A template for Using libadapta with Python

The scripts.

• ìnstall.sh to add the .deb files, pip install and python syntax symbols. Also it rebuilds and fixes a few issues if you rename or move the repo clone directory. gum is included to make nicer scripts.
• build.sh to build the .gz, .whl and install the package. This causes all python packages to be updated and makes ready for hatch publish.
• freeze.sh to make requirements.txt before a git add if more python dependencies become included in the project.
• pydo.sh to make a virtual environment launch icon. Just in case you don't have pipx or admin rights, and maybe need a .desktop file. Try ./pydo.sh adapta_test for example, and see cat ./adapta_test.desktop. Yes, icons are made after launch test. You might want to edit it a little. Something to do with GUI launch and
  setting the virtual environment not being automatic, or maybe you don't need a debug terminal. You also get an icon made. For other icons, don't use a command name, just make a .svg and pick a unique ID. It also makes a mimetype for the application icon. Put the application icon in ìcon_base before ./pydo.sh so the document mimetype icon is also made.
• lang.sh makes a localization file based on the user's language setting. Also updates the localizations if new things to translate are found.
• lazy-type.sh to monkey patch Neovim LazyVim for a _G.type new type definition function, which also return "object" and "class" as sub-types of "table".

Change the paths from xapp_adapta, as the name must be unique on PyPI, if you intend to upload a derivative work there. This is important.

The C++ module .so must be edited in xapp_adapta/cpp with so.in.cpp being edited to auto generate so.cpp. So, don't edit so.cpp. Add python interface definitions to so.pyi in xapp_adapta.

Thanks

Simon Jackson

Installing

1. git clone [email protected]:jackokring/mint-python-adapta.git
2. cd mint-python-adapta
3. ./install.sh (following the instructions for optional things)
4. If you've changed the name from xapp_adapta, ./build.sh will make a python module which can be uploaded to PyPI by hatch publish (check how in the hatch documentation), making pip install xapp-adapta work maybe?
5. A pip install would need a once àdapta_make_local for desktop Linux installs. Fancy icons and mimetypes

Interesting Things

The 3-way split of C++/Python/LuaJIT-5.1 can solve most problems. The C++ takes the role of ultimate fast without going rusty. The Python takes the role of glue and any ML via extensive modules from its community. Lua takes the role of a compact and reasonably fast interface to an alternate small scene due to a very compact language embedding, and documentation that you can read within a night.