friday-framework-speech 0.1.0a0__tar.gz

friday_framework_speech-0.1.0a0/.gitignore

@@ -0,0 +1,24 @@
+__pycache__/
+*.pyc
+.env
+.venv/
+.idea/
+.vscode/
+dist/
+build/
+*.egg-info/
+chroma_db/
+.friday/
+/vector/
+/exports/
+tests/integration/memory/vector/*/
+tests/integration/memory/vector/backups/*/
+tests/integration/memory/vector/reorganization_logs/*
+*.gpickle
+keys.txt
+experiments/*/config/runtime.local.yaml
+experiments/*/artifacts/*
+!experiments/*/artifacts/.gitkeep
+scripts/source-friday-chat-example-env.sh
+scripts/pypi.sh
+/docs/reference/PyPI-Recovery-Codes-cichuck-2026-07-20T15_36_49.709334.txt

friday_framework_speech-0.1.0a0/PKG-INFO

@@ -0,0 +1,56 @@
+Metadata-Version: 2.4
+Name: friday-framework-speech
+Version: 0.1.0a0
+Summary: Speech services (TTS, wake word) for Friday
+Project-URL: Homepage, https://github.com/CIChuck/agent-framework
+Project-URL: Repository, https://github.com/CIChuck/agent-framework
+Project-URL: Issues, https://github.com/CIChuck/agent-framework/issues
+Project-URL: Documentation, https://github.com/CIChuck/agent-framework/tree/main/docs
+Project-URL: Source, https://github.com/CIChuck/agent-framework/tree/main/packages/friday-speech
+Author: Friday Team
+License-Expression: MIT
+Keywords: assistant,audio,speech,tts,wakeword
+Requires-Python: >=3.10
+Requires-Dist: friday-framework-core==0.1.0a0
+Requires-Dist: pydantic>=2.0.0
+Requires-Dist: pyyaml>=6.0.0
+Provides-Extra: dev
+Requires-Dist: pytest-asyncio>=0.23.0; extra == 'dev'
+Requires-Dist: pytest>=8.0.0; extra == 'dev'
+Provides-Extra: elevenlabs
+Requires-Dist: elevenlabs>=1.0.0; extra == 'elevenlabs'
+Provides-Extra: picovoice
+Requires-Dist: pvcobra>=2.0.0; extra == 'picovoice'
+Requires-Dist: pvleopard>=2.0.0; extra == 'picovoice'
+Requires-Dist: pvorca>=0.1.0; extra == 'picovoice'
+Requires-Dist: pvporcupine>=3.0.0; extra == 'picovoice'
+Requires-Dist: pvrecorder>=1.2.0; extra == 'picovoice'
+Requires-Dist: pvrhino>=2.0.0; extra == 'picovoice'
+Requires-Dist: pvspeaker>=0.1.0; extra == 'picovoice'
+Description-Content-Type: text/markdown
+
+# Friday Framework Speech
+
+Speech service package for the Friday Framework.
+
+This package provides speech service abstractions and optional text-to-speech,
+wake-word, recording, playback, and speech-provider integrations.
+
+## Install
+
+```bash
+pip install friday-framework-speech
+```
+
+Optional speech providers are available through extras such as `elevenlabs` and
+`picovoice`.
+
+## Import Package
+
+```python
+import friday_speech
+```
+
+## License
+
+MIT

friday_framework_speech-0.1.0a0/README.md

@@ -0,0 +1,25 @@
+# Friday Framework Speech
+
+Speech service package for the Friday Framework.
+
+This package provides speech service abstractions and optional text-to-speech,
+wake-word, recording, playback, and speech-provider integrations.
+
+## Install
+
+```bash
+pip install friday-framework-speech
+```
+
+Optional speech providers are available through extras such as `elevenlabs` and
+`picovoice`.
+
+## Import Package
+
+```python
+import friday_speech
+```
+
+## License
+
+MIT

friday_framework_speech-0.1.0a0/pyproject.toml

@@ -0,0 +1,55 @@
+[project]
+name = "friday-framework-speech"
+version = "0.1.0a0"
+description = "Speech services (TTS, wake word) for Friday"
+readme = "README.md"
+requires-python = ">=3.10"
+license = "MIT"
+authors = [{ name = "Friday Team" }]
+keywords = ["speech", "tts", "wakeword", "audio", "assistant"]
+
+dependencies = [
+    "pydantic>=2.0.0",
+    "friday-framework-core==0.1.0a0",
+    "pyyaml>=6.0.0",
+]
+
+[project.urls]
+Homepage = "https://github.com/CIChuck/agent-framework"
+Repository = "https://github.com/CIChuck/agent-framework"
+Issues = "https://github.com/CIChuck/agent-framework/issues"
+Documentation = "https://github.com/CIChuck/agent-framework/tree/main/docs"
+Source = "https://github.com/CIChuck/agent-framework/tree/main/packages/friday-speech"
+
+[project.optional-dependencies]
+elevenlabs = [
+    "elevenlabs>=1.0.0",
+]
+picovoice = [
+    "pvporcupine>=3.0.0",
+    "pvrecorder>=1.2.0",
+    "pvorca>=0.1.0",
+    "pvspeaker>=0.1.0",
+    "pvcobra>=2.0.0",
+    "pvleopard>=2.0.0",
+    "pvrhino>=2.0.0",
+]
+dev = [
+    "pytest>=8.0.0",
+    "pytest-asyncio>=0.23.0",
+]
+
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[tool.hatch.build.targets.wheel]
+packages = ["src/friday_speech"]
+
+[tool.pytest.ini_options]
+testpaths = ["tests_speech"]
+asyncio_mode = "auto"
+asyncio_default_fixture_loop_scope = "function"
+
+[tool.uv.sources]
+friday-framework-core = { workspace = true }

friday_framework_speech-0.1.0a0/src/friday_speech/__init__.py

@@ -0,0 +1,44 @@
+# mypy: ignore-errors
+"""Friday speech package."""
+
+from .config import SpeechConfig
+from .errors import (
+    AdapterError,
+    DependencyError,
+    PlaybackError,
+    SelectionError,
+    SpeechConfigError,
+    SpeechError,
+    STTError,
+    WakeWordError,
+)
+from .harness import SpeechHarness
+from .interfaces import (
+    AdapterRegistry,
+    AudioFormat,
+    AudioStream,
+    STTAdapter,
+    TTSAdapter,
+    WakeWordAdapter,
+)
+from .service import SpeechService
+
+__all__ = [
+    "AdapterError",
+    "AdapterRegistry",
+    "AudioFormat",
+    "AudioStream",
+    "DependencyError",
+    "PlaybackError",
+    "SelectionError",
+    "SpeechConfig",
+    "SpeechConfigError",
+    "SpeechError",
+    "SpeechHarness",
+    "SpeechService",
+    "STTAdapter",
+    "STTError",
+    "TTSAdapter",
+    "WakeWordAdapter",
+    "WakeWordError",
+]

friday_framework_speech-0.1.0a0/src/friday_speech/adapters/__init__.py

@@ -0,0 +1,43 @@
+# mypy: ignore-errors
+"""Adapter factory helpers for friday-speech."""
+
+from __future__ import annotations
+
+from ..config import SpeechConfig, TTSProvider, WakeWordProvider
+from ..errors import AdapterError
+from ..interfaces import STTAdapter, TTSAdapter, WakeWordAdapter
+
+
+def build_tts_adapter(config: SpeechConfig) -> TTSAdapter:
+    """Create a TTS adapter for the configured provider."""
+    if config.tts.provider == TTSProvider.ELEVENLABS:
+        from .elevenlabs import ElevenLabsAdapter
+
+        return ElevenLabsAdapter(config.elevenlabs)
+    if config.tts.provider == TTSProvider.PICOVOICE:
+        from .picovoice import PicovoiceOrcaAdapter
+
+        return PicovoiceOrcaAdapter(config.picovoice)
+    raise AdapterError(f"Unsupported TTS provider: {config.tts.provider}")
+
+
+def build_wakeword_adapter(config: SpeechConfig) -> WakeWordAdapter | None:
+    """Create a wake word adapter if enabled."""
+    if not config.wakeword.enabled:
+        return None
+    if config.wakeword.provider == WakeWordProvider.PICOVOICE:
+        from .picovoice import PicovoicePorcupineAdapter
+
+        return PicovoicePorcupineAdapter(config.picovoice)
+    raise AdapterError(f"Unsupported wake word provider: {config.wakeword.provider}")
+
+
+def build_stt_adapter(config: SpeechConfig) -> STTAdapter | None:
+    """Create a speech-to-text adapter if enabled."""
+    if not config.stt.enabled:
+        return None
+    if config.stt.provider.value == "picovoice":
+        from .picovoice import PicovoiceLeopardAdapter
+
+        return PicovoiceLeopardAdapter(config.picovoice)
+    raise AdapterError(f"Unsupported STT provider: {config.stt.provider}")

friday_framework_speech-0.1.0a0/src/friday_speech/adapters/elevenlabs.py

@@ -0,0 +1,183 @@
+# mypy: ignore-errors
+"""ElevenLabs TTS adapter (best-effort streaming)."""
+
+from __future__ import annotations
+
+import asyncio
+import re
+import threading
+from collections.abc import AsyncIterator, Iterable
+from typing import Any
+
+from ..config import ElevenLabsConfig, VoiceSettingsConfig
+from ..errors import AdapterError, DependencyError
+from ..interfaces import AudioFormat, AudioStream, TTSAdapter
+
+
+class ElevenLabsAdapter(TTSAdapter):
+    """ElevenLabs adapter using the official Python SDK when available."""
+
+    def __init__(self, config: ElevenLabsConfig) -> None:
+        self._config = config
+        self.output_format = _infer_audio_format(config.output_format)
+        self.sample_rate = _infer_sample_rate(config.output_format)
+        self.channels = None
+
+    async def stream(
+        self,
+        text: str,
+        voice_id: str | None,
+        settings: VoiceSettingsConfig,
+    ) -> AudioStream:
+        if not self._config.api_key:
+            raise AdapterError("ElevenLabs API key is required")
+
+        elevenlabs = _import_elevenlabs()
+        client = _build_client(elevenlabs, self._config.api_key)
+        payload = _build_voice_settings(settings)
+
+        text_to_speech = getattr(client, "text_to_speech", None)
+        if text_to_speech is None:
+            raise AdapterError("ElevenLabs client missing text_to_speech interface")
+
+        if hasattr(text_to_speech, "convert_as_stream"):
+            sync_stream = text_to_speech.convert_as_stream(
+                text=text,
+                voice_id=voice_id or self._config.voice_id,
+                model_id=self._config.model_id,
+                output_format=self._config.output_format,
+                voice_settings=payload or None,
+            )
+            chunks = _async_from_sync_iterator(sync_stream)
+            return AudioStream(
+                format=self.output_format,
+                chunks=chunks,
+                sample_rate=self.sample_rate,
+                channels=self.channels,
+            )
+
+        if hasattr(text_to_speech, "convert"):
+            audio = await asyncio.to_thread(
+                text_to_speech.convert,
+                text=text,
+                voice_id=voice_id or self._config.voice_id,
+                model_id=self._config.model_id,
+                output_format=self._config.output_format,
+                voice_settings=payload or None,
+            )
+            chunks = _async_single_chunk(audio)
+            return AudioStream(
+                format=self.output_format,
+                chunks=chunks,
+                sample_rate=self.sample_rate,
+                channels=self.channels,
+            )
+
+        raise AdapterError(
+            "Unsupported ElevenLabs SDK: expected text_to_speech.convert[_as_stream]"
+        )
+
+
+def _import_elevenlabs() -> Any:
+    try:
+        import elevenlabs  # type: ignore
+
+        return elevenlabs
+    except Exception as exc:  # pragma: no cover - dependency gate
+        raise DependencyError(
+            "ElevenLabs SDK not installed. Install with: pip install elevenlabs"
+        ) from exc
+
+
+def _build_client(module: Any, api_key: str) -> Any:
+    for path in ("client", None):
+        try:
+            if path:
+                client_mod = __import__(f"elevenlabs.{path}", fromlist=["ElevenLabs"])
+                client_cls = getattr(client_mod, "ElevenLabs", None)
+            else:
+                client_cls = getattr(module, "ElevenLabs", None)
+            if client_cls:
+                return client_cls(api_key=api_key)
+        except Exception:
+            continue
+    raise AdapterError("Unable to initialize ElevenLabs client")
+
+
+def _build_voice_settings(settings: VoiceSettingsConfig) -> dict[str, Any]:
+    payload: dict[str, Any] = {}
+    if settings.stability is not None:
+        payload["stability"] = settings.stability
+    if settings.similarity_boost is not None:
+        payload["similarity_boost"] = settings.similarity_boost
+    if settings.style is not None:
+        payload["style"] = settings.style
+    if settings.speed is not None:
+        payload["speed"] = settings.speed
+    return payload
+
+
+def _infer_audio_format(output_format: str) -> AudioFormat:
+    value = output_format.lower()
+    if value.startswith("mp3"):
+        return AudioFormat.MP3
+    if value.startswith("pcm"):
+        return AudioFormat.PCM
+    raise AdapterError(f"Unsupported ElevenLabs output_format: {output_format}")
+
+
+def _infer_sample_rate(output_format: str) -> int | None:
+    match = re.search(r"(\d{4,6})", output_format)
+    if not match:
+        return None
+    try:
+        return int(match.group(1))
+    except ValueError:
+        return None
+
+
+def _async_single_chunk(audio: Any) -> AsyncIterator[bytes]:
+    async def _generator() -> AsyncIterator[bytes]:
+        if isinstance(audio, (bytes, bytearray)):
+            yield bytes(audio)
+            return
+        if isinstance(audio, Iterable):
+            for chunk in audio:
+                if isinstance(chunk, (bytes, bytearray)):
+                    yield bytes(chunk)
+                else:
+                    raise AdapterError("ElevenLabs stream yielded non-bytes chunk")
+            return
+        raise AdapterError("ElevenLabs audio payload is not bytes or iterable")
+
+    return _generator()
+
+
+def _async_from_sync_iterator(stream: Iterable[Any]) -> AsyncIterator[bytes]:
+    async def _generator() -> AsyncIterator[bytes]:
+        queue: asyncio.Queue[Any] = asyncio.Queue()
+        sentinel = object()
+        loop = asyncio.get_running_loop()
+
+        def _worker() -> None:
+            try:
+                for chunk in stream:
+                    loop.call_soon_threadsafe(queue.put_nowait, chunk)
+            except Exception as exc:
+                loop.call_soon_threadsafe(queue.put_nowait, exc)
+            finally:
+                loop.call_soon_threadsafe(queue.put_nowait, sentinel)
+
+        threading.Thread(target=_worker, daemon=True).start()
+
+        while True:
+            item = await queue.get()
+            if item is sentinel:
+                break
+            if isinstance(item, Exception):
+                raise item
+            if not isinstance(item, (bytes, bytearray)):
+                raise AdapterError("ElevenLabs stream yielded non-bytes chunk")
+            yield bytes(item)
+
+    return _generator()