PDF redaction
Foldr can permanently redact regions of a PDF: the bytes are stripped from the file, not just visually covered. After a redaction, the underlying text and image data is gone: searching the redacted PDF for the redacted content returns nothing, copy-paste returns nothing, and a forensic tool inspecting the PDF object stream finds nothing. That’s the right behaviour for releasing documents publicly with sensitive data removed (names, addresses, account numbers, signatures).
This is different from the visual annotation tools in the file viewer, which draw on top of the document without modifying the underlying bytes. Annotation is appropriate for review and feedback; redaction is appropriate for compliance and release.
How to redact a PDF
- Open the PDF in the Foldr web app file viewer.
- Activate Redact mode from the viewer’s toolbar.
- Drag rectangles over each region you want to remove. You can place multiple rectangles per page and span multiple pages. They all apply in one operation.
- Review the marked regions. Anything covered by a rectangle will be permanently destroyed.
- Click Apply redactions. The viewer warns you that the action is destructive and not reversible. Confirm to proceed.
The redacted PDF replaces the original at the same path. There’s no copy / undo path: if you needed to keep the un-redacted version, save a copy of the file before redacting.
Where it runs
Redaction is processed server-side using mutool (from the MuPDF project). The viewer sends the rectangles to the Foldr server, which downloads the original PDF, runs the redaction operation, and writes the redacted output back to the original location.
Because redaction is server-side, the operation works the same whether the file lives on an SMB share, OneDrive, SharePoint, Google Drive, or any other Foldr-supported storage backend.
Licensing
PDF redaction is gated by the redact feature flag.
- Foldr appliance: licensed add-on.
- Foldr SaaS: included in Business and Enterprise tiers.
When the licence is present, the Redact tool appears in the file viewer toolbar for PDF documents. When it’s absent, the tool is hidden.
Permission model
Redaction respects the same per-share permissions as any other write operation. To redact a file you need write access to the share. The same access you’d need to overwrite the file by uploading a new version. Read-only users see the redaction tool disabled with a tooltip explaining why.
External users (Secure Link recipients) and public-link visitors cannot redact, regardless of the underlying licence. The file viewer disables the tool on external and public surfaces by design. Redaction is an internal-user operation.
What can and can’t be redacted
- ✅ PDFs with a real text layer (born-digital PDFs from Word, PowerPoint, Pages, etc.). Text removal is exact, character-level.
- ✅ PDFs that contain rasterised pages (scanned documents). Image regions inside the redaction rectangles are masked or removed.
- ✅ Mixed-content PDFs. Text and image streams handled appropriately within the same file.
- ❌ Non-PDF formats. Word documents, images, etc. are not redacted by this tool. Convert to PDF first if you need to redact a non-PDF.
- ⚠️ Text outside the visible page. Some PDFs include hidden or off-page text. Redaction removes content within the rectangles you draw on the visible page; it doesn’t sweep hidden content unless you explicitly mark it.
Verifying a redaction
After applying, search the redacted PDF for the content you removed (using your local PDF viewer’s Find or Foldr’s search index). Any match indicates content that wasn’t covered by a rectangle. Re-redact those regions.
For high-stakes documents, run a third-party tool that inspects the PDF object stream directly to confirm the bytes are gone. This catches any edge cases where the source PDF stored sensitive data in metadata or attachments rather than visible content.
See also
- Web app file viewer. The surface that hosts the redact tool
- Custom Fields. For tagging redacted documents (e.g. “released”) so they’re findable later