TTS services improvements.

[filipi87]

TTS services improvements.

• Refactoring TTS services to handle audio context.
• Adding a parameter to TTSSpeakFrame to conditionally add it to the context.

[codecov]

Codecov Report

❌ Patch coverage is 13.54582% with 217 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
src/pipecat/services/resembleai/tts.py	0.00%	35 Missing ⚠️
src/pipecat/services/playht/tts.py	0.00%	17 Missing ⚠️
src/pipecat/services/inworld/tts.py	0.00%	14 Missing ⚠️
src/pipecat/services/azure/tts.py	0.00%	13 Missing ⚠️
src/pipecat/services/cartesia/tts.py	0.00%	13 Missing ⚠️
src/pipecat/services/rime/tts.py	0.00%	13 Missing ⚠️
src/pipecat/services/elevenlabs/tts.py	0.00%	12 Missing ⚠️
src/pipecat/services/tts_service.py	60.71%	11 Missing ⚠️
src/pipecat/services/google/tts.py	25.00%	9 Missing ⚠️
src/pipecat/services/asyncai/tts.py	0.00%	8 Missing ⚠️
... and 19 more

Files with missing lines	Coverage Δ
src/pipecat/frames/frames.py	88.83% <100.00%> (+0.01%)	⬆️
src/pipecat/processors/aggregators/llm_response.py	73.80% <100.00%> (ø)
...t/processors/aggregators/llm_response_universal.py	77.57% <66.66%> (-0.01%)	⬇️
src/pipecat/services/piper/tts.py	50.00% <66.66%> (ø)
src/pipecat/services/camb/tts.py	0.00% <0.00%> (ø)
src/pipecat/services/fish/tts.py	0.00% <0.00%> (ø)
src/pipecat/services/hathora/tts.py	0.00% <0.00%> (ø)
src/pipecat/services/hume/tts.py	0.00% <0.00%> (ø)
src/pipecat/services/kokoro/tts.py	0.00% <0.00%> (ø)
src/pipecat/services/lmnt/tts.py	0.00% <0.00%> (ø)
... and 21 more

🚀 New features to boost your workflow:

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

[markbackman]

@filipi87 I started work on this here:
#3465

Aleix raised this issue that prompted me to stop work on this:
#3465 (comment)

[filipi87]

@markbackman, @aconchillo,

I am halfway through validating that all the TTS services are working after this refactor. I am only missing tests for 12 more TTS services.

However, I believe this PR already addresses the main concern Aleix mentioned in the other PR:
https://github.com/pipecat-ai/pipecat/pull/3465/changes#r2696581006

If you could already provide a review for this one, it would be greatly appreciated.

I will also bring in any other commits that Mark made there which I haven’t addressed in this PR yet.

[filipi87]

@markbackman, @aconchillo, I think it’s now fully ready for review.
I’ll add the changelog later, once we’re happy with all the changes here.

[markbackman]

Should context_id be Optional here? Without that, this is a breaking change for run_tts(). While you've covered all cases, there are developers that have built their own services that we should support.

[filipi87]

Makes sense. 👍

[filipi87]

@mark, to be honest, I am having second thoughts about making this optional.

I know this would allow TTS services in third party projects to remain compatible with the current signature, but it would also introduce a downside: all the TTS services here would need to add extra logic to handle an optional context_id even though, in practice, it would never be optional.

So I think we should introduce this breaking change now, before 1.0.0. All the TTS services inside Pipecat will already be refactored to handle this anyway.

cc: @aconchillo for thoughts.