MLC Kernel Core

Features

• Persistent Python, Lua and JavaScript workers.
• Reused runtimes for repeated short executions.
• Stable long running execution through isolated worker queues.
• Dependency graph discovery for Python, Lua and JavaScript local modules.
• Absolute and relative path normalization.
• Windows and Unix path input support.
• Inline code execution support.

Running

cargo run

The demo starts:

1. Python file execution from python_app/main.py.
2. Lua file execution from lua_app/main.lua.
3. Inline Python execution.
4. Inline Lua execution.
5. Auto import libs

Path handling

The kernel accepts:

• python_app/main.py
• ./python_app/main.py
• python_app\\main.py
• absolute paths

Paths are normalized before loading and checked against the filesystem.

Worker model

Python and Lua use dedicated long-lived threads:

• startup cost is paid once;
• runtime state can be reused;
• task queues prevent unlimited memory growth;
• repeated executions avoid recreating interpreters.

Examples

Python:

kernel.run_python("examples/python_app/main.py")?;

Lua:

kernel.run_lua("examples/lua_app/main.lua")?;

JavaScript:

kernel.run_js("examples/js_app/main.js")?;

Language workers supports inline code JS worker run time is not full complete at this time.