feat(speech_recognition): add faster-whisper engine backend

[jatinkrmalik]

Summary

This PR adds faster-whisper as a new speech recognition engine for Vocalinux, delivering significantly faster CPU inference via CTranslate2 and INT8 quantization. It also lays the groundwork for future engine additions by introducing a minimal Engine Protocol and an engines/ registry.

https://huggingface.co/collections/Systran/faster-whisper

What changed

Phase 0 — engine abstraction

• Added EngineType enum and Engine Protocol to src/vocalinux/common_types.py.
• Added src/vocalinux/speech_recognition/engines/__init__.py as a registry for engine implementations.

Phase 1 — faster-whisper backend

• New src/vocalinux/speech_recognition/engines/faster_whisper_engine.py implementing the Engine Protocol.
• New src/vocalinux/utils/faster_whisper_model_info.py for model metadata, download-cache detection, and recommendations.
• Wired the engine into SpeechRecognitionManager (init, transcribe, reconfigure, reinitialize-after-resume).
• Added "faster_whisper" to CLI --engine choices.
• Added engine/model entries to the Settings UI and ConfigManager defaults.
• Added faster-whisper>=1.0.0 as an optional dependency ([faster_whisper]).
• Added vocalinux.speech_recognition.engines to setuptools packages.
• Added tests/test_faster_whisper_engine.py with 17 unit tests.
• Updated tests/test_suspend_handler.py fixture for the new _faster_whisper_engine attribute.

How to try it

pip install "vocalinux[faster_whisper]"
python -m vocalinux.main --engine faster_whisper --model tiny

Verification

• pytest tests/test_faster_whisper_engine.py — 17/17 passed.
• Affected regression tests (config, settings, recognition manager init, suspend handler) — 179/179 passed.
• make lint (black, isort, flake8) passed via pre-commit hooks.
• Note: a handful of pre-existing IBus environment tests fail locally and are unrelated to this change.

[codecov]

Codecov Report

❌ Patch coverage is 82.87037% with 37 lines in your changes missing coverage. Please review.
✅ Project coverage is 84.76%. Comparing base (3d65e5c) to head (0184b8e).
⚠️ Report is 3 commits behind head on main.

Files with missing lines	Patch %	Lines
...ocalinux/speech_recognition/recognition_manager.py	30.43%	26 Missing and 6 partials ⚠️
src/vocalinux/common_types.py	73.33%	4 Missing ⚠️
...peech_recognition/engines/faster_whisper_engine.py	98.55%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #543      +/-   ##
==========================================
- Coverage   84.83%   84.76%   -0.07%     
==========================================
  Files          33       36       +3     
  Lines        5861     6077     +216     
  Branches      951      979      +28     
==========================================
+ Hits         4972     5151     +179     
- Misses        688      719      +31     
- Partials      201      207       +6     

Files with missing lines	Coverage Δ
src/vocalinux/main.py	91.47% <ø> (ø)
...c/vocalinux/speech_recognition/engines/__init__.py	100.00% <100.00%> (ø)
src/vocalinux/ui/config_manager.py	98.70% <ø> (ø)
src/vocalinux/utils/faster_whisper_model_info.py	100.00% <100.00%> (ø)
...peech_recognition/engines/faster_whisper_engine.py	98.55% <98.55%> (ø)
src/vocalinux/common_types.py	74.28% <73.33%> (-0.72%)	⬇️
...ocalinux/speech_recognition/recognition_manager.py	81.63% <30.43%> (-1.43%)	⬇️

... and 1 file with indirect coverage changes

🚀 New features to boost your workflow:

• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.