Using the stdlib.

[FerrisWasTaken]

I am trying to embed python in rust with RustPython.
I have the code

use rustpython_vm as vm;
use vm::{Settings, stdlib::get_module_inits};

fn main() -> vm::PyResult<()> {
    vm::Interpreter::with_init(Settings::default(), |vm| {vm.add_native_modules(get_module_inits())}).enter(|vm| {
        let code_obj = vm
            .compile(
                "import platform\nprint(platform.release())",
                vm::compiler::Mode::BlockExpr,
                "<embedded>".to_owned(),
            )
            .map_err(|err| vm.new_syntax_error(&err)).unwrap();

        vm.run_code_obj(code_obj, vm.new_scope_with_builtins())?;

        Ok(())
    })
}

This code works when the python does not import anything but it fails with this.
If I need to download the file, is there a way to embed the stdlib?

[Rafael-Conde]

I was currently researching a way to achieving the embedding of external libraries together with the embedding of RustPython into a Rust application. So far, apparently, one can use vm.add_frozen(rustpython_pylib::frozen_stdlib()) to add the standard library frozen, which I believe it means you'll be adding a library that can't be changed, like you would normally use it from a Python program. I think that you could also use the feature freeze-stdlib to achieve the same result with less code.

I think there is another way to freeze libraries which I read here, apparently there is a macro called rustpython_vm::py_freeze!(dir = "path/to/dir") that automagically adds the dependencies in a frozen way for you. to see the comment I mentioning, just access this page and search the page for py_freeze this took me straight to the comment.

I believe I'll try this methods soon and I can come back here to more feedback.