I gave it a shot on my fork.

To test it, first clone my branch and enter the repo:

$ git clone -b cmake https://github.com/wllm-rbnt/squashfs-tools/
$ cd squashfs-tools/squashfs-tools

Then, edit user defined variables & options in CMakeLists.txt (top section).

Then you can test it using Ninja:

$ sudo apt install ninja-build
$ cmake -B build -G Ninja
$ ninja -C build
$ ninja -C build install

or make

$ cmake -B build
$ cd build
$ make
$ make install

DATE, VERSION and YEAR have to be edited manually.
Only pre-built manpages installation will work for now.
Building with LZMA1 support is not supported.

William

Is there some specific reason why you think CMake would be a better fit for this project than say Meson other than personally being used to CMake? Both have similar deptrees in majority of distributions and Meson is quite trivial to use. That said, the Makefile is not that bad.

My main itch with it is that it uses various variables in not-very-standard way (eg not using DESTDIR as installation directory such that content is installed to $DESTDIR/$PREFIX and not using the common PREFIX but instead having INSTALL_PREFIX, INSTALL_DIR and INSTALL_MANPAGES_DIR) so doesn't work with typical common Makefile abstractions but every integrator has to figure out how the variables work together.

No specific reason on my side. It was just a request I could fulfill, and I had time to work on it.