should pypdf be merged with PyMuPDF?

[johnkw]

PyMuPDF is faster, with more features, and creates smaller output files. I'm not sure what advantage pypdf might have.

But these projects have a ridiculous amount of overlap.

See also pymupdf/PyMuPDF#4978

[stefan6419846]

Yes, they might have similar code, as they are both processing PDF files. We have some basic page about the differences inside our documentation: https://pypdf.readthedocs.io/en/stable/meta/comparisons.html

The main difference is that pypdf is written in pure Python, while PyMuPDF depends on native code. Additionally, PyMuPDF is mostly impossible to integrate into commercial applications unless you pay for a commercial license. pypdf with its BSD-3-Clause license is much more freely licensed.

PyMuPDF is faster, with more features, and creates smaller output files.

Native code tends to be faster, although pypdf most likely has further potential to be improved in this regard. Useful features directly related to PDF processing can always be added, as well as exploring other optimizations. Everyone is invited to actively look into issues and provide corresponding fixes and improvements to pypdf - this is how (mostly unpaid) FOSS development works in contrast to paid support contracts.

Apart from this, you are always free to choose the libraries fitting your use case the most of course.

[shartzog]

Goes to show most folks don't read the license ;)

[stefan6419846]

This is the discussion section for pypdf, not pymupdf, thus you might want to post this in their repository.