add caching to save time on CodeQL analysis

[ljubon]

After #392 this PR should save time on CodeQL run

Example run on forked repo: https://github.com/ljubon/ParquetSharp/actions/runs/7493616704/job/20399707365

[github-advanced-security]

This pull request sets up GitHub code scanning for this repository. Once the scans have completed and the checks have passed, the analysis results for this pull request branch will appear on this overview. Once you merge this pull request, the 'Security' tab will show more code scanning analysis results (for example, for the default branch). Depending on your configuration and choice of analysis tool, future pull requests will be annotated with code scanning analysis results. For more information about GitHub code scanning, check out the documentation.

[adamreeve]

Can you also please split up the "sudo apt-get update..." step, moving the "apt-get" commands before "InitializeCodeQL", and make sure all the steps have a name

[adamreeve]

I'm seeing CodeQL issues show up in the security tab for things that are in our C++ dependencies, which I don't think we should include here. Those will probably be excluded if we get prebuilt dependencies from the cache, but that won't work every time.

From the CodeQL docs (https://docs.github.com/en/code-security/code-scanning/creating-an-advanced-setup-for-code-scanning/customizing-your-advanced-setup-for-code-scanning#specifying-directories-to-scan), you can only ignore paths for interpreted languages, so we need to somehow split out the build of the C++ dependencies from the build of the ParquetSharp C++ library. Currently that all happens together in build_unix.sh when we run the CMake build. I don't know if there's an easy way to first build all the dependencies.

If this is too hard to fix it's probably fine to leave this as is.