feat: add static voice catalog to PluginManifest

- Add  to PluginManifest
  - None = not declared (use VoiceLister fallback)
  - () = explicitly no static voices
  - Non-empty = static catalog available without Engine instantiation

- Update get_voices() to check manifest first, fall back to Engine
- Declare 54 Kokoro voices and 10 SuperTonic voices in manifests
- Remove hardcoded voice lists from engine.py files
- Engine.listVoices() now returns [] (manifest is source of truth)

- Clean up dead create_pipeline() kwargs (sample_rate, auto_download, total_steps)
  - SuperTonic plugin uses internal defaults
  - total_steps is per-request parameter via Pipeline.__call__() kwargs

- Add clear_preview_pipelines() for resource cleanup
- Fix test mocks to override listVoices()
- Update Architecture Amendment #1 doc

plugins/kokoro/__init__.py

@@ -1,121 +1,178 @@
-"""Kokoro TTS Plugin for the TTS Plugin Architecture.
-
-This plugin provides a Kokoro-based TTS engine that implements the
-Plugin API contract. It wraps the existing Kokoro backend in the
-new Engine/EngineSession architecture.
-
-Exports:
-    - PLUGIN_MANIFEST: PluginManifest
-    - MODEL_REQUIREMENTS: list[ModelManifest]
-    - create_engine: Factory function
-"""
-
-from __future__ import annotations
-
-from pathlib import Path
-from typing import Any
-
-from abogen.tts_plugin.engine import Engine
-from abogen.tts_plugin.host_context import HostContext
-from abogen.tts_plugin.manifest import (
-    AudioFormatManifest,
-    EngineManifest,
-    ModelManifest,
-    ParameterManifest,
-    PluginManifest,
-    RequirementManifest,
-    VoiceSourceManifest,
+"""Kokoro TTS Plugin for the TTS Plugin Architecture.
+
+This plugin provides a Kokoro-based TTS engine that implements the
+Plugin API contract. It wraps the existing Kokoro backend in the
+new Engine/EngineSession architecture.
+
+Exports:
+    - PLUGIN_MANIFEST: PluginManifest
+    - MODEL_REQUIREMENTS: list[ModelManifest]
+    - create_engine: Factory function
+"""
+
+from __future__ import annotations
+
+from pathlib import Path
+from typing import Any
+
+from abogen.tts_plugin.engine import Engine
+from abogen.tts_plugin.host_context import HostContext
+from abogen.tts_plugin.manifest import (
+    AudioFormatManifest,
+    EngineManifest,
+    ModelManifest,
+    ParameterManifest,
+    PluginManifest,
+    RequirementManifest,
+    VoiceManifest,
+    VoiceSourceManifest,
 )
-from abogen.tts_plugin.types import EngineConfig
-
-from .engine import KokoroEngine
-
-
-def _load_kpipeline() -> Any:
-    """Lazy-load Kokoro dependencies."""
-    from kokoro import KPipeline  # type: ignore[import-not-found]
-    return KPipeline
-
-
-PLUGIN_MANIFEST = PluginManifest(
-    id="kokoro",
-    name="Kokoro",
-    version="0.9.4",
-    api_version="1.0",
-    description="Kokoro TTS engine - high quality multilingual text-to-speech",
-    author="Kokoro Team",
-    capabilities=("voice_list",),
-    requires=RequirementManifest(
-        internet=False,
-    ),
-    engine=EngineManifest(
-        voiceSources=(
-            VoiceSourceManifest(
-                id="builtin",
-                name="Built-in Voices",
-                type="list",
-                config={"voices": "See listVoices()"},
-            ),
-        ),
-        parameters=(
-            ParameterManifest(
-                id="speed",
-                name="Speed",
-                description="Speech speed multiplier",
-                type="float",
-                default=1.0,
-                min=0.5,
-                max=2.0,
-                step=0.1,
-            ),
-        ),
-        audioFormats=(
-            AudioFormatManifest(mime="audio/wav", extension="wav"),
-        ),
-    ),
-)
-
-MODEL_REQUIREMENTS: list[ModelManifest] = []
-
-
-def create_engine(
-    context: HostContext,
-    model_path: Path | None,
-    config: EngineConfig,
-) -> Engine:
-    """Create a Kokoro engine instance.
-
-    This function is the plugin entry point. It must be atomic:
-    succeed fully or raise EngineError and clean up.
-
-    Args:
-        context: Host services (config dir, logger, http client).
-        model_path: Resolved model path, or None for default.
-        config: Engine initialization settings (device, etc.).
-
-    Returns:
-        A fully initialized KokoroEngine instance.
-
-    Raises:
-        EngineError: On failure. Cleans up partially created resources.
-    """
-    try:
-        KPipeline = _load_kpipeline()
-
-        # Determine repo_id from model_path or use default
-        repo_id = "hexgrad/Kokoro-82M"
-        if model_path is not None:
-            # If a specific model path is provided, use it as repo_id
-            repo_id = str(model_path)
-
-        pipeline = KPipeline(
-            lang_code="a",  # Default language code
-            repo_id=repo_id,
-            device=config.device,
-        )
-
-        engine = KokoroEngine(pipeline, lang_code="a")
-        return engine
-    except Exception as e:
-        from abogen.tts_plugin.errors import EngineError as EngineErrorClass
-        raise EngineErrorClass(f"Failed to create Kokoro engine: {e}") from e
+from abogen.tts_plugin.types import EngineConfig
+
+from .engine import KokoroEngine
+
+
+def _load_kpipeline() -> Any:
+    """Lazy-load Kokoro dependencies."""
+    from kokoro import KPipeline  # type: ignore[import-not-found]
+    return KPipeline
+
+
+PLUGIN_MANIFEST = PluginManifest(
+    id="kokoro",
+    name="Kokoro",
+    version="0.9.4",
+    api_version="1.0",
+    description="Kokoro TTS engine - high quality multilingual text-to-speech",
+    author="Kokoro Team",
+    capabilities=("voice_list",),
+    requires=RequirementManifest(
+        internet=False,
+    ),
+    engine=EngineManifest(
+        voiceSources=(
+            VoiceSourceManifest(
+                id="builtin",
+                name="Built-in Voices",
+                type="list",
+                config={"voices": "See listVoices()"},
+            ),
+        ),
+        parameters=(
+            ParameterManifest(
+                id="speed",
+                name="Speed",
+                description="Speech speed multiplier",
+                type="float",
+                default=1.0,
+                min=0.5,
+                max=2.0,
+                step=0.1,
+            ),
+        ),
+        audioFormats=(
+            AudioFormatManifest(mime="audio/wav", extension="wav"),
+        ),
+    ),
+    voices=(
+        VoiceManifest(id="af_alloy", name="Alloy", tags=("en", "female")),
+        VoiceManifest(id="af_aoede", name="Aoede", tags=("en", "female")),
+        VoiceManifest(id="af_bella", name="Bella", tags=("en", "female")),
+        VoiceManifest(id="af_heart", name="Heart", tags=("en", "female")),
+        VoiceManifest(id="af_jessica", name="Jessica", tags=("en", "female")),
+        VoiceManifest(id="af_kore", name="Kore", tags=("en", "female")),
+        VoiceManifest(id="af_nicole", name="Nicole", tags=("en", "female")),
+        VoiceManifest(id="af_nova", name="Nova", tags=("en", "female")),
+        VoiceManifest(id="af_river", name="River", tags=("en", "female")),
+        VoiceManifest(id="af_sarah", name="Sarah", tags=("en", "female")),
+        VoiceManifest(id="af_sky", name="Sky", tags=("en", "female")),
+        VoiceManifest(id="am_adam", name="Adam", tags=("en", "male")),
+        VoiceManifest(id="am_echo", name="Echo", tags=("en", "male")),
+        VoiceManifest(id="am_eric", name="Eric", tags=("en", "male")),
+        VoiceManifest(id="am_fenrir", name="Fenrir", tags=("en", "male")),
+        VoiceManifest(id="am_liam", name="Liam", tags=("en", "male")),
+        VoiceManifest(id="am_michael", name="Michael", tags=("en", "male")),
+        VoiceManifest(id="am_onyx", name="Onyx", tags=("en", "male")),
+        VoiceManifest(id="am_puck", name="Puck", tags=("en", "male")),
+        VoiceManifest(id="am_santa", name="Santa", tags=("en", "male")),
+        VoiceManifest(id="bf_alice", name="Alice", tags=("en", "female")),
+        VoiceManifest(id="bf_emma", name="Emma", tags=("en", "female")),
+        VoiceManifest(id="bf_isabella", name="Isabella", tags=("en", "female")),
+        VoiceManifest(id="bf_lily", name="Lily", tags=("en", "female")),
+        VoiceManifest(id="bm_daniel", name="Daniel", tags=("en", "male")),
+        VoiceManifest(id="bm_fable", name="Fable", tags=("en", "male")),
+        VoiceManifest(id="bm_george", name="George", tags=("en", "male")),
+        VoiceManifest(id="bm_lewis", name="Lewis", tags=("en", "male")),
+        VoiceManifest(id="ef_dora", name="Dora", tags=("es", "female")),
+        VoiceManifest(id="em_alex", name="Alex", tags=("es", "male")),
+        VoiceManifest(id="em_santa", name="Santa", tags=("es", "male")),
+        VoiceManifest(id="ff_siwis", name="Siwis", tags=("fr", "female")),
+        VoiceManifest(id="hf_alpha", name="Alpha", tags=("hi", "female")),
+        VoiceManifest(id="hf_beta", name="Beta", tags=("hi", "female")),
+        VoiceManifest(id="hm_omega", name="Omega", tags=("hi", "male")),
+        VoiceManifest(id="hm_psi", name="Psi", tags=("hi", "male")),
+        VoiceManifest(id="if_sara", name="Sara", tags=("it", "female")),
+        VoiceManifest(id="im_nicola", name="Nicola", tags=("it", "male")),
+        VoiceManifest(id="jf_alpha", name="Alpha", tags=("ja", "female")),
+        VoiceManifest(id="jf_gongitsune", name="Gongitsune", tags=("ja", "female")),
+        VoiceManifest(id="jf_nezumi", name="Nezumi", tags=("ja", "female")),
+        VoiceManifest(id="jf_tebukuro", name="Tebukuro", tags=("ja", "female")),
+        VoiceManifest(id="jm_kumo", name="Kumo", tags=("ja", "male")),
+        VoiceManifest(id="pf_dora", name="Dora", tags=("pt", "female")),
+        VoiceManifest(id="pm_alex", name="Alex", tags=("pt", "male")),
+        VoiceManifest(id="pm_santa", name="Santa", tags=("pt", "male")),
+        VoiceManifest(id="zf_xiaobei", name="Xiaobei", tags=("zh", "female")),
+        VoiceManifest(id="zf_xiaoni", name="Xiaoni", tags=("zh", "female")),
+        VoiceManifest(id="zf_xiaoxiao", name="Xiaoxiao", tags=("zh", "female")),
+        VoiceManifest(id="zf_xiaoyi", name="Xiaoyi", tags=("zh", "female")),
+        VoiceManifest(id="zm_yunjian", name="Yunjian", tags=("zh", "male")),
+        VoiceManifest(id="zm_yunxi", name="Yunxi", tags=("zh", "male")),
+        VoiceManifest(id="zm_yunxia", name="Yunxia", tags=("zh", "female")),
+        VoiceManifest(id="zm_yunyang", name="Yunyang", tags=("zh", "male")),
+    ),
+)
+
+MODEL_REQUIREMENTS: list[ModelManifest] = []
+
+
+def create_engine(
+    context: HostContext,
+    model_path: Path | None,
+    config: EngineConfig,
+) -> Engine:
+    """Create a Kokoro engine instance.
+
+    This function is the plugin entry point. It must be atomic:
+    succeed fully or raise EngineError and clean up.
+
+    Args:
+        context: Host services (config dir, logger, http client).
+        model_path: Resolved model path, or None for default.
+        config: Engine initialization settings (device, etc.).
+
+    Returns:
+        A fully initialized KokoroEngine instance.
+
+    Raises:
+        EngineError: On failure. Cleans up partially created resources.
+    """
+    try:
+        KPipeline = _load_kpipeline()
+
+        # Determine repo_id from model_path or use default
+        repo_id = "hexgrad/Kokoro-82M"
+        if model_path is not None:
+            # If a specific model path is provided, use it as repo_id
+            repo_id = str(model_path)
+
+        pipeline = KPipeline(
+            lang_code=config.lang_code,
+            repo_id=repo_id,
+            device=config.device,
+        )
+
+        engine = KokoroEngine(pipeline)
+        return engine
+    except Exception as e:
+        from abogen.tts_plugin.errors import EngineError as EngineErrorClass
+        raise EngineErrorClass(f"Failed to create Kokoro engine: {e}") from e

plugins/kokoro/engine.py

@@ -1,186 +1,118 @@
-"""Kokoro Engine adapter for the TTS Plugin Architecture.
-
-This module adapts the existing Kokoro backend to the new Engine/EngineSession
-protocol. It wraps the KokoroBackend without modifying it.
-"""
-
-from __future__ import annotations
-
-import logging
-from typing import Any, Iterator
-
-import numpy as np
-
-from abogen.tts_plugin.capabilities import VoiceLister
-from abogen.tts_plugin.engine import Engine, EngineSession
-from abogen.tts_plugin.errors import EngineError, InvalidInputError
-from abogen.tts_plugin.manifest import VoiceManifest
-from abogen.tts_plugin.types import (
-    AudioFormat,
-    Duration,
-    ParameterValues,
-    SynthesisRequest,
-    SynthesizedAudio,
-    VoiceSelection,
-)
-
-logger = logging.getLogger(__name__)
-
-# Kokoro voice list - source of truth
-_KOKORO_VOICES = (
-    "af_alloy", "af_aoede", "af_bella", "af_heart", "af_jessica",
-    "af_kore", "af_nicole", "af_nova", "af_river", "af_sarah",
-    "af_sky", "am_adam", "am_echo", "am_eric", "am_fenrir",
-    "am_liam", "am_michael", "am_onyx", "am_puck", "am_santa",
-    "bf_alice", "bf_emma", "bf_isabella", "bf_lily",
-    "bm_daniel", "bm_fable", "bm_george", "bm_lewis",
-    "ef_dora", "em_alex", "em_santa",
-    "ff_siwis", "hf_alpha", "hf_beta", "hm_omega", "hm_psi",
-    "if_sara", "im_nicola",
-    "jf_alpha", "jf_gongitsune", "jf_nezumi", "jf_tebukuro", "jm_kumo",
-    "pf_dora", "pm_alex", "pm_santa",
-    "zf_xiaobei", "zf_xiaoni", "zf_xiaoxiao", "zf_xiaoyi",
-    "zm_yunjian", "zm_yunxi", "zm_yunxia", "zm_yunyang",
-)
-
-# Voice display names mapping
-_VOICE_DISPLAY_NAMES: dict[str, str] = {
-    "af_alloy": "Alloy", "af_aoede": "Aoede", "af_bella": "Bella",
-    "af_heart": "Heart", "af_jessica": "Jessica", "af_kore": "Kore",
-    "af_nicole": "Nicole", "af_nova": "Nova", "af_river": "River",
-    "af_sarah": "Sarah", "af_sky": "Sky", "am_adam": "Adam",
-    "am_echo": "Echo", "am_eric": "Eric", "am_fenrir": "Fenrir",
-    "am_liam": "Liam", "am_michael": "Michael", "am_onyx": "Onyx",
-    "am_puck": "Puck", "am_santa": "Santa", "bf_alice": "Alice",
-    "bf_emma": "Emma", "bf_isabella": "Isabella", "bf_lily": "Lily",
-    "bm_daniel": "Daniel", "bm_fable": "Fable", "bm_george": "George",
-    "bm_lewis": "Lewis", "ef_dora": "Dora", "em_alex": "Alex",
-    "em_santa": "Santa", "ff_siwis": "Siwis", "hf_alpha": "Alpha",
-    "hf_beta": "Beta", "hm_omega": "Omega", "hm_psi": "Psi",
-    "if_sara": "Sara", "im_nicola": "Nicola",
-    "jf_alpha": "Alpha", "jf_gongitsune": "Gongitsune",
-    "jf_nezumi": "Nezumi", "jf_tebukuro": "Tebukuro", "jm_kumo": "Kumo",
-    "pf_dora": "Dora", "pm_alex": "Alex", "pm_santa": "Santa",
-    "zf_xiaobei": "Xiaobei", "zf_xiaoni": "Xiaoni",
-    "zf_xiaoxiao": "Xiaoxiao", "zf_xiaoyi": "Xiaoyi",
-    "zm_yunjian": "Yunjian", "zm_yunxi": "Yunxi",
-    "zm_yunxia": "Yunxia", "zm_yunyang": "Yunyang",
-}
-
-# Sample rate for Kokoro audio
-_KOKORO_SAMPLE_RATE = 24000
-
-
-class KokoroSession:
-    """EngineSession implementation for Kokoro.
-
-    Owns mutable execution state for synthesis.
-    NOT thread-safe.
-    """
-
-    def __init__(self, pipeline: Any, lang_code: str) -> None:
-        self._pipeline = pipeline
-        self._lang_code = lang_code
-        self._disposed = False
-
-    def synthesize(self, request: SynthesisRequest) -> SynthesizedAudio:
-        """Synthesize audio from text using Kokoro."""
-        if self._disposed:
-            raise EngineError("Session disposed")
-
-        try:
-            voice = request.voice.key
-            speed = request.parameters.values.get("speed", 1.0)
-            split_pattern = request.parameters.values.get("split_pattern", None)
-
-            audio_parts: list[np.ndarray] = []
-            for segment in self._pipeline(
-                request.text,
-                voice=voice,
-                speed=speed,
-                split_pattern=split_pattern,
-            ):
-                audio = segment.audio
-                if hasattr(audio, "numpy"):
-                    audio = audio.numpy()
-                audio_parts.append(np.asarray(audio, dtype="float32"))
-
-            if not audio_parts:
-                return SynthesizedAudio(
-                    data=b"",
-                    format=AudioFormat(mime="audio/wav", extension="wav"),
-                    duration=Duration(seconds=0.0),
-                )
-
-            combined = np.concatenate(audio_parts).astype("float32", copy=False)
-            audio_bytes = combined.tobytes()
-            duration_seconds = len(combined) / _KOKORO_SAMPLE_RATE
-
-            return SynthesizedAudio(
-                data=audio_bytes,
-                format=AudioFormat(mime="audio/wav", extension="wav"),
-                duration=Duration(seconds=duration_seconds),
-            )
-        except EngineError:
-            raise
-        except Exception as e:
-            raise EngineError(f"Synthesis failed: {e}") from e
-
-    def dispose(self) -> None:
-        """Release session resources. Idempotent."""
-        self._disposed = True
-
-
-class KokoroEngine:
-    """Engine implementation for Kokoro.
-
-    Factory for KokoroSession instances. Stateless and thread-safe.
-    """
-
-    def __init__(self, pipeline: Any, lang_code: str) -> None:
-        self._pipeline = pipeline
-        self._lang_code = lang_code
-        self._disposed = False
-
-    def createSession(self) -> KokoroSession:
-        """Create a new KokoroSession."""
-        if self._disposed:
-            raise EngineError("Engine disposed")
-        return KokoroSession(self._pipeline, self._lang_code)
-
-    def dispose(self) -> None:
-        """Release engine resources. Idempotent."""
-        self._disposed = True
-
-    def listVoices(self, sourceId: str) -> list[VoiceManifest]:
-        """List available Kokoro voices. Implements VoiceLister capability."""
-        if self._disposed:
-            raise EngineError("Engine disposed")
-        return [
-            VoiceManifest(
-                id=voice_id,
-                name=_VOICE_DISPLAY_NAMES.get(voice_id, voice_id),
-                tags=(_get_language_tag(voice_id), _get_gender_tag(voice_id)),
-            )
-            for voice_id in _KOKORO_VOICES
-        ]
-
-
-def _get_language_tag(voice_id: str) -> str:
-    """Extract language tag from voice ID."""
-    prefix = voice_id.split("_")[0]
-    lang_map = {
-        "af": "en-us", "am": "en-us", "bf": "en-gb", "bm": "en-gb",
-        "ef": "es", "em": "es", "ff": "fr", "hf": "hi", "hm": "hi",
-        "if": "it", "im": "it", "jf": "ja", "jm": "ja",
-        "pf": "pt", "pm": "pt", "zf": "zh", "zm": "zh",
-    }
-    return lang_map.get(prefix, "unknown")
-
-
-def _get_gender_tag(voice_id: str) -> str:
-    """Extract gender tag from voice ID."""
-    prefix = voice_id.split("_")[0]
-    if prefix.startswith("a") or prefix.startswith("b") or prefix.startswith("e"):
-        return "female" if prefix[1] == "f" else "male"
-    return "unknown"
+"""Kokoro Engine adapter for the TTS Plugin Architecture.
+
+This module adapts the existing Kokoro backend to the new Engine/EngineSession
+protocol. It wraps the KokoroBackend without modifying it.
+"""
+
+from __future__ import annotations
+
+import logging
+from typing import Any
+
+import numpy as np
+
+from abogen.tts_plugin.capabilities import VoiceLister
+from abogen.tts_plugin.engine import Engine, EngineSession
+from abogen.tts_plugin.errors import EngineError
+from abogen.tts_plugin.manifest import VoiceManifest
+from abogen.tts_plugin.types import (
+    AudioFormat,
+    Duration,
+    SynthesisRequest,
+    SynthesizedAudio,
+)
+
+logger = logging.getLogger(__name__)
+
+# Sample rate for Kokoro audio
+_KOKORO_SAMPLE_RATE = 24000
+
+
+class KokoroSession:
+    """EngineSession implementation for Kokoro.
+
+    Owns mutable execution state for synthesis.
+    NOT thread-safe.
+    """
+
+    def __init__(self, pipeline: Any) -> None:
+        self._pipeline = pipeline
+        self._disposed = False
+
+    def synthesize(self, request: SynthesisRequest) -> SynthesizedAudio:
+        """Synthesize audio from text using Kokoro."""
+        if self._disposed:
+            raise EngineError("Session disposed")
+
+        try:
+            voice = request.voice.key
+            speed = request.parameters.values.get("speed", 1.0)
+            split_pattern = request.parameters.values.get("split_pattern", None)
+
+            audio_parts: list[np.ndarray] = []
+            for segment in self._pipeline(
+                request.text,
+                voice=voice,
+                speed=speed,
+                split_pattern=split_pattern,
+            ):
+                audio = segment.audio
+                if hasattr(audio, "numpy"):
+                    audio = audio.numpy()
+                audio_parts.append(np.asarray(audio, dtype="float32"))
+
+            if not audio_parts:
+                return SynthesizedAudio(
+                    data=b"",
+                    format=AudioFormat(mime="audio/wav", extension="wav"),
+                    duration=Duration(seconds=0.0),
+                )
+
+            combined = np.concatenate(audio_parts).astype("float32", copy=False)
+            audio_bytes = combined.tobytes()
+            duration_seconds = len(combined) / _KOKORO_SAMPLE_RATE
+
+            return SynthesizedAudio(
+                data=audio_bytes,
+                format=AudioFormat(mime="audio/wav", extension="wav"),
+                duration=Duration(seconds=duration_seconds),
+            )
+        except EngineError:
+            raise
+        except Exception as e:
+            raise EngineError(f"Synthesis failed: {e}") from e
+
+    def dispose(self) -> None:
+        """Release session resources. Idempotent."""
+        self._disposed = True
+
+
+class KokoroEngine:
+    """Engine implementation for Kokoro.
+
+    Factory for KokoroSession instances. Stateless and thread-safe.
+    """
+
+    def __init__(self, pipeline: Any) -> None:
+        self._pipeline = pipeline
+        self._disposed = False
+
+    def createSession(self) -> KokoroSession:
+        """Create a new KokoroSession."""
+        if self._disposed:
+            raise EngineError("Engine disposed")
+        return KokoroSession(self._pipeline)
+
+    def dispose(self) -> None:
+        """Release engine resources. Idempotent."""
+        self._disposed = True
+
+    def listVoices(self, sourceId: str) -> list[VoiceManifest]:
+        """List available Kokoro voices. Implements VoiceLister capability.
+
+        Note: Static voices are declared in the plugin manifest.
+        This method is a fallback for dynamic plugins.
+        """
+        if self._disposed:
+            raise EngineError("Engine disposed")
+        return []

plugins/supertonic/__init__.py

@@ -1,123 +1,136 @@
-"""SuperTonic TTS Plugin for the TTS Plugin Architecture.
-
-This plugin provides a SuperTonic-based TTS engine that implements the
-Plugin API contract. It wraps the existing SuperTonic backend in the
-new Engine/EngineSession architecture.
-
-Exports:
-    - PLUGIN_MANIFEST: PluginManifest
-    - MODEL_REQUIREMENTS: list[ModelManifest]
-    - create_engine: Factory function
-"""
-
-from __future__ import annotations
-
-from pathlib import Path
-from typing import Any
-
-from abogen.tts_plugin.engine import Engine
-from abogen.tts_plugin.host_context import HostContext
-from abogen.tts_plugin.manifest import (
-    AudioFormatManifest,
-    EngineManifest,
-    ModelManifest,
-    ParameterManifest,
-    PluginManifest,
-    RequirementManifest,
-    VoiceSourceManifest,
-)
-from abogen.tts_plugin.types import EngineConfig
-
-from .engine import SuperTonicEngine
-
-
-def _load_supertonic_pipeline(sample_rate: int = 24000, auto_download: bool = True, total_steps: int = 5) -> Any:
-    """Lazy-load SuperTonic dependencies and create pipeline."""
-    from plugins.supertonic.pipeline import SupertonicPipeline
-
-    return SupertonicPipeline(
-        sample_rate=sample_rate,
-        auto_download=auto_download,
-        total_steps=total_steps,
-    )
-
-
-PLUGIN_MANIFEST = PluginManifest(
-    id="supertonic",
-    name="SuperTonic",
-    version="0.1.0",
-    api_version="1.0",
-    description="SuperTonic TTS engine - fast high-quality text-to-speech",
-    author="SuperTonic Team",
-    capabilities=("voice_list",),
-    requires=RequirementManifest(
-        internet=False,
-    ),
-    engine=EngineManifest(
-        voiceSources=(
-            VoiceSourceManifest(
-                id="builtin",
-                name="Built-in Voices",
-                type="list",
-                config={"voices": "See listVoices()"},
-            ),
-        ),
-        parameters=(
-            ParameterManifest(
-                id="speed",
-                name="Speed",
-                description="Speech speed multiplier",
-                type="float",
-                default=1.0,
-                min=0.7,
-                max=2.0,
-                step=0.1,
-            ),
-            ParameterManifest(
-                id="total_steps",
-                name="Quality Steps",
-                description="Inference steps (higher = better quality, slower)",
-                type="int",
-                default=5,
-                min=2,
-                max=15,
-                step=1,
-            ),
-        ),
-        audioFormats=(
-            AudioFormatManifest(mime="audio/wav", extension="wav"),
-        ),
-    ),
-)
-
-MODEL_REQUIREMENTS: list[ModelManifest] = []
-
-
-def create_engine(
-    context: HostContext,
-    model_path: Path | None,
-    config: EngineConfig,
-) -> Engine:
-    """Create a SuperTonic engine instance.
-
-    This function is the plugin entry point. It must be atomic:
-    succeed fully or raise EngineError and clean up.
-
-    Args:
-        context: Host services (config dir, logger, http client).
-        model_path: Resolved model path, or None for default.
-        config: Engine initialization settings (device, etc.).
-
-    Returns:
-        A fully initialized SuperTonicEngine instance.
-
-    Raises:
-        EngineError: On failure. Cleans up partially created resources.
-    """
-    try:
-        pipeline = _load_supertonic_pipeline()
-        engine = SuperTonicEngine(pipeline)
-        return engine
-    except Exception as e:
-        from abogen.tts_plugin.errors import EngineError as EngineErrorClass
-        raise EngineErrorClass(f"Failed to create SuperTonic engine: {e}") from e
+"""SuperTonic TTS Plugin for the TTS Plugin Architecture.
+
+This plugin provides a SuperTonic-based TTS engine that implements the
+Plugin API contract. It wraps the existing SuperTonic backend in the
+new Engine/EngineSession architecture.
+
+Exports:
+    - PLUGIN_MANIFEST: PluginManifest
+    - MODEL_REQUIREMENTS: list[ModelManifest]
+    - create_engine: Factory function
+"""
+
+from __future__ import annotations
+
+from pathlib import Path
+from typing import Any
+
+from abogen.tts_plugin.engine import Engine
+from abogen.tts_plugin.host_context import HostContext
+from abogen.tts_plugin.manifest import (
+    AudioFormatManifest,
+    EngineManifest,
+    ModelManifest,
+    ParameterManifest,
+    PluginManifest,
+    RequirementManifest,
+    VoiceManifest,
+    VoiceSourceManifest,
+)
+from abogen.tts_plugin.types import EngineConfig
+
+from .engine import SuperTonicEngine
+
+
+def _load_supertonic_pipeline() -> Any:
+    """Lazy-load SuperTonic dependencies and create pipeline."""
+    from plugins.supertonic.pipeline import SupertonicPipeline
+
+    return SupertonicPipeline(
+        sample_rate=24000,
+        auto_download=True,
+        total_steps=5,
+    )
+
+
+PLUGIN_MANIFEST = PluginManifest(
+    id="supertonic",
+    name="SuperTonic",
+    version="0.1.0",
+    api_version="1.0",
+    description="SuperTonic TTS engine - fast high-quality text-to-speech",
+    author="SuperTonic Team",
+    capabilities=("voice_list",),
+    requires=RequirementManifest(
+        internet=False,
+    ),
+    engine=EngineManifest(
+        voiceSources=(
+            VoiceSourceManifest(
+                id="builtin",
+                name="Built-in Voices",
+                type="list",
+                config={"voices": "See listVoices()"},
+            ),
+        ),
+        parameters=(
+            ParameterManifest(
+                id="speed",
+                name="Speed",
+                description="Speech speed multiplier",
+                type="float",
+                default=1.0,
+                min=0.7,
+                max=2.0,
+                step=0.1,
+            ),
+            ParameterManifest(
+                id="total_steps",
+                name="Quality Steps",
+                description="Inference steps (higher = better quality, slower)",
+                type="int",
+                default=5,
+                min=2,
+                max=15,
+                step=1,
+            ),
+        ),
+        audioFormats=(
+            AudioFormatManifest(mime="audio/wav", extension="wav"),
+        ),
+    ),
+    voices=(
+        VoiceManifest(id="M1", name="Male 1", tags=("male",)),
+        VoiceManifest(id="M2", name="Male 2", tags=("male",)),
+        VoiceManifest(id="M3", name="Male 3", tags=("male",)),
+        VoiceManifest(id="M4", name="Male 4", tags=("male",)),
+        VoiceManifest(id="M5", name="Male 5", tags=("male",)),
+        VoiceManifest(id="F1", name="Female 1", tags=("female",)),
+        VoiceManifest(id="F2", name="Female 2", tags=("female",)),
+        VoiceManifest(id="F3", name="Female 3", tags=("female",)),
+        VoiceManifest(id="F4", name="Female 4", tags=("female",)),
+        VoiceManifest(id="F5", name="Female 5", tags=("female",)),
+    ),
+)
+
+MODEL_REQUIREMENTS: list[ModelManifest] = []
+
+
+def create_engine(
+    context: HostContext,
+    model_path: Path | None,
+    config: EngineConfig,
+) -> Engine:
+    """Create a SuperTonic engine instance.
+
+    This function is the plugin entry point. It must be atomic:
+    succeed fully or raise EngineError and clean up.
+
+    Args:
+        context: Host services (config dir, logger, http client).
+        model_path: Resolved model path, or None for default.
+        config: Engine initialization settings (device, etc.).
+
+    Returns:
+        A fully initialized SuperTonicEngine instance.
+
+    Raises:
+        EngineError: On failure. Cleans up partially created resources.
+    """
+    try:
+        pipeline = _load_supertonic_pipeline()
+        engine = SuperTonicEngine(pipeline)
+        return engine
+    except Exception as e:
+        from abogen.tts_plugin.errors import EngineError as EngineErrorClass
+        raise EngineErrorClass(f"Failed to create SuperTonic engine: {e}") from e

plugins/supertonic/engine.py

@@ -1,156 +1,125 @@
-"""SuperTonic Engine adapter for the TTS Plugin Architecture.
-
-This module adapts the existing SuperTonic backend to the new Engine/EngineSession
-protocol. It wraps the SupertonicPipeline without modifying it.
-"""
-
-from __future__ import annotations
-
-import io
-import logging
-from typing import Any, Optional
-
-import numpy as np
-
-from abogen.tts_plugin.capabilities import VoiceLister
-from abogen.tts_plugin.engine import Engine, EngineSession
-from abogen.tts_plugin.errors import EngineError, InvalidInputError
-from abogen.tts_plugin.manifest import VoiceManifest
-from abogen.tts_plugin.types import (
-    AudioFormat,
-    Duration,
-    ParameterValues,
-    SynthesisRequest,
-    SynthesizedAudio,
-    VoiceSelection,
-)
-
-logger = logging.getLogger(__name__)
-
-# SuperTonic voice list - source of truth
-_SUPERTONIC_VOICES = ("M1", "M2", "M3", "M4", "M5", "F1", "F2", "F3", "F4", "F5")
-
-# Voice display names mapping
-_VOICE_DISPLAY_NAMES: dict[str, str] = {
-    "M1": "Male 1",
-    "M2": "Male 2",
-    "M3": "Male 3",
-    "M4": "Male 4",
-    "M5": "Male 5",
-    "F1": "Female 1",
-    "F2": "Female 2",
-    "F3": "Female 3",
-    "F4": "Female 4",
-    "F5": "Female 5",
-}
-
-# Sample rate for SuperTonic audio
-_SUPERTONIC_SAMPLE_RATE = 24000
-
-
-class SuperTonicSession:
-    """EngineSession implementation for SuperTonic.
-
-    Owns mutable execution state for synthesis.
-    NOT thread-safe.
-    """
-
-    def __init__(self, pipeline: Any) -> None:
-        self._pipeline = pipeline
-        self._disposed = False
-
-    def synthesize(self, request: SynthesisRequest) -> SynthesizedAudio:
-        """Synthesize audio from text using SuperTonic."""
-        if self._disposed:
-            raise EngineError("Session disposed")
-
-        try:
-            import soundfile as sf
-
-            voice = request.voice.key
-            speed = float(request.parameters.values.get("speed", 1.0))
-            total_steps = request.parameters.values.get("total_steps", None)
-            split_pattern = request.parameters.values.get("split_pattern", None)
-
-            if total_steps is not None:
-                total_steps = int(total_steps)
-
-            audio_parts: list[np.ndarray] = []
-            for segment in self._pipeline(
-                request.text,
-                voice=voice,
-                speed=speed,
-                split_pattern=split_pattern,
-                total_steps=total_steps,
-            ):
-                audio_parts.append(segment.audio)
-
-            if not audio_parts:
-                return SynthesizedAudio(
-                    data=b"",
-                    format=AudioFormat(mime="audio/wav", extension="wav"),
-                    duration=Duration(seconds=0.0),
-                )
-
-            combined = np.concatenate(audio_parts).astype("float32", copy=False)
-            buf = io.BytesIO()
-            sf.write(buf, combined, self._pipeline.sample_rate, format="WAV")
-            audio_bytes = buf.getvalue()
-            duration_seconds = len(combined) / self._pipeline.sample_rate
-
-            return SynthesizedAudio(
-                data=audio_bytes,
-                format=AudioFormat(mime="audio/wav", extension="wav"),
-                duration=Duration(seconds=duration_seconds),
-            )
-        except EngineError:
-            raise
-        except Exception as e:
-            raise EngineError(f"Synthesis failed: {e}") from e
-
-    def dispose(self) -> None:
-        """Release session resources. Idempotent."""
-        self._disposed = True
-
-
-class SuperTonicEngine:
-    """Engine implementation for SuperTonic.
-
-    Factory for SuperTonicSession instances. Stateless and thread-safe.
-    """
-
-    def __init__(self, pipeline: Any) -> None:
-        self._pipeline = pipeline
-        self._disposed = False
-
-    def createSession(self) -> SuperTonicSession:
-        """Create a new SuperTonicSession."""
-        if self._disposed:
-            raise EngineError("Engine disposed")
-        return SuperTonicSession(self._pipeline)
-
-    def dispose(self) -> None:
-        """Release engine resources. Idempotent."""
-        self._disposed = True
-
-    def listVoices(self, sourceId: str) -> list[VoiceManifest]:
-        """List available SuperTonic voices. Implements VoiceLister capability."""
-        if self._disposed:
-            raise EngineError("Engine disposed")
-        return [
-            VoiceManifest(
-                id=voice_id,
-                name=_VOICE_DISPLAY_NAMES.get(voice_id, voice_id),
-                tags=(_get_gender_tag(voice_id),),
-            )
-            for voice_id in _SUPERTONIC_VOICES
-        ]
-
-
-def _get_gender_tag(voice_id: str) -> str:
-    """Extract gender tag from voice ID."""
-    if voice_id.startswith("M"):
-        return "male"
-    elif voice_id.startswith("F"):
-        return "female"
-    return "unknown"
+"""SuperTonic Engine adapter for the TTS Plugin Architecture.
+
+This module adapts the existing SuperTonic backend to the new Engine/EngineSession
+protocol. It wraps the SupertonicPipeline without modifying it.
+"""
+
+from __future__ import annotations
+
+import io
+import logging
+from typing import Any
+
+import numpy as np
+
+from abogen.tts_plugin.capabilities import VoiceLister
+from abogen.tts_plugin.engine import Engine, EngineSession
+from abogen.tts_plugin.errors import EngineError
+from abogen.tts_plugin.manifest import VoiceManifest
+from abogen.tts_plugin.types import (
+    AudioFormat,
+    Duration,
+    SynthesisRequest,
+    SynthesizedAudio,
+)
+
+logger = logging.getLogger(__name__)
+
+# Sample rate for SuperTonic audio
+_SUPERTONIC_SAMPLE_RATE = 24000
+
+
+class SuperTonicSession:
+    """EngineSession implementation for SuperTonic.
+
+    Owns mutable execution state for synthesis.
+    NOT thread-safe.
+    """
+
+    def __init__(self, pipeline: Any) -> None:
+        self._pipeline = pipeline
+        self._disposed = False
+
+    def synthesize(self, request: SynthesisRequest) -> SynthesizedAudio:
+        """Synthesize audio from text using SuperTonic."""
+        if self._disposed:
+            raise EngineError("Session disposed")
+
+        try:
+            import soundfile as sf
+
+            voice = request.voice.key
+            speed = float(request.parameters.values.get("speed", 1.0))
+            total_steps = request.parameters.values.get("total_steps", None)
+            split_pattern = request.parameters.values.get("split_pattern", None)
+
+            if total_steps is not None:
+                total_steps = int(total_steps)
+
+            audio_parts: list[np.ndarray] = []
+            for segment in self._pipeline(
+                request.text,
+                voice=voice,
+                speed=speed,
+                split_pattern=split_pattern,
+                total_steps=total_steps,
+            ):
+                audio_parts.append(segment.audio)
+
+            if not audio_parts:
+                return SynthesizedAudio(
+                    data=b"",
+                    format=AudioFormat(mime="audio/wav", extension="wav"),
+                    duration=Duration(seconds=0.0),
+                )
+
+            combined = np.concatenate(audio_parts).astype("float32", copy=False)
+            buf = io.BytesIO()
+            sf.write(buf, combined, self._pipeline.sample_rate, format="WAV")
+            audio_bytes = buf.getvalue()
+            duration_seconds = len(combined) / self._pipeline.sample_rate
+
+            return SynthesizedAudio(
+                data=audio_bytes,
+                format=AudioFormat(mime="audio/wav", extension="wav"),
+                duration=Duration(seconds=duration_seconds),
+            )
+        except EngineError:
+            raise
+        except Exception as e:
+            raise EngineError(f"Synthesis failed: {e}") from e
+
+    def dispose(self) -> None:
+        """Release session resources. Idempotent."""
+        self._disposed = True
+
+
+class SuperTonicEngine:
+    """Engine implementation for SuperTonic.
+
+    Factory for SuperTonicSession instances. Stateless and thread-safe.
+    """
+
+    def __init__(self, pipeline: Any) -> None:
+        self._pipeline = pipeline
+        self._disposed = False
+
+    def createSession(self) -> SuperTonicSession:
+        """Create a new SuperTonicSession."""
+        if self._disposed:
+            raise EngineError("Engine disposed")
+        return SuperTonicSession(self._pipeline)
+
+    def dispose(self) -> None:
+        """Release engine resources. Idempotent."""
+        self._disposed = True
+
+    def listVoices(self, sourceId: str) -> list[VoiceManifest]:
+        """List available SuperTonic voices. Implements VoiceLister capability.
+        
+        Note: Static voice catalog is declared in plugin manifest.
+        This method is retained for VoiceLister interface compliance.
+        """
+        if self._disposed:
+            raise EngineError("Engine disposed")
+        return []