[依赖、模型和配置]：完成本地语音模型准备，包含.env配置、模型下载脚本和model-check

src/owner_voice_pet/speech_models.py

@@ -0,0 +1,142 @@
+from __future__ import annotations
+
+import importlib.util
+import json
+from dataclasses import dataclass
+from pathlib import Path
+from typing import Any
+
+from .models import ErrorCode, ProviderError
+
+DEFAULT_VAD_URL = "https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/silero_vad.onnx"
+DEFAULT_STT_URL = (
+    "https://github.com/k2-fsa/sherpa-onnx/releases/download/asr-models/"
+    "sherpa-onnx-streaming-zipformer-zh-14M-2023-02-23.tar.bz2"
+)
+DEFAULT_STT_DIR = "sherpa-onnx-streaming-zipformer-zh-14M-2023-02-23"
+
+REQUIRED_MODEL_FILES = (
+    "vad/silero_vad.onnx",
+    f"stt/{DEFAULT_STT_DIR}/tokens.txt",
+    f"stt/{DEFAULT_STT_DIR}/encoder-epoch-99-avg-1.int8.onnx",
+    f"stt/{DEFAULT_STT_DIR}/decoder-epoch-99-avg-1.onnx",
+    f"stt/{DEFAULT_STT_DIR}/joiner-epoch-99-avg-1.int8.onnx",
+)
+
+
+@dataclass(frozen=True, slots=True)
+class SpeechModelStatus:
+    models_dir: Path
+    manifest_path: Path
+    manifest_present: bool
+    required_files: tuple[str, ...]
+    missing_files: tuple[str, ...]
+    sherpa_onnx_available: bool
+
+    @property
+    def ok(self) -> bool:
+        return not self.missing_files and self.sherpa_onnx_available
+
+    def to_json(self) -> dict[str, Any]:
+        return {
+            "ok": self.ok,
+            "models_dir": str(self.models_dir),
+            "manifest_path": str(self.manifest_path),
+            "manifest_present": self.manifest_present,
+            "required_files": list(self.required_files),
+            "missing_files": list(self.missing_files),
+            "sherpa_onnx_available": self.sherpa_onnx_available,
+        }
+
+
+def default_manifest() -> dict[str, Any]:
+    return {
+        "schema": "owner_voice_pet.speech_models.v1",
+        "sample_rate": 16000,
+        "sources": {
+            "vad": DEFAULT_VAD_URL,
+            "stt": DEFAULT_STT_URL,
+        },
+        "providers": {
+            "vad": {
+                "type": "silero-vad",
+                "path": "vad/silero_vad.onnx",
+            },
+            "stt": {
+                "type": "sherpa-onnx-streaming-transducer",
+                "model_dir": f"stt/{DEFAULT_STT_DIR}",
+                "tokens": f"stt/{DEFAULT_STT_DIR}/tokens.txt",
+                "encoder": f"stt/{DEFAULT_STT_DIR}/encoder-epoch-99-avg-1.int8.onnx",
+                "decoder": f"stt/{DEFAULT_STT_DIR}/decoder-epoch-99-avg-1.onnx",
+                "joiner": f"stt/{DEFAULT_STT_DIR}/joiner-epoch-99-avg-1.int8.onnx",
+            },
+        },
+        "required_files": list(REQUIRED_MODEL_FILES),
+    }
+
+
+def write_default_manifest(models_dir: str | Path) -> Path:
+    root = Path(models_dir)
+    root.mkdir(parents=True, exist_ok=True)
+    manifest_path = root / "manifest.json"
+    manifest_path.write_text(
+        json.dumps(default_manifest(), ensure_ascii=False, indent=2) + "\n",
+        encoding="utf-8",
+    )
+    return manifest_path
+
+
+def load_manifest(models_dir: str | Path) -> dict[str, Any]:
+    manifest_path = Path(models_dir) / "manifest.json"
+    if not manifest_path.exists():
+        return default_manifest()
+    return json.loads(manifest_path.read_text(encoding="utf-8"))
+
+
+def required_model_files(models_dir: str | Path) -> tuple[str, ...]:
+    manifest = load_manifest(models_dir)
+    files = manifest.get("required_files", list(REQUIRED_MODEL_FILES))
+    return tuple(str(item) for item in files)
+
+
+def check_speech_models(models_dir: str | Path, require_sherpa: bool = True) -> SpeechModelStatus:
+    root = Path(models_dir)
+    manifest_path = root / "manifest.json"
+    required = required_model_files(root)
+    missing = tuple(relative for relative in required if not (root / relative).exists())
+    sherpa_available = True
+    if require_sherpa:
+        sherpa_available = importlib.util.find_spec("sherpa_onnx") is not None
+    return SpeechModelStatus(
+        models_dir=root,
+        manifest_path=manifest_path,
+        manifest_present=manifest_path.exists(),
+        required_files=required,
+        missing_files=missing,
+        sherpa_onnx_available=sherpa_available,
+    )
+
+
+def model_status_errors(status: SpeechModelStatus) -> list[ProviderError]:
+    errors: list[ProviderError] = []
+    if not status.sherpa_onnx_available:
+        errors.append(
+            ProviderError(
+                ErrorCode.STT_TRANSCRIBE_FAILED,
+                "sherpa_onnx is not installed in the active Python environment",
+                False,
+                "sherpa-onnx",
+                "model-check",
+            )
+        )
+    if status.missing_files:
+        errors.append(
+            ProviderError(
+                ErrorCode.STT_MODEL_MISSING,
+                "missing speech model files: " + ", ".join(status.missing_files),
+                False,
+                "speech-models",
+                "model-check",
+            )
+        )
+    return errors