Enhancement: navbar upload button

Fix: retry a search index rebuild that was interrupted (#14379 )
An interrupted rebuild left an empty index stamped as current, so the next start reported it as up to date. Mark the rebuild as in progress and only clear the marker once it completes.
2026-10-13 19:47:13 +00:00 · 2026-10-13 12:00:05 -07:00 · 2026-10-07 08:54:36 -07:00 · 2026-10-06 15:13:08 +00:00 · 2026-10-06 08:11:44 -07:00
36 changed files with 296 additions and 145 deletions
@@ -80,7 +80,7 @@ jobs:
    needs: changes
    if: needs.changes.outputs.backend_changed == 'true'
    name: "Python ${{ matrix.python-version }}"
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    permissions:
      contents: read
    strategy:
@@ -114,7 +114,7 @@ jobs:
          packages: unpaper tesseract-ocr imagemagick ghostscript poppler-utils
      - name: Configure ImageMagick
        run: |
-          sudo cp docker/rootfs/etc/ImageMagick-6/paperless-policy.xml /etc/ImageMagick-6/policy.xml
+          sudo cp docker/rootfs/etc/ImageMagick-6/paperless-policy.xml /etc/ImageMagick-7/policy.xml
      - name: Install Python dependencies
        env:
          PYTHON_VERSION: ${{ steps.setup-python.outputs.python-version }}
@@ -158,7 +158,7 @@ jobs:
    needs: changes
    if: needs.changes.outputs.backend_changed == 'true'
    name: Check project typing
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    permissions:
      contents: read
    env:
@@ -24,10 +24,10 @@ jobs:
      fail-fast: false
      matrix:
        include:
-          - runner: ubuntu-24.04
+          - runner: ubuntu-26.04
            arch: amd64
            platform: linux/amd64
-          - runner: ubuntu-24.04-arm
+          - runner: ubuntu-26.04-arm
            arch: arm64
            platform: linux/arm64
    runs-on: ${{ matrix.runner }}
@@ -163,7 +163,7 @@ jobs:
          archive: false
  merge-and-push:
    name: Merge and Push Manifest
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    needs: build-arch
    if: needs.build-arch.outputs.should-push == 'true'
    environment: image-publishing
@@ -65,7 +65,7 @@ jobs:
    needs: changes
    if: needs.changes.outputs.docs_changed == 'true'
    name: Build Documentation
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    steps:
      - uses: actions/configure-pages@45bfe0192ca1faeb007ade9deae92b16b8254a0d # v6.0.0
      - name: Checkout
@@ -102,7 +102,7 @@ jobs:
    name: Deploy Documentation
    needs: [changes, build]
    if: github.event_name == 'push' && github.ref == 'refs/heads/main' && needs.changes.outputs.docs_changed == 'true'
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    permissions:
      pages: write
      id-token: write
@@ -72,7 +72,7 @@ jobs:
    needs: changes
    if: needs.changes.outputs.frontend_changed == 'true'
    name: Install Dependencies
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    permissions:
      contents: read
    steps:
@@ -104,7 +104,7 @@ jobs:
    name: Lint
    needs: [changes, install-dependencies]
    if: needs.changes.outputs.frontend_changed == 'true'
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    permissions:
      contents: read
    steps:
@@ -137,7 +137,7 @@ jobs:
    name: "Unit Tests (${{ matrix.shard-index }}/${{ matrix.shard-count }})"
    needs: [changes, install-dependencies]
    if: needs.changes.outputs.frontend_changed == 'true'
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    permissions:
      contents: read
    strategy:
@@ -188,10 +188,10 @@ jobs:
    name: E2E Tests
    needs: [changes, install-dependencies]
    if: needs.changes.outputs.frontend_changed == 'true'
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    permissions:
      contents: read
-    container: mcr.microsoft.com/playwright:v1.62.1-noble
+    container: mcr.microsoft.com/playwright:v1.62.1-resolute
    env:
      PLAYWRIGHT_BROWSERS_PATH: /ms-playwright
      PLAYWRIGHT_SKIP_BROWSER_DOWNLOAD: 1
@@ -246,7 +246,7 @@ jobs:
    name: Frontend Build
    needs: [changes, unit-tests, e2e-tests]
    if: needs.changes.outputs.frontend_changed == 'true'
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    permissions:
      contents: read
    steps:
@@ -14,7 +14,7 @@ permissions: {}
 jobs:
  wait-for-docker:
    name: Wait for Docker Build
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    permissions:
      checks: read
      statuses: read
@@ -30,7 +30,7 @@ jobs:
  build-release:
    name: Build Release
    needs: wait-for-docker
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    permissions:
      contents: read
    steps:
@@ -73,7 +73,7 @@ jobs:
        timeout-minutes: 12
        uses: $/.github/actions/apt-install
        with:
-          packages: gettext liblept5
+          packages: gettext libleptonica6
      # ---- Build Documentation ----
      - name: Build documentation
        env:
@@ -145,7 +145,7 @@ jobs:
  publish-release:
    name: Publish Release
    needs: build-release
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    permissions:
      contents: write
      pull-requests: write
@@ -197,7 +197,7 @@ jobs:
    name: Append Changelog
    needs: publish-release
    if: needs.publish-release.outputs.prerelease == 'false'
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    permissions:
      contents: write
      pull-requests: write
@@ -15,7 +15,7 @@ permissions:
 jobs:
  zizmor:
    name: Run zizmor
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    permissions:
      contents: read
      actions: read
@@ -29,7 +29,7 @@ jobs:
        uses: zizmorcore/zizmor-action@cc914d7f3750a2d13d75c7f184a1060aa0e9d482 # v0.6.4
  semgrep:
    name: Semgrep CE
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    container:
      image: semgrep/semgrep:1.155.0@sha256:cc869c685dcc0fe497c86258da9f205397d8108e56d21a86082ea4886e52784d
    if: github.actor != 'dependabot[bot]'
@@ -17,7 +17,7 @@ jobs:
  cleanup-images:
    name: Cleanup Image Tags for ${{ matrix.primary-name }}
    if: github.repository_owner == 'paperless-ngx'
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    environment: registry-maintenance
    strategy:
      fail-fast: false
@@ -42,7 +42,7 @@ jobs:
  cleanup-untagged-images:
    name: Cleanup Untagged Images Tags for ${{ matrix.primary-name }}
    if: github.repository_owner == 'paperless-ngx'
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    needs:
      - cleanup-images
    environment: registry-maintenance
@@ -21,7 +21,7 @@ on:
 jobs:
  analyze:
    name: Analyze
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    permissions:
      actions: read
      contents: read
@@ -13,7 +13,7 @@ jobs:
  synchronize-with-crowdin:
    name: Crowdin Sync
    if: github.repository_owner == 'paperless-ngx'
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    environment: translation-sync
    steps:
      - name: Checkout
@@ -7,7 +7,7 @@ jobs:
    # Note: peakoss/anti-slop does not support the `issues` event yet (all of its
    # issue inputs are still commented out upstream), so the checks that the PR Bot
    # workflow gets from the action are implemented manually here.
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-slim
    permissions:
      issues: write
    steps:
@@ -4,7 +4,7 @@ on:
    types: [opened]
 jobs:
  Anti-slop:
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-slim
    permissions:
      contents: read
      issues: read
@@ -24,7 +24,7 @@ jobs:
            ASLOP-PR-VERIFY
  pr-bot:
    name: Automated PR Bot
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-slim
    # Runs after Anti-slop so the welcome comment can see whether the PR was closed
    # instead of racing it. Still runs if that job fails, so labeling is not lost.
    needs: Anti-slop
@@ -12,7 +12,7 @@ permissions:
 jobs:
  pr_opened_or_reopened:
    name: pr_opened_or_reopened
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-slim
    permissions:
      # write permission is required for autolabeler
      pull-requests: write
@@ -9,7 +9,7 @@ jobs:
  stale:
    name: 'Stale'
    if: github.repository_owner == 'paperless-ngx'
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    permissions:
      issues: write
      pull-requests: write
@@ -34,7 +34,7 @@ jobs:
  lock-threads:
    name: 'Lock Old Threads'
    if: github.repository_owner == 'paperless-ngx'
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-26.04
    permissions:
      issues: write
      pull-requests: write
@@ -58,7 +58,7 @@ jobs:
  close-answered-discussions:
    name: 'Close Answered Discussions'
    if: github.repository_owner == 'paperless-ngx'
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-slim
    permissions:
      discussions: write
    steps:
@@ -117,7 +117,7 @@ jobs:
  close-outdated-discussions:
    name: 'Close Outdated Discussions'
    if: github.repository_owner == 'paperless-ngx'
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-slim
    permissions:
      discussions: write
    steps:
@@ -211,7 +211,7 @@ jobs:
  close-unsupported-feature-requests:
    name: 'Close Unsupported Feature Requests'
    if: github.repository_owner == 'paperless-ngx'
-    runs-on: ubuntu-24.04
+    runs-on: ubuntu-slim
    permissions:
      discussions: write
    steps:
@@ -8,7 +8,7 @@ env:
 jobs:
  generate-translate-strings:
    name: Generate Translation Strings
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-26.04
    environment: translation-sync
    permissions:
      contents: write
@@ -76,9 +76,7 @@ is not supported by any of the available parsers.

 **A:** Not by default. As of v3, a file whose contents match an existing document is still
 consumed, and the duplicate is flagged in the UI — open the document and check the
-**Duplicates** tab to review documents that share the same content, or filter the document
-list by **Duplicates** to find all of them (see
-[Duplicate documents](usage.md#duplicate-documents)). If you prefer the old
+**Duplicates** tab to review documents that share the same content. If you prefer the old
 behavior of rejecting duplicates during consumption, set
 [`PAPERLESS_CONSUMER_DELETE_DUPLICATES`](configuration.md#PAPERLESS_CONSUMER_DELETE_DUPLICATES)
 to `true`.
@@ -299,18 +299,19 @@ for details.
 ### Duplicate documents

 By default, Paperless-ngx **does not reject duplicates**. If you consume a file whose
-contents match an existing document (same original or archive checksum), the new copy is
-still consumed and a warning is logged.
+contents exactly match an existing document (same checksum), the new copy is still
+consumed and a warning is logged. The task entry for the upload also flags that a
+duplicate was detected and links to the existing document(s).

-When a document has duplicates, a **Duplicates** tab appears on its detail page, listing
-the other documents you can view that share the same content (including any in the trash).
-To find all documents with duplicates, choose **Duplicates** in the document list's text
-filter dropdown, or use `has_duplicates=true` in the REST API.
+To review duplicates, open a document and switch to the **Duplicates** tab on the
+document detail page. It lists other documents that share the same content, including any
+that are in the trash (shown with a badge), and links to each so you can decide which to
+keep.

 If you would rather reject duplicates at consumption time (the pre-v3 behavior), set
 [`PAPERLESS_CONSUMER_DELETE_DUPLICATES`](configuration.md#PAPERLESS_CONSUMER_DELETE_DUPLICATES)
 to `true`. The duplicate file is then deleted instead of consumed, and the task fails with
-a "Document already exists" message linking to the existing document.
+a "document already exists" message.

 ## Document Suggestions

@@ -1,6 +1,6 @@
 [project]
 name = "paperless-ngx"
-version = "3.3.0"
+version = "3.2.1"
 description = """\
  A community-supported supercharged document management system: scan, index and archive all your physical documents\
  """
@@ -1,6 +1,6 @@
 {
  "name": "paperless-ngx-ui",
-  "version": "3.3.0",
+  "version": "3.2.1",
  "scripts": {
    "preinstall": "npx only-allow pnpm",
    "ng": "ng",
@@ -29,8 +29,13 @@
  </div>
  <div class="search-container flex-grow-1 py-2 pb-3 pb-sm-2 me-sm-auto order-3 order-sm-1"
    [class.mobile-hidden]="mobileSearchHidden()">
-    <div class="col-12 header-search mx-auto">
-      <pngx-global-search></pngx-global-search>
+    <div class="col-12 header-search mx-auto d-flex align-items-stretch gap-3">
+      <pngx-global-search class="flex-grow-1"></pngx-global-search>
+      <button *pngxIfPermissions="{ action: PermissionAction.Add, type: PermissionType.Document }"
+        class="btn btn-sm header-upload border-0 d-flex align-items-center justify-content-center flex-shrink-0 py-0 px-2" type="button"
+        (click)="fileUpload.click()" aria-label="Upload documents" i18n-aria-label>
+        <i-bs class="d-flex" width="1.1em" height="1.1em" name="file-earmark-arrow-up"></i-bs><span class="ms-2 d-sm-none d-md-inline" i18n>Upload</span>
+      </button>
    </div>
  </div>
  <ul ngbNav class="order-sm-3">
@@ -70,6 +75,8 @@
  </ul>
 </nav>

+<input type="file" class="visually-hidden" (change)="onFileSelected($event)" multiple #fileUpload tabindex="-1" aria-hidden="true">
+
 <div class="container-fluid">
  <div class="row">
    <nav id="sidebarMenu" class="d-md-block bg-light sidebar collapse"
@@ -532,7 +532,23 @@ main {

 .header-search {
  width: 100%;
-  max-width: 55rem;
+  max-width: 50rem;
+}
+
+.header-search pngx-global-search {
+  min-width: 0;
+}
+
+// light w shadow to stand out against search
+.header-upload {
+  --bs-btn-bg: rgba(255, 255, 255, .15);
+  --bs-btn-hover-bg: rgba(255, 255, 255, .22);
+  --bs-btn-active-bg: rgba(255, 255, 255, .3);
+  color: var(--pngx-primary-text-contrast) !important; // .navbar .dropdown > .btn in styles.scss
+
+  &:not(:focus) {
+    box-shadow: 0 1px 3px rgba(0, 0, 0, .35), 0 1px 1px rgba(0, 0, 0, .2); // leave room for the .btn:focus ring
+  }
 }

@media screen and (max-width: 575.98px) {
@@ -662,3 +678,7 @@ main {
 ::ng-deep .navItemDrag .position-absolute i-bs {
  display: none;
 }
+
+.smaller {
+  font-size: 0.75rem;
+}
@@ -33,6 +33,7 @@ import { SearchService } from 'src/app/services/rest/search.service'
 import { SettingsService } from 'src/app/services/settings.service'
 import { TasksService } from 'src/app/services/tasks.service'
 import { ToastService } from 'src/app/services/toast.service'
+import { UploadDocumentsService } from 'src/app/services/upload-documents.service'
 import { environment } from 'src/environments/environment'
 import { ProfileEditDialogComponent } from '../common/profile-edit-dialog/profile-edit-dialog.component'
 import { DocumentDetailComponent } from '../document-detail/document-detail.component'
@@ -462,6 +463,29 @@ describe('AppFrameComponent', () => {
    expect(routerSpy).toHaveBeenCalled()
  })

+  it('should upload files selected from the navbar upload button', () => {
+    const uploadDocumentsService = TestBed.inject(UploadDocumentsService)
+    const uploadSpy = jest
+      .spyOn(uploadDocumentsService, 'uploadFiles')
+      .mockImplementation(() => {})
+    const toastSpy = jest.spyOn(toastService, 'showInfo')
+    const file = new File(['testing'], 'file.pdf')
+    const input = window.document.createElement('input')
+    jest.spyOn(input, 'files', 'get').mockReturnValue([file] as any)
+
+    component.onFileSelected({ target: input } as unknown as Event)
+    expect(toastSpy).toHaveBeenCalled()
+    expect(uploadSpy).toHaveBeenCalledWith([file])
+  })
+
+  it('should not upload when no files are selected from the navbar', () => {
+    const uploadDocumentsService = TestBed.inject(UploadDocumentsService)
+    const uploadSpy = jest.spyOn(uploadDocumentsService, 'uploadFiles')
+    const input = window.document.createElement('input')
+    component.onFileSelected({ target: input } as unknown as Event)
+    expect(uploadSpy).not.toHaveBeenCalled()
+  })
+
  it('should close all documents on logout', () => {
    const closeAllSpy = jest.spyOn(openDocumentsService, 'closeAll')
    component.onLogout()
@@ -48,12 +48,13 @@ import { SavedViewService } from 'src/app/services/rest/saved-view.service'
 import { SettingsService } from 'src/app/services/settings.service'
 import { TasksService } from 'src/app/services/tasks.service'
 import { ToastService } from 'src/app/services/toast.service'
+import { UploadDocumentsService } from 'src/app/services/upload-documents.service'
 import { environment } from 'src/environments/environment'
 import { ChatComponent } from '../chat/chat/chat.component'
+import { SwitchComponent } from '../common/input/switch/switch.component'
 import { BrandMarkComponent } from '../common/logo/brand-mark/brand-mark.component'
 import { LogoComponent } from '../common/logo/logo.component'
 import { ProfileEditDialogComponent } from '../common/profile-edit-dialog/profile-edit-dialog.component'
-import { SwitchComponent } from '../common/input/switch/switch.component'
 import { DocumentDetailComponent } from '../document-detail/document-detail.component'
 import { ComponentWithPermissions } from '../with-permissions/with-permissions.component'
 import { GlobalSearchComponent } from './global-search/global-search.component'
@@ -101,6 +102,7 @@ export class AppFrameComponent
  private modalService = inject(NgbModal)
  permissionsService = inject(PermissionsService)
  private djangoMessagesService = inject(DjangoMessagesService)
+  private uploadDocumentsService = inject(UploadDocumentsService)

  readonly appRemoteVersion = signal<AppRemoteVersion>(null)
  readonly isMenuCollapsed = signal(true)
@@ -511,6 +513,15 @@ export class AppFrameComponent
    }
  }

+  onFileSelected(event: Event) {
+    const input = event.target as HTMLInputElement
+    const files = Array.from(input.files ?? [])
+    input.value = ''
+    if (!files.length) return
+    this.toastService.showInfo($localize`Initiating upload...`, 3000)
+    this.uploadDocumentsService.uploadFiles(files)
+  }
+
  onLogout() {
    this.openDocumentsService.closeAll()
  }
@@ -85,15 +85,11 @@ describe('UploadFileWidgetComponent', () => {
      'file.pdf'
    )
    const fileInput = fixture.debugElement.query(By.css('input'))
-    jest.spyOn(fileInput.nativeElement, 'files', 'get').mockReturnValue({
-      item: () => file,
-      length: 1,
-      [Symbol.iterator]: () => ({
-        next: () => ({ done: false, value: file }),
-      }),
-    } as any)
+    jest
+      .spyOn(fileInput.nativeElement, 'files', 'get')
+      .mockReturnValue([file] as any)
    fileInput.nativeElement.dispatchEvent(new Event('change'))
-    expect(uploadSpy).toHaveBeenCalled()
+    expect(uploadSpy).toHaveBeenCalledWith(file)
  })

  it('should generate stats summary', () => {
@@ -130,11 +130,9 @@ export class UploadFileWidgetComponent extends ComponentWithPermissions {
  }

  public onFileSelected(event: Event) {
-    const files = (event.target as HTMLInputElement).files
-    for (let i = 0; i < files?.length; i++) {
-      const file = files.item(i)
-      file && this.uploadDocumentsService.uploadFile(file)
-    }
+    const input = event.target as HTMLInputElement
+    this.uploadDocumentsService.uploadFiles(Array.from(input.files ?? []))
+    input.value = ''
  }

  get slimSidebarEnabled(): boolean {
@@ -135,7 +135,7 @@ export class FileDropComponent {
        .then((results) => {
          files.push(...[].concat(...results))
          this.toastService.showInfo($localize`Initiating upload...`, 3000)
-          files.forEach((file) => this.uploadDocumentsService.uploadFile(file))
+          this.uploadDocumentsService.uploadFiles(files)
        })
        .catch((e) => {
          this.toastService.showError(
@@ -144,7 +144,7 @@ export class FileDropComponent {
        })
    } else if (files.length) {
      this.toastService.showInfo($localize`Initiating upload...`, 3000)
-      files.forEach((file) => this.uploadDocumentsService.uploadFile(file))
+      this.uploadDocumentsService.uploadFiles(files)
    }

    this.onDragLeave(event, true)
@@ -54,6 +54,18 @@ describe('UploadDocumentsService', () => {
    req[0].flush('123-456')
  })

+  it('uploads each file when given multiple files', () => {
+    const uploadSpy = jest.spyOn(uploadDocumentsService, 'uploadFile')
+    const file1 = new File(['one'], 'one.pdf')
+    const file2 = new File(['two'], 'two.pdf')
+    uploadDocumentsService.uploadFiles([file1, file2])
+    expect(uploadSpy).toHaveBeenCalledWith(file1)
+    expect(uploadSpy).toHaveBeenCalledWith(file2)
+    httpTestingController
+      .match(`${environment.apiBaseUrl}documents/post_document/`)
+      .forEach((req) => req.flush('123-456'))
+  })
+
  it('updates progress during upload and failure', () => {
    const file = new File(
      [new Blob(['testing'], { type: 'application/pdf' })],
@@ -16,6 +16,12 @@ export class UploadDocumentsService {

  private uploadSubscriptions: Array<Subscription> = []

+  public uploadFiles(files: Iterable<File>) {
+    for (const file of files) {
+      this.uploadFile(file)
+    }
+  }
+
  public uploadFile(file: File) {
    let formData = new FormData()
    formData.append('document', file, file.name)
@@ -8,7 +8,7 @@ export const environment = {
  apiVersion: '10', // match src/paperless/settings.py
  appTitle: DEFAULT_APP_TITLE,
  tag: 'prod',
-  version: '3.3.0',
+  version: '3.2.1',
  webSocketHost: window.location.host,
  webSocketProtocol: window.location.protocol == 'https:' ? 'wss:' : 'ws:',
  webSocketBaseUrl: base_url.pathname + 'ws/',
@@ -90,6 +90,7 @@ import {
  exclamationTriangleFill,
  eye,
  fileEarmark,
+  fileEarmarkArrowUp,
  fileEarmarkCheck,
  fileEarmarkDiff,
  fileEarmarkFill,
@@ -341,6 +342,7 @@ const icons = {
  exclamationTriangleFill,
  eye,
  fileEarmark,
+  fileEarmarkArrowUp,
  fileEarmarkCheck,
  fileEarmarkDiff,
  fileEarmarkFill,
@@ -31,6 +31,7 @@ from documents.search._query import parse_user_query
 from documents.search._schema import _write_sentinels
 from documents.search._schema import build_schema
 from documents.search._schema import open_or_rebuild_index
+from documents.search._schema import rebuild_in_progress
 from documents.search._schema import wipe_index
 from documents.search._tokenizer import ascii_fold
 from documents.search._tokenizer import autocomplete_tokens
@@ -1110,39 +1111,44 @@ class TantivyBackend:
                flushing a segment, deferring merge work; they do not avoid it.
        """
        wipe_index(self._path)
-        new_index = tantivy.Index(build_schema(), path=str(self._path))
-        _write_sentinels(self._path)
-        register_tokenizers(new_index, settings.SEARCH_LANGUAGE)
+        # The marker covers the window where the empty index is already stamped
+        # as current but not yet populated, so an interrupted rebuild is retried.
+        with rebuild_in_progress(self._path):
+            new_index = tantivy.Index(build_schema(), path=str(self._path))
+            _write_sentinels(self._path)
+            register_tokenizers(new_index, settings.SEARCH_LANGUAGE)

-        # Point instance at the new index so _build_tantivy_doc uses it
-        old_index, old_schema = self._raw_index, self._raw_schema
-        self._raw_index = new_index
-        self._raw_schema = new_index.schema
-        # Stream documents one-by-one (so the progress bar advances per
-        # document) while fetching viewer permissions one SQL query per chunk.
-        # The stream is Sized, so iter_wrapper can still discover the total.
-        documents_stream = _DocumentViewerStream(documents, chunk_size=1000)
-        try:
-            writer = new_index.writer(heap_size=writer_heap_bytes)
-            for document, (viewer_ids, viewer_group_ids) in iter_wrapper(
-                documents_stream,
-            ):
-                doc = self._build_tantivy_doc(
-                    document,
-                    viewer_ids=viewer_ids,
-                    viewer_group_ids=viewer_group_ids,
-                )
-                writer.add_document(doc)
-            writer.commit()
-            # Wait for background merge threads to finish so all segments are
-            # fully merged and persisted before the index is considered rebuilt.
-            writer.wait_merging_threads()
-            new_index.reload()
-        except BaseException:  # pragma: no cover
-            # Restore old index on failure so the backend remains usable
-            self._raw_index = old_index
-            self._raw_schema = old_schema
-            raise
+            # Point instance at the new index so _build_tantivy_doc uses it
+            old_index, old_schema = self._raw_index, self._raw_schema
+            self._raw_index = new_index
+            self._raw_schema = new_index.schema
+            # Stream documents one-by-one (so the progress bar advances per
+            # document) while fetching viewer permissions one SQL query per
+            # chunk. The stream is Sized, so iter_wrapper can still discover
+            # the total.
+            documents_stream = _DocumentViewerStream(documents, chunk_size=1000)
+            try:
+                writer = new_index.writer(heap_size=writer_heap_bytes)
+                for document, (viewer_ids, viewer_group_ids) in iter_wrapper(
+                    documents_stream,
+                ):
+                    doc = self._build_tantivy_doc(
+                        document,
+                        viewer_ids=viewer_ids,
+                        viewer_group_ids=viewer_group_ids,
+                    )
+                    writer.add_document(doc)
+                writer.commit()
+                # Wait for background merge threads to finish so all segments
+                # are fully merged and persisted before the index is considered
+                # rebuilt.
+                writer.wait_merging_threads()
+                new_index.reload()
+            except BaseException:  # pragma: no cover
+                # Restore old index on failure so the backend remains usable
+                self._raw_index = old_index
+                self._raw_schema = old_schema
+                raise


 def chunked(iterable, size):
@@ -4,6 +4,7 @@ import hashlib
 import json
 import logging
 import shutil
+from contextlib import contextmanager
 from typing import TYPE_CHECKING
 from typing import Final
 from typing import NamedTuple
@@ -16,6 +17,7 @@ from whoosh_compat import FieldKind
 from documents.search._fields import PUBLIC_FIELDS

 if TYPE_CHECKING:
+    from collections.abc import Iterator
    from pathlib import Path

 logger = logging.getLogger("paperless.search")
@@ -28,6 +30,11 @@ logger = logging.getLogger("paperless.search")
 # v3 - barcodes JSON field for stored barcode contents
 SCHEMA_VERSION: Final[int] = 3

+# Present in the index directory from the moment a full rebuild starts until it
+# finishes. If a rebuild is interrupted it is left behind, so the half-built
+# index is not mistaken for a complete one.
+REBUILD_MARKER: Final[str] = ".rebuilding"
+

 class FieldDescriptor(NamedTuple):
    """One tantivy field, in declaration order.
@@ -255,9 +262,9 @@ def needs_rebuild(index_dir: Path) -> bool:
    """
    Check if the search index needs rebuilding.

-    Reads .index_settings.json to compare the stored schema version, search
-    language and schema fingerprint against the current configuration. Returns
-    True if the file is missing, unparsable, or any value mismatches.
+    True if a previous full rebuild never finished (the rebuild marker is still
+    present), or if the index's stamped settings no longer match the current
+    configuration. See _settings_mismatch().

    Args:
        index_dir: Path to the search index directory
@@ -265,6 +272,40 @@ def needs_rebuild(index_dir: Path) -> bool:
    Returns:
        True if the index needs rebuilding, False if it's up to date
    """
+    if (index_dir / REBUILD_MARKER).exists():
+        logger.warning("Previous search index rebuild did not finish - rebuilding.")
+        return True
+    return _settings_mismatch(index_dir)
+
+
+@contextmanager
+def rebuild_in_progress(index_dir: Path) -> Iterator[None]:
+    """
+    Flag the index as incomplete for the duration of a full rebuild.
+
+    The marker is cleared only if the block exits cleanly. There is deliberately
+    no try/finally: an exception must leave the marker behind so the next
+    needs_rebuild() check retries the rebuild.
+    """
+    marker = index_dir / REBUILD_MARKER
+    marker.touch()
+    yield
+    marker.unlink(missing_ok=True)
+
+
+def _settings_mismatch(index_dir: Path) -> bool:
+    """
+    Check the stamped settings against the current configuration.
+
+    Reads .index_settings.json to compare the stored schema version, search
+    language and schema fingerprint. Returns True if the file is missing,
+    unparsable, or any value mismatches.
+
+    This deliberately ignores the rebuild marker: open_or_rebuild_index() uses it
+    so that a process opening the index while another process is mid-rebuild
+    (or after one died) does not wipe the partial index out from under it.
+    Repopulating is the job of ``document_index reindex``.
+    """
    settings_file = index_dir / ".index_settings.json"
    if not settings_file.exists():
        return True
@@ -333,7 +374,7 @@ def open_or_rebuild_index(index_dir: Path | None = None) -> tantivy.Index:
        index_dir = cast("Path", settings.INDEX_DIR)
    if not index_dir.exists():
        return tantivy.Index(build_schema())
-    if needs_rebuild(index_dir):
+    if _settings_mismatch(index_dir):
        wipe_index(index_dir)
        idx = tantivy.Index(build_schema(), path=str(index_dir))
        _write_sentinels(index_dir)
@@ -16,7 +16,10 @@ from documents.search._backend import TantivyBackend
 from documents.search._backend import WriteBatch
 from documents.search._backend import get_backend
 from documents.search._backend import reset_backend
+from documents.search._schema import REBUILD_MARKER
+from documents.search._schema import needs_rebuild
 from documents.signals.handlers import add_to_index
+from paperless_testing.dirs import PaperlessDirs
 from paperless_testing.factories import CorrespondentFactory
 from paperless_testing.factories import DocumentFactory
 from paperless_testing.factories import DocumentTypeFactory
@@ -823,6 +826,53 @@ class TestRebuild:
        backend.rebuild(Document.objects.all(), iter_wrapper=wrapper)
        assert 30 in seen

+    def test_successful_rebuild_leaves_index_up_to_date(
+        self,
+        backend: TantivyBackend,
+        paperless_dirs: PaperlessDirs,
+    ) -> None:
+        """
+        GIVEN:
+            - A backend and one document
+        WHEN:
+            - rebuild() completes
+        THEN:
+            - needs_rebuild() is False and no rebuild marker remains
+        """
+        DocumentFactory.create()
+
+        backend.rebuild(Document.objects.all())
+
+        assert needs_rebuild(paperless_dirs.index_dir) is False
+        assert not (paperless_dirs.index_dir / REBUILD_MARKER).exists()
+
+    def test_interrupted_rebuild_is_retried(
+        self,
+        backend: TantivyBackend,
+        paperless_dirs: PaperlessDirs,
+    ) -> None:
+        """
+        GIVEN:
+            - A rebuild that dies while indexing documents (e.g. the database
+              connection is lost)
+        WHEN:
+            - needs_rebuild() is checked afterwards
+        THEN:
+            - It is True, even though the empty index was already stamped with
+              current settings, so the next start rebuilds instead of reporting
+              the index as up to date
+        """
+        DocumentFactory.create()
+
+        def die(pairs):
+            raise RuntimeError("terminating connection due to administrator command")
+            yield  # pragma: no cover
+
+        with pytest.raises(RuntimeError):
+            backend.rebuild(Document.objects.all(), iter_wrapper=die)
+
+        assert needs_rebuild(paperless_dirs.index_dir) is True
+
    def test_includes_group_granted_viewers(self, backend: TantivyBackend) -> None:
        """Rebuild must index viewer ids for group-only grants, not just direct ones.

@@ -2,7 +2,7 @@ msgid ""
 msgstr ""
 "Project-Id-Version: paperless-ngx\n"
 "Report-Msgid-Bugs-To: \n"
-"POT-Creation-Date: 2026-10-05 16:26+0000\n"
+"POT-Creation-Date: 2026-10-06 15:12+0000\n"
 "PO-Revision-Date: 2022-02-17 04:17\n"
 "Last-Translator: \n"
 "Language-Team: English\n"
@@ -1941,25 +1941,8 @@ msgstr ""
 msgid "As a final step, please complete the following form:"
 msgstr ""

-#: documents/validators.py:24
-#, python-brace-format
-msgid "Unable to parse URI {value}, missing scheme"
-msgstr ""
-
-#: documents/validators.py:29
-#, python-brace-format
-msgid "Unable to parse URI {value}, missing net location or path"
-msgstr ""
-
 #: documents/validators.py:36
-msgid ""
-"URI scheme '{parts.scheme}' is not allowed. Allowed schemes: {', '."
-"join(allowed_schemes)}"
-msgstr ""
-
-#: documents/validators.py:45
-#, python-brace-format
-msgid "Unable to parse URI {value}"
+msgid ", "
 msgstr ""

 #: documents/views.py:336 documents/views.py:2729
@@ -137,9 +137,20 @@ class TestNginxService:
    reason="No Gotenberg/Tika servers to test with",
 )
 class TestParserLive:
-    @staticmethod
-    def imagehash(file: Path, hash_size: int = 18) -> str:
-        return f"{average_hash(Image.open(file), hash_size)}"
+    # Rasterizer versions shift a few pixels, so compare perceptual hashes by
+    # Hamming distance (out of 18 * 18 = 324 bits) rather than for equality
+    MAX_HASH_DISTANCE = 8
+
+    @classmethod
+    def assert_thumbnails_similar(cls, generated: Path, expected: Path) -> None:
+        distance = average_hash(Image.open(generated), 18) - average_hash(
+            Image.open(expected),
+            18,
+        )
+        assert distance <= cls.MAX_HASH_DISTANCE, (
+            f"Thumbnail {generated} differs from {expected} by {distance} bits "
+            f"(max {cls.MAX_HASH_DISTANCE})"
+        )

    def test_get_thumbnail(
        self,
@@ -168,12 +179,7 @@ class TestParserLive:
        assert thumb.exists()
        assert thumb.is_file()

-        assert self.imagehash(thumb) == self.imagehash(
-            simple_txt_email_thumbnail_file,
-        ), (
-            f"Created thumbnail {thumb} differs from expected file "
-            f"{simple_txt_email_thumbnail_file}"
-        )
+        self.assert_thumbnails_similar(thumb, simple_txt_email_thumbnail_file)

    def test_tika_parse_successful(self, mail_parser: MailDocumentParser) -> None:
        """
@@ -255,7 +261,7 @@ class TestParserLive:
        THEN:
            - Gotenberg shall be called to generate the PDF
            - The archive PDF shall contain the expected content
-            - The generated thumbnail shall match the expected image hash
+            - The generated thumbnail shall be perceptually close to the expected image
        """
        util_call_with_backoff(mail_parser.parse, [html_email_file, "message/rfc822"])

@@ -272,14 +278,4 @@ class TestParserLive:
            html_email_file,
            "message/rfc822",
        )
-        generated_thumbnail_hash = self.imagehash(generated_thumbnail)
-
-        # The created PDF is not reproducible, but the converted image
-        # should always look the same
-        expected_hash = self.imagehash(html_email_thumbnail_file)
-
-        assert generated_thumbnail_hash == expected_hash, (
-            f"PDF thumbnail differs from expected. "
-            f"Generated: {generated_thumbnail}, "
-            f"Hash: {generated_thumbnail_hash} vs {expected_hash}"
-        )
+        self.assert_thumbnails_similar(generated_thumbnail, html_email_thumbnail_file)
@@ -1,6 +1,6 @@
 from typing import Final

-__version__: Final[tuple[int, int, int]] = (3, 3, 0)
+__version__: Final[tuple[int, int, int]] = (3, 2, 1)
 # Version string like X.Y.Z
 __full_version_str__: Final[str] = ".".join(map(str, __version__))
 # Version string like X.Y
@@ -2971,7 +2971,7 @@ wheels = [

 [[package]]
 name = "paperless-ngx"
-version = "3.3.0"
+version = "3.2.1"
 source = { virtual = "." }
 dependencies = [
    { name = "azure-ai-documentintelligence" },
Author	SHA1	Message	Date
shamoon	05321b10d3	Enhancement: navbar upload button	2026-10-13 12:00:05 -07:00
Trenton H	d7a9894400	Fix: retry a search index rebuild that was interrupted (#14379 ) An interrupted rebuild left an empty index stamped as current, so the next start reported it as up to date. Mark the rebuild as in progress and only clear the marker once it completes.	2026-10-07 08:54:36 -07:00
GitHub Actions	ee34a6598e	Auto translate strings	2026-10-06 15:13:08 +00:00
Trenton H	3a3b3ef66a	Chore: Upgrade runners to Ubuntu 26.04 (#14347 ) * Moves runners to 26.04 and a few jobs to -slim variant * Probably fixing the imagemagik 7 problems and maybe the frontend playwright thing? * Compare thumbnails, but allow a little difference in the perceptual hash	2026-10-06 08:11:44 -07:00