enh: applying redactions with image cropping for currently unsupported colorspaces

[mahlzahn]

Description of the bug

Applying redactions with image cropping (images=pymupdf.PDF_REDACT_IMAGE_PIXELS) sometimes fail with the message only Gray, RGB, and CMYK colorspaces supported.

/usr/lib/python3.14/site-packages/pymupdf/__init__.py in ?(page, images, graphics, text)
  10766 
  10767         if redact_annots == []:  # any redactions on this page?
  10768             return False  # no redactions
  10769 
> 10770         rc = page._apply_redactions(text, images, graphics)  # call MuPDF
  10771         if not rc:  # should not happen really
  10772             raise ValueError("Error applying redactions.")
  10773 

/usr/lib/python3.14/site-packages/pymupdf/__init__.py in ?(self, text, images, graphics)
   9925         opts.black_boxes = 0  # no black boxes
   9926         opts.text = text  # how to treat text
   9927         opts.image_method = images  # how to treat images
   9928         opts.line_art = graphics  # how to treat vector graphics
-> 9929         success = mupdf.pdf_redact_page(page.doc(), page, opts)
   9930         return success

/usr/lib/python3.14/site-packages/mupdf/__init__.py in ?(doc, page, opts)
  57660 def pdf_redact_page(doc, page, opts):
  57661     r"""Class-aware wrapper for `::pdf_redact_page()`."""
> 57662     return _mupdf.pdf_redact_page(doc, page, opts)

FzErrorArgument: code=4: only Gray, RGB, and CMYK colorspaces supported

The error message is unexpected but clear and workarounds exist, e.g., redacting the image with total removal (images=pymupdf.PDF_REDACT_IMAGE_REMOVE). Thus, feel free to classify it as feature request or close it. Thanks.

How to reproduce the bug

1. Get sample from: https://github.com/ArtifexSoftware/tests/blob/master/pdf/foo0-ge-64k.pdf
2. Run following python code snippet:

document = pymupdf.Document("foo0-ge-64k.pdf")
page = document[0]
page.add_redact_annot(page.rect)
page.apply_redactions(images=pymupdf.PDF_REDACT_IMAGE_PIXELS)

PyMuPDF version

1.27.1

Operating system

Linux

Python version

3.14

[JorjMcKie]

Thanks for submitting this.
There is no way to make all possible image types in this world modifiable in the redaction process.
All we can potentially achieve is to either remove or ignore unsupported images.
I have submitted a respective suggestion to the MuPDF team here.
Please don't expect a prompt reaction ...