Messing around

paperless.conf.example

@@ -14,6 +14,7 @@
 # Paths and folders
 
 #PAPERLESS_CONSUMPTION_DIR=../consume
+#PAPERLESS_CONSUMPTION_FAILED_DIR=../consume/failed
 #PAPERLESS_DATA_DIR=../data
 #PAPERLESS_EMPTY_TRASH_DIR=
 #PAPERLESS_MEDIA_ROOT=../media

src/documents/consumer.py

@@ -174,6 +174,11 @@ class ConsumerPluginMixin:
     ):
         self._send_progress(100, 100, ProgressStatusOptions.FAILED, message)
         self.log.error(log_message or message, exc_info=exc_info)
+        # Move the file to the failed directory
+        if self.input_doc.original_file.exists():
+            self.input_doc.original_file.rename(
+                settings.CONSUMPTION_FAILED_DIR / self.input_doc.original_file.name,
+            )
         raise ConsumerError(f"{self.filename}: {log_message or message}") from exception
 
 
@@ -913,7 +918,6 @@ class ConsumerPlugin(
         except Exception:  # pragma: no cover
             pass
 
-
 class ConsumerPreflightPlugin(
     NoCleanupPluginMixin,
     NoSetupPluginMixin,

src/documents/tasks.py

@@ -37,6 +37,7 @@ from documents.consumer import ConsumerPreflightPlugin
 from documents.consumer import WorkflowTriggerPlugin
 from documents.data_models import ConsumableDocument
 from documents.data_models import DocumentMetadataOverrides
+from documents.data_models import DocumentSource
 from documents.double_sided import CollatePlugin
 from documents.file_handling import create_source_path_directory
 from documents.file_handling import generate_unique_filename
@@ -239,6 +240,23 @@ def consume_file(
     return msg
 
 
+@shared_task
+def retry_failed_file(task_id: str, clean: bool = False, skip_ocr: bool = False):
+    task = PaperlessTask.objects.get(task_id=task_id, status=states.FAILURE)
+    if task:
+        failed_file = settings.CONSUMPTION_FAILED_DIR / task.task_file_name
+        if not failed_file.exists():
+            logger.error(f"Failed file {failed_file} not found")
+            return
+
+        consume_file(
+            ConsumableDocument(
+                source=DocumentSource.ConsumeFolder,
+                original_file=failed_file,
+            ),
+        )
+
+
 @shared_task
 def sanity_check(*, scheduled=True, raise_on_error=True):
     messages = sanity_checker.check_sanity(scheduled=scheduled)

src/documents/tests/utils.py

@@ -37,6 +37,7 @@ def setup_directories():
     dirs.scratch_dir = Path(tempfile.mkdtemp()).resolve()
     dirs.media_dir = Path(tempfile.mkdtemp()).resolve()
     dirs.consumption_dir = Path(tempfile.mkdtemp()).resolve()
+    dirs.consumption_failed_dir = (dirs.consumption_dir / "failed").resolve()
     dirs.static_dir = Path(tempfile.mkdtemp()).resolve()
     dirs.index_dir = dirs.data_dir / "index"
     dirs.originals_dir = dirs.media_dir / "documents" / "originals"
@@ -58,6 +59,7 @@ def setup_directories():
         THUMBNAIL_DIR=dirs.thumbnail_dir,
         ARCHIVE_DIR=dirs.archive_dir,
         CONSUMPTION_DIR=dirs.consumption_dir,
+        CONSUMPTION_FAILED_DIR=dirs.consumption_failed_dir,
         LOGGING_DIR=dirs.logging_dir,
         INDEX_DIR=dirs.index_dir,
         STATIC_ROOT=dirs.static_dir,
@@ -74,6 +76,7 @@ def remove_dirs(dirs) -> None:
     shutil.rmtree(dirs.data_dir, ignore_errors=True)
     shutil.rmtree(dirs.scratch_dir, ignore_errors=True)
     shutil.rmtree(dirs.consumption_dir, ignore_errors=True)
+    shutil.rmtree(dirs.consumption_failed_dir, ignore_errors=True)
     shutil.rmtree(dirs.static_dir, ignore_errors=True)
     dirs.settings_override.disable()
 

src/paperless/checks.py

@@ -68,6 +68,10 @@ def paths_check(app_configs, **kwargs) -> list[Error]:
         + path_check("PAPERLESS_EMPTY_TRASH_DIR", settings.EMPTY_TRASH_DIR)
         + path_check("PAPERLESS_MEDIA_ROOT", settings.MEDIA_ROOT)
         + path_check("PAPERLESS_CONSUMPTION_DIR", settings.CONSUMPTION_DIR)
+        + path_check(
+            "PAPERLESS_CONSUMPTION_FAILED_DIR",
+            settings.CONSUMPTION_FAILED_DIR,
+        )
     )
 
 

src/paperless/settings/__init__.py

@@ -98,6 +98,11 @@ CONSUMPTION_DIR = get_path_from_env(
     BASE_DIR.parent / "consume",
 )
 
+CONSUMPTION_FAILED_DIR = __get_path(
+    "PAPERLESS_CONSUMPTION_FAILED_DIR",
+    CONSUMPTION_DIR / "failed",
+)
+
 # This will be created if it doesn't exist
 SCRATCH_DIR = get_path_from_env(
     "PAPERLESS_SCRATCH_DIR",
@@ -782,6 +787,8 @@ CONSUMER_IGNORE_PATTERNS = list(
         ),
     ),
 )
+if CONSUMPTION_DIR in CONSUMPTION_FAILED_DIR.parents:
+    CONSUMER_IGNORE_PATTERNS.append(CONSUMPTION_FAILED_DIR.name)
 
 # Directories to always ignore.  These are matched by directory name, not full path
 CONSUMER_IGNORE_DIRS = list(