PyMuPDF Pro 1.25.0: Image in .doc file unexpectedly overlaps with text when using get_pixmap()

[trianxy]

Description of the bug

Sometimes, an embedded image inside a .doc file overlaps with the text when creating an image of the document using get_pixmap(), although at other software (Google Docs, Mac Pages, libreoffice) this does not happen.

How to reproduce the bug

Uploading .doc files is not supported inside GitHub issues, so step 1 would be for you to tell me your email address, so that I can send you a test file image-unexpectedly-overlaps-with-text.doc.

Download attached image-unexpectedly-overlaps-with-text.doc.zip, unzip, and then:

After that, run:

from PIL import Image
import io, os
import pymupdf.pro
pymupdf.pro.unlock()  # add trial key if you want >3 pages

document = pymupdf.open("image-unexpectedly-overlaps-with-text.doc")
image_bytes = document.load_page(0).get_pixmap().tobytes(output="png")
img = Image.open(io.BytesIO(image_bytes))
img.save("tmp.png")

Now open tmp.png and observe that the image overlaps with the text, although this is not the case if I open the .doc file with e.g. Google Docs.

PyMuPDF version

1.25.0

Operating system

Linux

Python version

3.9

[JorjMcKie]

Please simply ZIP any files unsupported by GitHub. This will circumvent this restriction and can also be used for files that GitHub views as "executable", includeing .exe, .py and others.
But of course you are welcome to use my email address, too.
If files are not confidential attaching them (or their zipped version) here will however facilitate communication within our team.

[trianxy]

Please simply ZIP any files unsupported by GitHub

Thanks @JorjMcKie - I added a zip file above