search

adamsol / Pyxell

Multi-paradigm programming language compiled to C++, written in Python.
MIT License
54 stars 6 forks source link

Package it onto the PyPi #12

Open ThatXliner opened 3 years ago

ThatXliner commented 3 years ago

Package it with a setup.py and everything so we can install it on PyPi

adamsol commented 3 years ago

I've already had that idea, but failed at creating a package with all the necessary files. It's easy for .py source code, but Pyxell needs also some other files, and for some reason I couldn't make it work.

On second thought, I'm not sure whether PyPI is a good place for a programming language. Compilers are usually installed in some other ways. Also, it's only an implementation detail that Pyxell is written in Python. I'm already thinking about rewriting it to C++ to make parsing faster and make it possible to implement an interpreter in a fairly easy way.

ThatXliner commented 3 years ago

I've already had that idea, but failed at creating a package with all the necessary files. It's easy for .py source code, but Pyxell needs also some other files, and for some reason I couldn't make it work.

Have you seen the MANIFEST.in file? That's where you can include other files, too.

On second thought, I'm not sure whether PyPI is a good place for a programming language.

Pip is pre-installed on almost every system. I almost cry when packages say "get this from homebrew" because my HomeBrew barely works. Pip is the easiest way. Or conda.

adamsol commented 3 years ago

Have you seen the MANIFEST.in file? That' where you can include other files, too.

Yeah, I tried it, and something still didn't work. It seems to be a strange problem for many people, see e.g. https://stackoverflow.com/questions/1612733/including-non-python-files-with-setup-py. I just gave up after trying several approaches.

Pip is pre-installed on almost every system. I almost cry when packages say "get this from homebrew" because my HomeBrew barely works. Pip is the easiest way. Or conda.

Maybe I'll have better luck with conda. It also fits better, since it's language-agnostic. However, Pyxell can be easily downloaded from Github releases and run from source. It's the same way I've recently downloaded Kotlin.