Select sentence tokenizer language from TTS settings

[markbackman]

Summary

• Select the sentence tokenizer from the TTS language, with an optional text_aggregation_language override. Regional codes map to the corresponding Punkt model; unspecified and unsupported languages retain the English-model and Unicode-punctuation fallback.
• Capture the tokenizer language per generation so runtime updates apply to the next generation. Streaming audio contexts retain their own model, and explicitly associated RTVI observers use frame snapshots to handle delayed callbacks.
• Prepare models in background threads and document configuration, runtime updates, and observer association.

Stacked on #5726, targeting feat/bundled-punkt-data. Language selection uses the bundled models supplied by that PR, without runtime downloads or external NLTK data. This diff contains only the language-aware aggregation changes.

Testing

• Full Pyright run with the project virtual environment: 0 errors and 0 warnings.
• 424 tests passed covering bundled data, tokenization, TTS ordering and settings, interruptions, streaming contexts, and RTVI.
• Verified multilingual tokenization with external NLTK data disabled and downloads blocked.
• The unset-language type guard is preserved; Ruff and diff whitespace checks passed.

[codecov]

Codecov Report

❌ Patch coverage is 96.07843% with 4 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
src/pipecat/processors/frameworks/rtvi/observer.py	85.18%	4 Missing ⚠️

Files with missing lines	Coverage Δ
src/pipecat/services/tts_service.py	91.08% <100.00%> (+0.73%)	⬆️
...ipecat/utils/context/aggregated_frame_sequencer.py	98.82% <100.00%> (+<0.01%)	⬆️
src/pipecat/utils/prewarm.py	71.42% <100.00%> (ø)
src/pipecat/utils/string.py	90.83% <100.00%> (+0.65%)	⬆️
src/pipecat/utils/text/base_text_aggregator.py	82.22% <100.00%> (+3.27%)	⬆️
src/pipecat/utils/text/simple_text_aggregator.py	100.00% <100.00%> (ø)
src/pipecat/processors/frameworks/rtvi/observer.py	59.52% <85.18%> (+1.33%)	⬆️

... and 1 file with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.