mupdf: 1.23.0 -> 1.23.5; only build and install shared libraries

[lilyinstarlight]

Description of changes

Second attempt at #237199

Things done

• Built on platform(s)
  • x86_64-linux (only direct dependents of mupdf and tested that they function)
  • aarch64-linux (all reverse dependencies via nixpkgs-review and confirmed no regressions against master)
  • x86_64-darwin
  • aarch64-darwin
• For non-Linux: Is sandbox = true set in nix.conf? (See Nix manual)
• Tested, as applicable:
  • NixOS test(s) (look inside nixos/tests)
  • and/or package tests
  • or, for functions and "core" functionality, tests in lib/tests or pkgs/test
  • made sure NixOS tests are linked to the relevant packages
• Tested compilation of all packages that depend on this change using nix-shell -p nixpkgs-review --run "nixpkgs-review rev HEAD". Note: all changes have to be committed, also see nixpkgs-review usage
• Tested basic functionality of all binary files (usually in ./result/bin/)
• 23.11 Release Notes (or backporting 23.05 Release notes)
  • (Package updates) Added a release notes entry if the change is major or breaking
  • (Module updates) Added a release notes entry if the change is significant
  • (Module addition) Added a release notes entry if adding a new NixOS module
• Fits CONTRIBUTING.md.

[lilyinstarlight]

So, uh, it turns out pymupdf is broken on staging independent of this PR after it was bumped in the last python-updates cycle (my test builds were against these commits picked on top of master, so I didn't experience the breakage when running nixpkgs-review)

It's not really related to this PR, but I'll probably still push a fix in the next day or so (but it can go in a new PR since it is orthogonal to this one, mostly)

Edit: PyMuPDF is actually building from it's own vendored version right now anyway and can't use our MuPDF since it doesn't build C++ bindings...

[lilyinstarlight]

Okay so I've rebased this to fix a merge conflict, I added support for building the C++ and Python bindings to MuPDF and fixed PyMuPDF to use our own MuPDF build rather than building it's own

I think I've untangled as much MuPDF stuff as I can at this point and I added myself as a maintainer for the package

[happysalada]

Alright I've tried everything I could, so just writing my thoughts here in case someone has an idea.
pymupdf is broken on master because somehow when mupdf builds with ocr enabled, pkg-config is not able to find leptonica.
You can tell the compiler (USE_SYSTEM_LEPTONICA=no) to build with the thirdparty leptonica included, but somehow the build of pymupdf will fail with a circular import error. So I'm guessing that something else needs to be done for the build to actually work.
The best would be to have pkg-config detect leptonica, but I don't understand why it's not happening.
the build will conveniently fail quickly but the error message just points
https://github.com/ArtifexSoftware/mupdf/blob/b426d88c7f77ece0cca739af403d191066357b1f/Makethird#L57
and it's not immediately obvious that leptonica wasn't found.
I'm reaching out here as I'm guessing this is where someone more familiar with the build process than I am might be.
If everyone is busy, no worries of course.

[lilyinstarlight]

@happysalada That issue has nothing to do with pkg-config+leptonica (which works just fine if you actually test it) and has nothing to do with this PR. See #338680 (comment)