pw_console/docs.rst

.. _module-pw_console:

----------
pw_console
----------

The Pigweed Console provides a Python repl (read eval print loop) using
`ptpython`_ and a log message viewer in a single-window terminal based
interface. It is designed to be a replacement for `IPython's embed()`_ function.

.. warning::
   The Pigweed Console is under heavy development. A user manual and usage
   information will be documented as features near completion.

==========
Goals
==========

``pw_console`` is a complete solution for interacting with hardware devices
using :ref:`module-pw_rpc` over a :ref:`module-pw_hdlc` transport.

The repl allows interactive RPC sending while the log viewer provides immediate
feedback on device status.

- Interactive Python repl and log viewer in a single terminal window.

- Easily embeddable within a project's own custom console. This should allow
  users to define their own transport layer.

- Plugin framework to add custom status toolbars or window panes.

- Log viewer with searching and filtering.

- Daemon that provides a many-to-many mapping between consoles and devices.

=====
Usage
=====

``pw console`` is invoked by calling the ``embed()`` function in your own
Python script.

.. automodule:: pw_console.console_app
    :members: embed
    :undoc-members:
    :show-inheritance:

=========================
Implementation and Design
=========================

Detains on Pigweed Console internals follows.

Thread and Event Loop Design
----------------------------

In `ptpython`_ and `IPython`_ all user repl code is run in the foreground. This
allows interrupts like ``Ctrl-C`` and functions like ``print()`` and
``time.sleep()`` to work as expected. Pigweed's Console doesn't use this
approach as it would hide or freeze the `prompt_toolkit`_ user interface while
running repl code.

To get around this issue all user repl code is run in a dedicated thread with
stdout and stderr patched to capture output. This lets the user interface stay
responsive and new log messages to continue to be displayed.

Here's a diagram showing how ``pw_console`` threads and `asyncio`_ tasks are
organized.

.. mermaid::

   flowchart LR
       classDef eventLoop fill:#e3f2fd,stroke:#90caf9,stroke-width:1px;
       classDef thread fill:#fffde7,stroke:#ffeb3b,stroke-width:1px;
       classDef plugin fill:#fce4ec,stroke:#f06292,stroke-width:1px;
       classDef builtinFeature fill:#e0f2f1,stroke:#4db6ac,stroke-width:1px;

       %% Subgraphs are drawn in reverse order.

       subgraph pluginThread [Plugin Thread 1]
           subgraph pluginLoop [Plugin Event Loop 1]
               toolbarFunc-->|"Refresh<br/>UI Tokens"| toolbarFunc
               toolbarFunc[Toolbar Update Function]
           end
           class pluginLoop eventLoop;
       end
       class pluginThread thread;

       subgraph pluginThread2 [Plugin Thread 2]
           subgraph pluginLoop2 [Plugin Event Loop 2]
               paneFunc-->|"Refresh<br/>UI Tokens"| paneFunc
               paneFunc[Pane Update Function]
           end
           class pluginLoop2 eventLoop;
       end
       class pluginThread2 thread;

       subgraph replThread [Repl Thread]
           subgraph replLoop [Repl Event Loop]
               Task1 -->|Finished| Task2 -->|Cancel with Ctrl-C| Task3
           end
           class replLoop eventLoop;
       end
       class replThread thread;

       subgraph main [Main Thread]
           subgraph mainLoop [User Interface Event Loop]
               log[[Log Pane]]
               repl[[Python Repl]]
               pluginToolbar([User Toolbar Plugin])
               pluginPane([User Pane Plugin])
               class log,repl builtinFeature;
               class pluginToolbar,pluginPane plugin;
           end
           class mainLoop eventLoop;
       end
       class main thread;

       repl-.->|Run Code| replThread
       pluginToolbar-.->|Register Plugin| pluginThread
       pluginPane-.->|Register Plugin| pluginThread2

.. _IPython's embed(): https://ipython.readthedocs.io/en/stable/interactive/reference.html#embedding
.. _IPython: https://ipython.readthedocs.io/
.. _asyncio: https://docs.python.org/3/library/asyncio.html
.. _prompt_toolkit: https://python-prompt-toolkit.readthedocs.io/
.. _ptpython: https://github.com/prompt-toolkit/ptpython/