Normalize filenames and titles to NFC

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/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/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.