.data_models import ConsumableDocument
@@ -21,6 +23,7 @@ from documents.models import Workflow
 from documents.models import WorkflowTrigger
 from documents.permissions import get_objects_for_user_owner_aware
 from documents.regex import safe_regex_search
+from documents.utils import normalize_nfc
 
 if TYPE_CHECKING:
     from django.db.models import QuerySet
@@ -30,6 +33,34 @@ if TYPE_CHECKING:
 logger = logging.getLogger("paperless.matching")
 
 
+def _normalize_glob_value(value: str) -> str:
+    """Normalize strings for glob-style matching (case-insensitive)."""
+
+    return (normalize_nfc(value) or "").casefold()
+
+
+def _normalized_fnmatch(name: str, pattern: str) -> bool:
+    """Canonicalize Unicode and compare using fnmatch semantics."""
+
+    return fnmatch(_normalize_glob_value(name), _normalize_glob_value(pattern))
+
+
+def _glob_regex_variants(pattern: str) -> list[str]:
+    """
+    Build regex patterns that match both NFC and NFD forms of a glob pattern.
+    Using both forms lets DB prefilters remain Unicode-normalization agnostic.
+    """
+
+    regexes = set()
+    for normalized in {
+        normalize_nfc(pattern) or "",
+        unicodedata.normalize("NFD", pattern),
+    }:
+        regex = fnmatch_translate(normalized).lstrip("^").rstrip("$")
+        regexes.add(regex)
+    return list(regexes)
+
+
 def log_reason(
     matching_model: MatchingModel | WorkflowTrigger,
     document: Document,
@@ -305,9 +336,9 @@ def consumable_document_matches_workflow(
     if (
         trigger.filter_filename is not None
         and len(trigger.filter_filename) > 0
-        and not fnmatch(
-            document.original_file.name.lower(),
-            trigger.filter_filename.lower(),
+        and not _normalized_fnmatch(
+            document.original_file.name,
+            trigger.filter_filename,
         )
     ):
         reason = (
@@ -328,7 +359,7 @@ def consumable_document_matches_workflow(
     if (
         trigger.filter_path is not None
         and len(trigger.filter_path) > 0
-        and not fnmatch(
+        and not _normalized_fnmatch(
             match_against,
             trigger.filter_path,
         )
@@ -492,9 +523,9 @@ def existing_document_matches_workflow(
         trigger.filter_filename is not None
         and len(trigger.filter_filename) > 0
         and document.original_filename is not None
-        and not fnmatch(
-            document.original_filename.lower(),
-            trigger.filter_filename.lower(),
+        and not _normalized_fnmatch(
+            document.original_filename,
+            trigger.filter_filename,
         )
     ):
         return (
@@ -573,8 +604,11 @@ def prefilter_documents_by_workflowtrigger(
         documents = documents.annotate(**annotations).filter(custom_field_q)
 
     if trigger.filter_filename:
-        regex = fnmatch_translate(trigger.filter_filename).lstrip("^").rstrip("$")
-        documents = documents.filter(original_filename__iregex=regex)
+        regexes = _glob_regex_variants(trigger.filter_filename)
+        filename_q = Q()
+        for regex in regexes:
+            filename_q |= Q(original_filename__iregex=regex)
+        documents = documents.filter(filename_q)
 
     return documents
 

src/documents/tests/test_api_search.py

@@ -89,6 +89,23 @@ class TestDocumentSearchApi(DirectoriesMixin, APITestCase):
         self.assertEqual(len(results), 0)
         self.assertCountEqual(response.data["all"], [])
 
+    def test_search_handles_diacritics_normalization(self):
+        doc = Document.objects.create(
+            title="certida\u0303o de nascimento",
+            content="birth record without keyword",
+            checksum="D",
+            pk=10,
+        )
+        with AsyncWriter(index.open_index()) as writer:
+            index.update_document(writer, doc)
+
+        response = self.client.get("/api/documents/?query=certidão")
+        self.assertEqual(response.status_code, status.HTTP_200_OK)
+        results = response.data["results"]
+        self.assertEqual(response.data["count"], 1)
+        self.assertEqual(len(results), 1)
+        self.assertEqual(results[0]["id"], doc.id)
+
     def test_search_custom_field_ordering(self):
         custom_field = CustomField.objects.create(
             name="Sortable field",

src/documents/tests/test_consumer.py

@@ -290,6 +290,23 @@ class TestConsumer(
 
         self._assert_first_last_send_progress()
 
+    def test_override_filename_normalized(self):
+        filename = self.get_test_file()
+        override_filename = "Inhaltsu\u0308bersicht.pdf"
+
+        with self.get_consumer(
+            filename,
+            DocumentMetadataOverrides(filename=override_filename),
+        ) as consumer:
+            consumer.run()
+
+            document = Document.objects.first()
+
+        self.assertIsNotNone(document)
+        self.assertEqual(document.original_filename, "Inhaltsübersicht.pdf")
+        self.assertEqual(document.title, "Inhaltsübersicht")
+        self._assert_first_last_send_progress()
+
     def testOverrideTitle(self):
         with self.get_consumer(
             self.get_test_file(),
@@ -304,6 +321,25 @@ class TestConsumer(
         self.assertEqual(document.title, "Override Title")
         self._assert_first_last_send_progress()
 
+    @override_settings(FILENAME_FORMAT="{{ title }}")
+    def test_filename_format_normalized(self):
+        filename = self.get_test_file()
+        title = "Inhaltsu\u0308bersicht Faszination"
+
+        with self.get_consumer(
+            filename,
+            DocumentMetadataOverrides(title=title),
+        ) as consumer:
+            consumer.run()
+
+            document = Document.objects.first()
+
+        self.assertIsNotNone(document)
+        self.assertEqual(document.title, "Inhaltsübersicht Faszination")
+        self.assertEqual(document.filename, "Inhaltsübersicht Faszination.pdf")
+        self.assertIsFile(document.source_path)
+        self._assert_first_last_send_progress()
+
     def testOverrideCorrespondent(self):
         c = Correspondent.objects.create(name="test")
 

src/documents/tests/test_workflows.py

@@ -557,6 +557,50 @@ class TestWorkflows(
         expected_str = f"Document filename {test_file.name} does not match"
         self.assertIn(expected_str, cm.output[1])
 
+    def test_workflow_match_filename_diacritics_normalized(self):
+        """
+        GIVEN:
+            - Consumption workflow filtering on filename with diacritics
+        WHEN:
+            - File with decomposed Unicode filename is consumed
+        THEN:
+            - Workflow still matches and applies overrides
+        """
+        trigger = WorkflowTrigger.objects.create(
+            type=WorkflowTrigger.WorkflowTriggerType.CONSUMPTION,
+            sources=f"{DocumentSource.ApiUpload},{DocumentSource.ConsumeFolder},{DocumentSource.MailFetch}",
+            filter_filename="*račun*",
+        )
+        action = WorkflowAction.objects.create(
+            assign_title="Diacritics matched",
+        )
+        action.save()
+
+        w = Workflow.objects.create(
+            name="Workflow 1",
+            order=0,
+        )
+        w.triggers.add(trigger)
+        w.actions.add(action)
+        w.save()
+
+        decomposed_name = "rac\u030cun.pdf"
+        test_file = shutil.copy(
+            self.SAMPLE_DIR / "simple.pdf",
+            self.dirs.scratch_dir / decomposed_name,
+        )
+
+        with mock.patch("documents.tasks.ProgressManager", DummyProgressManager):
+            tasks.consume_file(
+                ConsumableDocument(
+                    source=DocumentSource.ConsumeFolder,
+                    original_file=test_file,
+                ),
+                None,
+            )
+            document = Document.objects.first()
+            self.assertEqual(document.title, "Diacritics matched")
+
     def test_workflow_no_match_path(self):
         """
         GIVEN:
@@ -946,6 +990,35 @@ class TestWorkflows(
         self.assertEqual(doc.correspondent, self.c2)
         self.assertEqual(doc.title, f"Doc created in {created.year}")
 
+    def test_document_added_filename_diacritics_normalized(self):
+        trigger = WorkflowTrigger.objects.create(
+            type=WorkflowTrigger.WorkflowTriggerType.DOCUMENT_ADDED,
+            filter_filename="*račun*",
+        )
+        action = WorkflowAction.objects.create(
+            assign_title="Matched diacritics",
+        )
+        w = Workflow.objects.create(
+            name="Workflow 1",
+            order=0,
+        )
+        w.triggers.add(trigger)
+        w.actions.add(action)
+        w.save()
+
+        doc = Document.objects.create(
+            title="sample test",
+            correspondent=self.c,
+            original_filename="rac\u030cun.pdf",
+        )
+
+        document_consumption_finished.send(
+            sender=self.__class__,
+            document=doc,
+        )
+
+        self.assertEqual(doc.title, "Matched diacritics")
+
     def test_document_added_no_match_filename(self):
         trigger = WorkflowTrigger.objects.create(
             type=WorkflowTrigger.WorkflowTriggerType.DOCUMENT_ADDED,

src/documents/utils.py

@@ -1,5 +1,7 @@
 import logging
 import shutil
+import unicodedata
+from os import PathLike
 from os import utime
 from pathlib import Path
 from subprocess import CompletedProcess
@@ -16,6 +18,14 @@ def _coerce_to_path(
     return Path(source).resolve(), Path(dest).resolve()
 
 
+def normalize_nfc(value: str | PathLike[str] | None) -> str | None:
+    """Return NFC-normalized string for filesystem-safe comparisons."""
+
+    if value is None:
+        return None
+    return unicodedata.normalize("NFC", str(value))
+
+
 def copy_basic_file_stats(source: Path | str, dest: Path | str) -> None:
     """
     Copies only the m_time and a_time attributes from source to destination.