Add Python output modules

[daxmawal]

Summary

This PR is a follow-up to #1104 and adds reusable helpers for implementing Speech Dispatcher output modules in Python.
Disclosure: Codex assisted with code review and testing.

Follow-up work

Planned follow-up PRs will:

1. Python output module protocol implementation (module_main and module_process) and server-side audio transport.
2. add the Pocket TTS output module.
3. implement the remaining features, including sound icons, STOP, PAUSE, and complete custom debug logging.

[daxmawal]

I didn't find equivalent to strtol()

[sthibaul]

We can simply call int() and return -1 if an exception is caught

[daxmawal]

The Python tests have been successfully run and are available here.

[sthibaul]

Mmm, module_readline is already available in libspeechd_module, I'd say we can just use it instead of reimplementing it (and thus having to maintain both).

[sthibaul]

Note that this is different from module_strip_ssml, as module_readline is available under a BSD license while the existing implementation of module_strip_ssml is available under LGPL.

What I'm thinking is that the python module helper can leverage libspeechd_module to manage the protocol etc. and translate the eventual calls into python calls, plus some helpers to make things easier for module implementors.

[sthibaul]

this, however, we indeed want to have for module_process.c to call it

[sthibaul]

I don't think we want to mimic module_util's way of emitting debug information. Better use a pythonic way

[sthibaul]

Python output module protocol implementation (module_main and module_process)

As mentioned above, module_process can be taken from libspeechd_module. We can just make the python module expose the few C functions module_* functions to interact with it. That'll avoid having to maintain several implementations of the core protocol.