This library provides "asyncio REPL", a simple REPL that can be used even while your program is running, allowing you to inspect program state, create tasks, and await asynchronous functions.

This is inspired by Python's asyncio module when run via python -m asyncio.

The MicroPython REPL is unavailable while your program is running. This library runs a background REPL using the asyncio scheduler.

Furthermore, it is not possible to await at the main REPL because it does not know about the asyncio scheduler.

To use this library, you need to import the library and then start the REPL task.

For example, in main.py:

importasyncioimportaioreplasyncdefdemo(): awaitasyncio.sleep_ms(1000) print("async demo") state=20asyncdeftask1(): whilestate: #print("task 1")awaitasyncio.sleep_ms(500) print("done") asyncdefmain(): print("Starting tasks...") # Start other program tasks.t1=asyncio.create_task(task1()) # Start the aiorepl task.repl=asyncio.create_task(aiorepl.task()) awaitasyncio.gather(t1, repl) asyncio.run(main())

An optional globals dictionary can be passed to aiorepl.task(), which allows you to specify what will be in scope for the REPL. By default it uses the globals dictionary from the __main__ module, which is the same scope as the regular REPL (and main.py). In the example above, the REPL will be able to call the demo() function as well as get/set the state variable.

You can also provide your own dictionary, e.g. aiorepl.task({"obj": obj }), or use the globals dict from the current module, e.g. aiorepl.task(globals()). Note that you cannot use a class instance's members dictionary, e.g. aiorepl.task(obj.__dict__), as this is read-only in MicroPython.

Instead of the regular >>> prompt, the asyncio REPL will show --> .

--> 1+1 2 --> await demo() async demo --> state 20 --> import myapp.core --> state = await myapp.core.query_state() --> 1/0 ZeroDivisionError: divide by zero --> def foo(x): return x + 1 --> await asyncio.sleep(foo(3)) --> 

History is supported via the up/down arrow keys.

During command editing (the "R" phase), pressing Ctrl-C will cancel the current command and display a new prompt, like the regular REPL.

While a command is being executed, Ctrl-C will cancel the task that is executing the command. This will have no effect on blocking code (e.g. time.sleep()), but this should be rare in an asyncio-based program.

Ctrl-D at the asyncio REPL command prompt will terminate the current event loop, which will stop the running program and return to the regular REPL.

The following features are unsupported:

• Tab completion is not supported (also unsupported in python -m asyncio).
• Multi-line continuation. However you can do single-line definitions of functions, see demo above.
• Exception tracebacks. Only the exception type and message is shown, see demo above.
• Emacs shortcuts (e.g. Ctrl-A, Ctrl-E, to move to start/end of line).
• Unicode handling for input.