Basic option selection

paperless.conf.example

@@ -18,6 +18,7 @@
 # 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-ui/messages.xlf

@@ -1994,120 +1994,141 @@
           <context context-type="linenumber">72</context>
         </context-group>
       </trans-unit>
+      <trans-unit id="7934833136974560675" datatype="html">
+        <source>Retry</source>
+        <context-group purpose="location">
+          <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.html</context>
+          <context context-type="linenumber">86</context>
+        </context-group>
+      </trans-unit>
       <trans-unit id="1536087519743707362" datatype="html">
         <source>Dismiss</source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.html</context>
-          <context context-type="linenumber">85</context>
+          <context context-type="linenumber">90</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.ts</context>
-          <context context-type="linenumber">68</context>
+          <context context-type="linenumber">71</context>
         </context-group>
       </trans-unit>
       <trans-unit id="2134950584701094962" datatype="html">
         <source>Open Document</source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.html</context>
-          <context context-type="linenumber">90</context>
+          <context context-type="linenumber">95</context>
         </context-group>
       </trans-unit>
       <trans-unit id="428536141871853903" datatype="html">
         <source>{VAR_PLURAL, plural, =1 {One <x id="INTERPOLATION"/> task} other {<x id="INTERPOLATION_1"/> total <x id="INTERPOLATION"/> tasks}}</source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.html</context>
-          <context context-type="linenumber">109</context>
+          <context context-type="linenumber">114</context>
         </context-group>
       </trans-unit>
       <trans-unit id="1943508481059904274" datatype="html">
         <source> (<x id="INTERPOLATION" equiv-text="{{selectedTasks.size}}"/> selected)</source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.html</context>
-          <context context-type="linenumber">111</context>
+          <context context-type="linenumber">116</context>
         </context-group>
       </trans-unit>
       <trans-unit id="5639839509673911668" datatype="html">
         <source>Failed<x id="START_BLOCK_IF" equiv-text="@if (tasksService.failedFileTasks.length &gt; 0) {"/><x id="START_TAG_SPAN" ctype="x-span" equiv-text="&lt;span class=&quot;badge bg-danger ms-2&quot;&gt;"/><x id="INTERPOLATION" equiv-text="{{tasksService.failedFileTasks.length}}"/><x id="CLOSE_TAG_SPAN" ctype="x-span" equiv-text="&lt;/span&gt;"/><x id="CLOSE_BLOCK_IF" equiv-text="}"/></source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.html</context>
-          <context context-type="linenumber">123,125</context>
+          <context context-type="linenumber">128,130</context>
         </context-group>
       </trans-unit>
       <trans-unit id="8210778930307085868" datatype="html">
         <source>Complete<x id="START_BLOCK_IF" equiv-text="@if (tasksService.completedFileTasks.length &gt; 0) {"/><x id="START_TAG_SPAN" ctype="x-span" equiv-text="&lt;span class=&quot;badge bg-secondary ms-2&quot;&gt;"/><x id="INTERPOLATION" equiv-text="{{tasksService.completedFileTasks.length}}"/><x id="CLOSE_TAG_SPAN" ctype="x-span" equiv-text="&lt;/span&gt;"/><x id="CLOSE_BLOCK_IF" equiv-text="}"/></source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.html</context>
-          <context context-type="linenumber">131,133</context>
+          <context context-type="linenumber">136,138</context>
         </context-group>
       </trans-unit>
       <trans-unit id="3522801015717851360" datatype="html">
         <source>Started<x id="START_BLOCK_IF" equiv-text="@if (tasksService.startedFileTasks.length &gt; 0) {"/><x id="START_TAG_SPAN" ctype="x-span" equiv-text="&lt;span class=&quot;badge bg-secondary ms-2&quot;&gt;"/><x id="INTERPOLATION" equiv-text="{{tasksService.startedFileTasks.length}}"/><x id="CLOSE_TAG_SPAN" ctype="x-span" equiv-text="&lt;/span&gt;"/><x id="CLOSE_BLOCK_IF" equiv-text="}"/></source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.html</context>
-          <context context-type="linenumber">139,141</context>
+          <context context-type="linenumber">144,146</context>
         </context-group>
       </trans-unit>
       <trans-unit id="2341807459308874922" datatype="html">
         <source>Queued<x id="START_BLOCK_IF" equiv-text="@if (tasksService.queuedFileTasks.length &gt; 0) {"/><x id="START_TAG_SPAN" ctype="x-span" equiv-text="&lt;span class=&quot;badge bg-secondary ms-2&quot;&gt;"/><x id="INTERPOLATION" equiv-text="{{tasksService.queuedFileTasks.length}}"/><x id="CLOSE_TAG_SPAN" ctype="x-span" equiv-text="&lt;/span&gt;"/><x id="CLOSE_BLOCK_IF" equiv-text="}"/></source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.html</context>
-          <context context-type="linenumber">147,149</context>
+          <context context-type="linenumber">152,154</context>
         </context-group>
       </trans-unit>
       <trans-unit id="5404910960991552159" datatype="html">
         <source>Dismiss selected</source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.ts</context>
-          <context context-type="linenumber">31</context>
+          <context context-type="linenumber">33</context>
         </context-group>
       </trans-unit>
       <trans-unit id="8829078752502782653" datatype="html">
         <source>Dismiss all</source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.ts</context>
-          <context context-type="linenumber">32</context>
+          <context context-type="linenumber">34</context>
         </context-group>
       </trans-unit>
       <trans-unit id="1323591410517879795" datatype="html">
         <source>Confirm Dismiss All</source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.ts</context>
-          <context context-type="linenumber">65</context>
+          <context context-type="linenumber">68</context>
         </context-group>
       </trans-unit>
       <trans-unit id="4157200209636243740" datatype="html">
         <source>Dismiss all <x id="PH" equiv-text="tasks.size"/> tasks?</source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.ts</context>
-          <context context-type="linenumber">66</context>
+          <context context-type="linenumber">69</context>
+        </context-group>
+      </trans-unit>
+      <trans-unit id="7611027432301841688" datatype="html">
+        <source>Retrying task...</source>
+        <context-group purpose="location">
+          <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.ts</context>
+          <context context-type="linenumber">92</context>
+        </context-group>
+      </trans-unit>
+      <trans-unit id="5445438607105804721" datatype="html">
+        <source>Failed to retry task</source>
+        <context-group purpose="location">
+          <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.ts</context>
+          <context context-type="linenumber">95</context>
         </context-group>
       </trans-unit>
       <trans-unit id="9011556615675272238" datatype="html">
         <source>queued</source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.ts</context>
-          <context context-type="linenumber">135</context>
+          <context context-type="linenumber">149</context>
         </context-group>
       </trans-unit>
       <trans-unit id="6415892379431855826" datatype="html">
         <source>started</source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.ts</context>
-          <context context-type="linenumber">137</context>
+          <context context-type="linenumber">151</context>
         </context-group>
       </trans-unit>
       <trans-unit id="7510279840486540181" datatype="html">
         <source>completed</source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.ts</context>
-          <context context-type="linenumber">139</context>
+          <context context-type="linenumber">153</context>
         </context-group>
       </trans-unit>
       <trans-unit id="4083337005045748464" datatype="html">
         <source>failed</source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/admin/tasks/tasks.component.ts</context>
-          <context context-type="linenumber">141</context>
+          <context context-type="linenumber">155</context>
         </context-group>
       </trans-unit>
       <trans-unit id="3418677553313974490" datatype="html">

src-ui/src/app/components/admin/tasks/tasks.component.html

@@ -81,6 +81,9 @@
           </td>
           <td scope="row">
             <div class="btn-group" role="group">
+              @if (task.status === PaperlessTaskStatus.Failed) {
+                <ng-container *ngTemplateOutlet="retryDropdown; context: { task: task }"></ng-container>
+              }
               <button class="btn btn-sm btn-outline-secondary" (click)="dismissTask(task); $event.stopPropagation();" *pngxIfPermissions="{ action: PermissionAction.Change, type: PermissionType.PaperlessTask }">
                 <i-bs name="check"></i-bs>&nbsp;<ng-container i18n>Dismiss</ng-container>
               </button>
@@ -153,3 +156,25 @@
   </li>
 </ul>
 <div [ngbNavOutlet]="nav"></div>
+
+<ng-template #retryDropdown let-task="task">
+  <div ngbDropdown>
+    <button class="btn btn-sm btn-outline-primary" (click)="$event.stopImmediatePropagation()" ngbDropdownToggle>
+      <i-bs name="arrow-repeat"></i-bs>&nbsp;<ng-container i18n>Retry</ng-container>
+    </button>
+    <div ngbDropdownMenu class="shadow retry-dropdown">
+      <div class="p-2">
+        <ul class="list-group list-group-flush">
+            <li class="list-group-item small" i18n>
+              <pngx-input-check [(ngModel)]="retryClean" i18n-title title="Attempt to clean pdf"></pngx-input-check>
+            </li>
+        </ul>
+        <div class="d-flex justify-content-end">
+          <button class="btn btn-sm btn-outline-primary" (click)="retryTask(task); $event.stopPropagation();">
+            <ng-container i18n>Proceed</ng-container>
+          </button>
+        </div>
+      </div>
+    </div>
+  </div>
+</ng-template>

src-ui/src/app/components/admin/tasks/tasks.component.scss

@@ -26,3 +26,7 @@ pre {
         max-width: 150px;
     }
 }
+
+.retry-dropdown {
+    width: 300px;
+}

src-ui/src/app/components/admin/tasks/tasks.component.spec.ts

@@ -31,6 +31,9 @@ import { PermissionsGuard } from 'src/app/guards/permissions.guard'
 import { NgxBootstrapIconsModule, allIcons } from 'ngx-bootstrap-icons'
 import { FormsModule } from '@angular/forms'
 import { provideHttpClient, withInterceptorsFromDi } from '@angular/common/http'
+import { ToastService } from 'src/app/services/toast.service'
+import { of, throwError } from 'rxjs'
+import { CheckComponent } from '../../common/input/check/check.component'
 
 const tasks: PaperlessTask[] = [
   {
@@ -115,6 +118,7 @@ describe('TasksComponent', () => {
   let modalService: NgbModal
   let router: Router
   let httpTestingController: HttpTestingController
+  let toastService: ToastService
   let reloadSpy
 
   beforeEach(async () => {
@@ -125,6 +129,7 @@ describe('TasksComponent', () => {
         IfPermissionsDirective,
         CustomDatePipe,
         ConfirmDialogComponent,
+        CheckComponent,
       ],
       imports: [
         NgbModule,
@@ -152,6 +157,7 @@ describe('TasksComponent', () => {
     httpTestingController = TestBed.inject(HttpTestingController)
     modalService = TestBed.inject(NgbModal)
     router = TestBed.inject(Router)
+    toastService = TestBed.inject(ToastService)
     fixture = TestBed.createComponent(TasksComponent)
     component = fixture.componentInstance
     jest.useFakeTimers()
@@ -173,8 +179,10 @@ describe('TasksComponent', () => {
       `Failed${currentTasksLength}`
     )
     expect(
-      fixture.debugElement.queryAll(By.css('table input[type="checkbox"]'))
+      fixture.debugElement.queryAll(
-    ).toHaveLength(currentTasksLength + 1)
+        By.css('table td > .form-check input[type="checkbox"]')
+      )
+    ).toHaveLength(currentTasksLength)
 
     currentTasksLength = tasks.filter(
       (t) => t.status === PaperlessTaskStatus.Complete
@@ -289,4 +297,20 @@ describe('TasksComponent', () => {
     jest.advanceTimersByTime(6000)
     expect(reloadSpy).toHaveBeenCalledTimes(2)
   })
+
+  it('should retry a task, show toast on error or success', () => {
+    const retrySpy = jest.spyOn(tasksService, 'retryTask')
+    const toastInfoSpy = jest.spyOn(toastService, 'showInfo')
+    const toastErrorSpy = jest.spyOn(toastService, 'showError')
+    retrySpy.mockReturnValueOnce(of({ task_id: '123' }))
+    component.retryTask(tasks[0])
+    expect(retrySpy).toHaveBeenCalledWith(tasks[0], false)
+    expect(toastInfoSpy).toHaveBeenCalledWith('Retrying task...')
+    retrySpy.mockReturnValueOnce(throwError(() => new Error('test')))
+    component.retryTask(tasks[0])
+    expect(toastErrorSpy).toHaveBeenCalledWith(
+      'Failed to retry task',
+      new Error('test')
+    )
+  })
 })

src-ui/src/app/components/admin/tasks/tasks.component.ts

@@ -2,10 +2,11 @@ import { Component, OnInit, OnDestroy } from '@angular/core'
 import { Router } from '@angular/router'
 import { NgbModal } from '@ng-bootstrap/ng-bootstrap'
 import { first } from 'rxjs'
-import { PaperlessTask } from 'src/app/data/paperless-task'
+import { PaperlessTask, PaperlessTaskStatus } from 'src/app/data/paperless-task'
 import { TasksService } from 'src/app/services/tasks.service'
 import { ConfirmDialogComponent } from '../../common/confirm-dialog/confirm-dialog.component'
 import { ComponentWithPermissions } from '../../with-permissions/with-permissions.component'
+import { ToastService } from 'src/app/services/toast.service'
 
 @Component({
   selector: 'pngx-tasks',
@@ -16,6 +17,7 @@ export class TasksComponent
   extends ComponentWithPermissions
   implements OnInit, OnDestroy
 {
+  public PaperlessTaskStatus = PaperlessTaskStatus
   public activeTab: string
   public selectedTasks: Set<number> = new Set()
   public togggleAll: boolean = false
@@ -26,6 +28,8 @@ export class TasksComponent
 
   public autoRefreshInterval: any
 
+  public retryClean: boolean = false
+
   get dismissButtonText(): string {
     return this.selectedTasks.size > 0
       ? $localize`Dismiss selected`
@@ -35,6 +39,7 @@ export class TasksComponent
   constructor(
     public tasksService: TasksService,
     private modalService: NgbModal,
+    private toastService: ToastService,
     private readonly router: Router
   ) {
     super()
@@ -83,6 +88,17 @@ export class TasksComponent
     this.router.navigate(['documents', task.related_document])
   }
 
+  retryTask(task: PaperlessTask) {
+    this.tasksService.retryTask(task, this.retryClean).subscribe({
+      next: () => {
+        this.toastService.showInfo($localize`Retrying task...`)
+      },
+      error: (e) => {
+        this.toastService.showError($localize`Failed to retry task`, e)
+      },
+    })
+  }
+
   expandTask(task: PaperlessTask) {
     this.expandedTask = this.expandedTask == task.id ? undefined : task.id
   }

src-ui/src/app/services/tasks.service.spec.ts

@@ -118,4 +118,29 @@ describe('TasksService', () => {
     expect(tasksService.queuedFileTasks).toHaveLength(1)
     expect(tasksService.startedFileTasks).toHaveLength(1)
   })
+
+  it('should call retry task api endpoint', () => {
+    const task = {
+      id: 1,
+      type: PaperlessTaskType.File,
+      status: PaperlessTaskStatus.Failed,
+      acknowledged: false,
+      task_id: '1234',
+      task_file_name: 'file1.pdf',
+      date_created: new Date(),
+    }
+
+    tasksService.retryTask(task, true).subscribe()
+    const reloadSpy = jest.spyOn(tasksService, 'reload')
+    const req = httpTestingController.expectOne(
+      `${environment.apiBaseUrl}tasks/${task.id}/retry/`
+    )
+    expect(req.request.method).toEqual('POST')
+    expect(req.request.body).toEqual({
+      clean: true,
+    })
+    req.flush({ task_id: 12345 })
+    expect(reloadSpy).toHaveBeenCalled()
+    httpTestingController.expectOne(`${environment.apiBaseUrl}tasks/`).flush([])
+  })
 })

src-ui/src/app/services/tasks.service.ts

@@ -1,7 +1,7 @@
 import { HttpClient } from '@angular/common/http'
 import { Injectable } from '@angular/core'
-import { Subject } from 'rxjs'
+import { Observable, Subject } from 'rxjs'
-import { first, takeUntil } from 'rxjs/operators'
+import { first, takeUntil, tap } from 'rxjs/operators'
 import {
   PaperlessTask,
   PaperlessTaskStatus,
@@ -73,6 +73,20 @@ export class TasksService {
       })
   }
 
+  public retryTask(task: PaperlessTask, clean: boolean): Observable<any> {
+    return this.http
+      .post(`${this.baseUrl}tasks/${task.id}/retry/`, {
+        clean,
+      })
+      .pipe(
+        takeUntil(this.unsubscribeNotifer),
+        first(),
+        tap(() => {
+          this.reload()
+        })
+      )
+  }
+
   public cancelPending(): void {
     this.unsubscribeNotifer.next(true)
   }

src/documents/consumer.py

@@ -148,6 +148,17 @@ class ConsumerPlugin(
     ):
         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()
+            and not Path(
+                settings.CONSUMPTION_FAILED_DIR / self.input_doc.original_file.name,
+            ).exists()
+        ):
+            copy_file_with_basic_stats(
+                self.input_doc.original_file,
+                settings.CONSUMPTION_FAILED_DIR / self.input_doc.original_file.name,
+            )
         raise ConsumerError(f"{self.filename}: {log_message or message}") from exception
 
     def pre_check_file_exists(self):

src/documents/models.py

@@ -679,24 +679,28 @@ class PaperlessTask(models.Model):
         verbose_name=_("Task State"),
         help_text=_("Current state of the task being run"),
     )
+
     date_created = models.DateTimeField(
         null=True,
         default=timezone.now,
         verbose_name=_("Created DateTime"),
         help_text=_("Datetime field when the task result was created in UTC"),
     )
+
     date_started = models.DateTimeField(
         null=True,
         default=None,
         verbose_name=_("Started DateTime"),
         help_text=_("Datetime field when the task was started in UTC"),
     )
+
     date_done = models.DateTimeField(
         null=True,
         default=None,
         verbose_name=_("Completed DateTime"),
         help_text=_("Datetime field when the task was completed in UTC"),
     )
+
     result = models.TextField(
         null=True,
         default=None,

src/documents/serialisers.py

@@ -1585,6 +1585,14 @@ class TasksViewSerializer(serializers.ModelSerializer):
         return result
 
 
+class RetryTaskSerializer(serializers.Serializer):
+    clean = serializers.BooleanField(
+        default=False,
+        write_only=True,
+        required=False,
+    )
+
+
 class AcknowledgeTasksViewSerializer(serializers.Serializer):
     tasks = serializers.ListField(
         required=True,

src/documents/signals/handlers.py

@@ -1,6 +1,7 @@
 import logging
 import os
 import shutil
+from pathlib import Path
 
 from celery import states
 from celery.signals import before_task_publish
@@ -520,6 +521,19 @@ def update_filename_and_move_files(
             )
 
 
+@receiver(models.signals.post_save, sender=PaperlessTask)
+def cleanup_failed_documents(sender, instance: PaperlessTask, **kwargs):
+    if instance.status != states.FAILURE or not instance.acknowledged:
+        return
+
+    if instance.task_file_name:
+        try:
+            Path(settings.CONSUMPTION_FAILED_DIR / instance.task_file_name).unlink()
+            logger.debug(f"Cleaned up failed file {instance.task_file_name}")
+        except FileNotFoundError:
+            logger.warning(f"Failed to clean up failed file {instance.task_file_name}")
+
+
 def set_log_entry(sender, document: Document, logging_group=None, **kwargs):
     ct = ContentType.objects.get(model="document")
     user = User.objects.get(username="consumer")

src/documents/tasks.py

@@ -9,6 +9,7 @@ from tempfile import TemporaryDirectory
 import tqdm
 from celery import Task
 from celery import shared_task
+from celery import states
 from django.conf import settings
 from django.db import models
 from django.db import transaction
@@ -27,12 +28,14 @@ from documents.consumer import ConsumerPlugin
 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
 from documents.models import Correspondent
 from documents.models import Document
 from documents.models import DocumentType
+from documents.models import PaperlessTask
 from documents.models import StoragePath
 from documents.models import Tag
 from documents.parsers import DocumentParser
@@ -44,6 +47,8 @@ from documents.plugins.helpers import ProgressStatusOptions
 from documents.sanity_checker import SanityCheckFailedException
 from documents.signals import document_updated
 from documents.signals.handlers import cleanup_document_deletion
+from documents.utils import copy_file_with_basic_stats
+from documents.utils import run_subprocess
 
 if settings.AUDIT_LOG_ENABLED:
     from auditlog.models import LogEntry
@@ -169,6 +174,48 @@ 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"File {failed_file} not found")
+            raise FileNotFoundError(f"File {failed_file} not found")
+        working_copy = settings.SCRATCH_DIR / failed_file.name
+        copy_file_with_basic_stats(failed_file, working_copy)
+
+        if clean:
+            try:
+                result = run_subprocess(
+                    [
+                        "qpdf",
+                        "--replace-input",
+                        "--warning-exit-0",
+                        working_copy,
+                    ],
+                    logger=logger,
+                )
+                if result.returncode != 0:
+                    raise Exception(
+                        f"qpdf failed with exit code {result.returncode}, error: {result.stderr}",
+                    )
+                else:
+                    logger.debug("PDF cleaned successfully")
+            except Exception as e:
+                logger.error(f"Error while cleaning PDF: {e}")
+                raise e
+
+        task = consume_file.delay(
+            ConsumableDocument(
+                source=DocumentSource.ConsumeFolder,
+                original_file=working_copy,
+            ),
+        )
+
+        return task.id
+
+
 @shared_task
 def sanity_check():
     messages = sanity_checker.check_sanity()

src/documents/tests/test_tasks.py

@@ -1,5 +1,8 @@
 import os
+import shutil
+import uuid
 from datetime import timedelta
+from pathlib import Path
 from unittest import mock
 
 from django.conf import settings
@@ -7,15 +10,22 @@ from django.test import TestCase
 from django.utils import timezone
 
 from documents import tasks
+from documents.data_models import ConsumableDocument
+from documents.data_models import DocumentSource
 from documents.models import Correspondent
 from documents.models import Document
 from documents.models import DocumentType
+from documents.models import PaperlessTask
 from documents.models import Tag
 from documents.sanity_checker import SanityCheckFailedException
 from documents.sanity_checker import SanityCheckMessages
+from documents.signals.handlers import before_task_publish_handler
+from documents.signals.handlers import task_failure_handler
 from documents.tests.test_classifier import dummy_preprocess
 from documents.tests.utils import DirectoriesMixin
+from documents.tests.utils import DummyProgressManager
 from documents.tests.utils import FileSystemAssertsMixin
+from documents.tests.utils import SampleDirMixin
 
 
 class TestIndexReindex(DirectoriesMixin, TestCase):
@@ -184,3 +194,68 @@ class TestEmptyTrashTask(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
 
         tasks.empty_trash()
         self.assertEqual(Document.global_objects.count(), 0)
+
+
+class TestRetryConsumeTask(
+    DirectoriesMixin,
+    SampleDirMixin,
+    FileSystemAssertsMixin,
+    TestCase,
+):
+    def do_failed_task(self, test_file: Path) -> PaperlessTask:
+        temp_copy = self.dirs.scratch_dir / test_file.name
+        shutil.copy(test_file, temp_copy)
+
+        headers = {
+            "id": str(uuid.uuid4()),
+            "task": "documents.tasks.consume_file",
+        }
+        body = (
+            # args
+            (
+                ConsumableDocument(
+                    source=DocumentSource.ConsumeFolder,
+                    original_file=str(temp_copy),
+                ),
+                None,
+            ),
+            # kwargs
+            {},
+            # celery stuff
+            {"callbacks": None, "errbacks": None, "chain": None, "chord": None},
+        )
+        before_task_publish_handler(headers=headers, body=body)
+
+        with mock.patch("documents.tasks.ProgressManager", DummyProgressManager):
+            with self.assertRaises(Exception):
+                tasks.consume_file(
+                    ConsumableDocument(
+                        source=DocumentSource.ConsumeFolder,
+                        original_file=temp_copy,
+                    ),
+                )
+
+        task_failure_handler(
+            task_id=headers["id"],
+            exception="Example failure",
+        )
+
+        task = PaperlessTask.objects.first()
+        # Ensure the file is moved to the failed dir
+        self.assertIsFile(settings.CONSUMPTION_FAILED_DIR / task.task_file_name)
+        return task
+
+    @mock.patch("documents.tasks.consume_file.delay")
+    @mock.patch("documents.tasks.run_subprocess")
+    def test_retry_consume_clean(self, m_subprocess, m_consume_file):
+        task = self.do_failed_task(self.SAMPLE_DIR / "corrupted.pdf")
+        m_subprocess.return_value.returncode = 0
+        task_id = tasks.retry_failed_file(task_id=task.task_id, clean=True)
+        self.assertIsNotNone(task_id)
+        m_consume_file.assert_called_once()
+
+    def test_cleanup(self):
+        task = self.do_failed_task(self.SAMPLE_DIR / "corrupted.pdf")
+        task.acknowledged = True
+        task.save()  # simulate the task being acknowledged
+        self.assertIsNotFile(settings.CONSUMPTION_FAILED_DIR / task.task_file_name)

src/documents/tests/utils.py

@@ -35,6 +35,7 @@ def setup_directories():
     dirs.scratch_dir = Path(tempfile.mkdtemp())
     dirs.media_dir = Path(tempfile.mkdtemp())
     dirs.consumption_dir = Path(tempfile.mkdtemp())
+    dirs.consumption_failed_dir = Path(tempfile.mkdtemp("failed"))
     dirs.static_dir = Path(tempfile.mkdtemp())
     dirs.index_dir = dirs.data_dir / "index"
     dirs.originals_dir = dirs.media_dir / "documents" / "originals"
@@ -56,6 +57,7 @@ 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,
@@ -72,6 +74,7 @@ def remove_dirs(dirs):
     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/documents/views.py

@@ -136,6 +136,7 @@ from documents.serialisers import DocumentListSerializer
 from documents.serialisers import DocumentSerializer
 from documents.serialisers import DocumentTypeSerializer
 from documents.serialisers import PostDocumentSerializer
+from documents.serialisers import RetryTaskSerializer
 from documents.serialisers import SavedViewSerializer
 from documents.serialisers import SearchResultSerializer
 from documents.serialisers import ShareLinkSerializer
@@ -152,6 +153,7 @@ from documents.serialisers import WorkflowTriggerSerializer
 from documents.signals import document_updated
 from documents.tasks import consume_file
 from documents.tasks import empty_trash
+from documents.tasks import retry_failed_file
 from documents.templating.filepath import validate_filepath_template_and_render
 from paperless import version
 from paperless.celery import app as celery_app
@@ -1718,6 +1720,25 @@ class TasksViewSet(ReadOnlyModelViewSet):
             queryset = PaperlessTask.objects.filter(task_id=task_id)
         return queryset
 
+    @action(methods=["post"], detail=True)
+    def retry(self, request, pk=None):
+        task = self.get_object()
+
+        serializer = RetryTaskSerializer(data=request.data)
+        serializer.is_valid(raise_exception=True)
+        clean = serializer.validated_data.get("clean")
+
+        try:
+            new_task_id = retry_failed_file(task.task_id, clean)
+            return Response({"task_id": new_task_id})
+        except FileNotFoundError:
+            return HttpResponseBadRequest("Original file not found")
+        except Exception as e:
+            logger.warning(f"An error occurred retrying task: {e!s}")
+            return HttpResponseBadRequest(
+                "Error retrying task, check logs for more detail.",
+            )
+
 
 class AcknowledgeTasksView(GenericAPIView):
     permission_classes = (IsAuthenticated,)

src/paperless/checks.py

@@ -65,6 +65,10 @@ def paths_check(app_configs, **kwargs):
         + 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.py

@@ -281,6 +281,11 @@ CONSUMPTION_DIR = __get_path(
     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(
     "PAPERLESS_SCRATCH_DIR",
@@ -890,6 +895,8 @@ CONSUMER_IGNORE_PATTERNS = list(
         ),
     ),
 )
+if CONSUMPTION_DIR in CONSUMPTION_FAILED_DIR.parents:
+    CONSUMER_IGNORE_PATTERNS.append(CONSUMPTION_FAILED_DIR.name)
 
 CONSUMER_SUBDIRS_AS_TAGS = __get_boolean("PAPERLESS_CONSUMER_SUBDIRS_AS_TAGS")
 

src/paperless/tests/test_checks.py

@@ -29,10 +29,11 @@ class TestChecks(DirectoriesMixin, TestCase):
         MEDIA_ROOT="uuh",
         DATA_DIR="whatever",
         CONSUMPTION_DIR="idontcare",
+        CONSUMPTION_FAILED_DIR="nope",
     )
     def test_paths_check_dont_exist(self):
         msgs = paths_check(None)
-        self.assertEqual(len(msgs), 3, str(msgs))
+        self.assertEqual(len(msgs), 4, str(msgs))
 
         for msg in msgs:
             self.assertTrue(msg.msg.endswith("is set but doesn't exist."))
@@ -41,13 +42,15 @@ class TestChecks(DirectoriesMixin, TestCase):
         os.chmod(self.dirs.data_dir, 0o000)
         os.chmod(self.dirs.media_dir, 0o000)
         os.chmod(self.dirs.consumption_dir, 0o000)
+        os.chmod(self.dirs.consumption_failed_dir, 0o000)
 
         self.addCleanup(os.chmod, self.dirs.data_dir, 0o777)
         self.addCleanup(os.chmod, self.dirs.media_dir, 0o777)
         self.addCleanup(os.chmod, self.dirs.consumption_dir, 0o777)
+        self.addCleanup(os.chmod, self.dirs.consumption_failed_dir, 0o777)
 
         msgs = paths_check(None)
-        self.assertEqual(len(msgs), 3)
+        self.assertEqual(len(msgs), 4)
 
         for msg in msgs:
             self.assertTrue(msg.msg.endswith("is not writeable"))