[独立唤醒与转写显示]：完成运行时本地唤醒分离，包含wake注入、转写输出和污染回归测试

openspec/changes/separate-wake-and-realtime-transcript/tasks.md

@@ -18,11 +18,11 @@
 
 ## 3. Runtime 独立唤醒与实时转写
 
-- [ ] 3.1 实现 `SherpaOnnxKeywordWakeWordProvider`；前置条件：KWS 路径 helper 完成；验收标准：可加载 KWS 模型，缺文件结构化失败；测试要点：missing/fake provider 测试；优先级：P0；预计：60 分钟。
+- [x] 3.1 实现 `SherpaOnnxKeywordWakeWordProvider`；前置条件：KWS 路径 helper 完成；验收标准：可加载 KWS 模型，缺文件结构化失败；测试要点：missing/fake provider 测试；优先级：P0；预计：60 分钟。
-- [ ] 3.2 修改 `LiveVoiceRuntime` 注入并使用 wake provider；前置条件：3.1 完成；验收标准：wake 阶段不调用 STT，唤醒命中后录正式问题；测试要点：两轮 runtime STT 调用次数；优先级：P0；预计：60 分钟。
+- [x] 3.2 修改 `LiveVoiceRuntime` 注入并使用 wake provider；前置条件：3.1 完成；验收标准：wake 阶段不调用 STT，唤醒命中后录正式问题；测试要点：两轮 runtime STT 调用次数；优先级：P0；预计：60 分钟。
-- [ ] 3.3 增加 `RuntimeReporter.transcript`；前置条件：3.2 完成；验收标准：转写结果在 LLM 前输出；测试要点：reporter 状态顺序；优先级：P0；预计：40 分钟。
+- [x] 3.3 增加 `RuntimeReporter.transcript`；前置条件：3.2 完成；验收标准：转写结果在 LLM 前输出；测试要点：reporter 状态顺序；优先级：P0；预计：40 分钟。
-- [ ] 3.4 增加唤醒词污染回归测试；前置条件：3.2 完成；验收标准：LLM user message 不含 wake 音频文本；测试要点：fake wake metadata；优先级：P0；预计：30 分钟。
+- [x] 3.4 增加唤醒词污染回归测试；前置条件：3.2 完成；验收标准：LLM user message 不含 wake 音频文本；测试要点：fake wake metadata；优先级：P0；预计：30 分钟。
-- [ ] 3.5 验证并提交“独立唤醒与转写显示”模块；前置条件：3.1 至 3.4 完成；验收标准：compileall、unittest、security-check、OpenSpec 通过后 commit；优先级：P0；预计：20 分钟。
+- [x] 3.5 验证并提交“独立唤醒与转写显示”模块；前置条件：3.1 至 3.4 完成；验收标准：compileall、unittest、security-check、OpenSpec 通过后 commit；优先级：P0；预计：20 分钟。
 
 ## 4. 文档、真实验收与归档
 

src/owner_voice_pet/runtime.py

@@ -8,17 +8,21 @@ from .config import AppConfig
 from .conversation import ConversationContext
 from .llm import OpenAICompatibleLlmProvider
 from .models import AudioSegment, ErrorCode, PipelineState, ProviderError
-from .protocols import AudioTransport, LlmProvider, SttProvider, TtsProvider
+from .protocols import AudioTransport, LlmProvider, SttProvider, TtsProvider, WakeWordProvider
 from .stt import CloudAsrSttProvider, SherpaOnnxSttProvider, is_valid_transcript_text
 from .transport import SoundDeviceAudioTransport
 from .tts import CloudTtsProvider, MacSayTtsProvider, SentenceBuffer
 from .vad import EnergyVadProvider, VadRecorder
+from .wakeword import SherpaOnnxKeywordWakeWordProvider
 
 
 class RuntimeReporter(Protocol):
     def status(self, state: str, message: str, *, turn_id: int | None = None) -> None:
         ...
 
+    def transcript(self, text: str, *, final: bool, turn_id: int | None = None) -> None:
+        ...
+
     def error(self, stage: str, code: str, message: str, *, turn_id: int | None = None) -> None:
         ...
 
@@ -28,6 +32,11 @@ class TerminalRuntimeReporter:
         prefix = f"[第{turn_id}轮] " if turn_id is not None else ""
         print(f"{prefix}{message}", flush=True)
 
+    def transcript(self, text: str, *, final: bool, turn_id: int | None = None) -> None:
+        prefix = f"[第{turn_id}轮] " if turn_id is not None else ""
+        label = "转写结果" if final else "转写中"
+        print(f"{prefix}{label}：{text}", flush=True)
+
     def error(self, stage: str, code: str, message: str, *, turn_id: int | None = None) -> None:
         prefix = f"[第{turn_id}轮] " if turn_id is not None else ""
         print(f"{prefix}{stage}失败：{code} {message}", file=sys.stderr, flush=True)
@@ -56,6 +65,7 @@ class LiveVoiceRuntime:
         *,
         config: AppConfig,
         transport: AudioTransport,
+        wakeword: WakeWordProvider,
         vad_recorder: VadRecorder,
         stt: SttProvider,
         llm: LlmProvider,
@@ -66,6 +76,7 @@ class LiveVoiceRuntime:
     ) -> None:
         self.config = config
         self.transport = transport
+        self.wakeword = wakeword
         self.vad_recorder = vad_recorder
         self.stt = stt
         self.llm = llm
@@ -76,6 +87,7 @@ class LiveVoiceRuntime:
         self._states: list[PipelineState] = []
 
     def load(self) -> None:
+        self.wakeword.load()
         self.vad_recorder.provider.load()
         self.stt.load()
         self.tts.load()
@@ -126,27 +138,10 @@ class LiveVoiceRuntime:
         self.transport.stop()
 
     def _wait_for_wake_and_user_text(self, turn_id: int) -> str | ProviderError:
-        while True:
+        wake_error = self._wait_for_local_wake(turn_id)
-            wake_segment = self._capture_segment(turn_id, state_message="待机：检测到语音，正在判断唤醒词")
+        if wake_error is not None:
-            if isinstance(wake_segment, ProviderError):
+            return wake_error
-                if wake_segment.code == ErrorCode.VAD_TIMEOUT_NO_SPEECH:
-                    self._state(PipelineState.WAKE_LISTENING, "待机：继续等待唤醒词“小杰小杰”", turn_id=turn_id)
-                    continue
-                return wake_segment
-            try:
-                wake_transcript = self.stt.transcribe(wake_segment).normalized_text
-            except ProviderError as exc:
-                if exc.code == ErrorCode.STT_EMPTY_TRANSCRIPT:
-                    self._state(PipelineState.WAKE_LISTENING, "恢复待机：未听清唤醒词", turn_id=turn_id)
-                    continue
-                raise
-            remainder = _text_after_wake_word(wake_transcript, self.config.wake_word)
-            if remainder is None:
-                self._state(PipelineState.WAKE_LISTENING, "恢复待机：未命中唤醒词", turn_id=turn_id)
-                continue
         self._state(PipelineState.SPEECH_DETECTING, "唤醒命中：请说出问题", turn_id=turn_id)
-            if is_valid_transcript_text(remainder):
-                return remainder
         user_segment = self._capture_segment(turn_id, state_message="录音中：正在听取问题")
         if isinstance(user_segment, ProviderError):
             return user_segment
@@ -161,8 +156,21 @@ class LiveVoiceRuntime:
                 "live-runtime",
                 "stt",
             )
+        self.reporter.transcript(user_text, final=True, turn_id=turn_id)
         return user_text
 
+    def _wait_for_local_wake(self, turn_id: int) -> ProviderError | None:
+        self.wakeword.reset()
+        while True:
+            frames = self.transport.read_frames(timeout_ms=100)
+            if not frames:
+                continue
+            for frame in frames:
+                event = self.wakeword.detect(frame)
+                if event is not None:
+                    self.wakeword.reset()
+                    return None
+
     def _capture_segment(self, turn_id: int, *, state_message: str) -> AudioSegment | ProviderError:
         self.vad_recorder.reset()
         self.vad_recorder.provider.reset()
@@ -180,7 +188,7 @@ class LiveVoiceRuntime:
 
     def _reply_to_user(self, user_text: str, turn_id: int) -> TurnResult:
         self.context.append_user(user_text)
-        self._state(PipelineState.THINKING, f"思考中：{user_text}", turn_id=turn_id)
+        self._state(PipelineState.THINKING, "思考中：正在生成回复", turn_id=turn_id)
         assistant_text = ""
         try:
             for delta in self.llm.stream_reply(self.context.build_llm_messages()):
@@ -237,6 +245,13 @@ def build_live_runtime(config: AppConfig, reporter: RuntimeReporter | None = Non
     return LiveVoiceRuntime(
         config=config,
         transport=SoundDeviceAudioTransport(output_device=config.audio_output_device),
+        wakeword=SherpaOnnxKeywordWakeWordProvider(
+            config.speech_models_dir,
+            keyword=config.wake_word,
+            keywords_file=config.wake_keywords_file,
+            threshold=config.wake_kws_threshold,
+            score=config.wake_kws_score,
+        ),
         vad_recorder=VadRecorder(EnergyVadProvider(), min_duration_ms=300, end_silence_ms=500, no_speech_timeout_ms=8000),
         stt=stt,
         llm=OpenAICompatibleLlmProvider(config),
@@ -247,21 +262,3 @@ def build_live_runtime(config: AppConfig, reporter: RuntimeReporter | None = Non
         ),
         reporter=reporter,
     )
-
-
-def _text_after_wake_word(text: str, wake_word: str) -> str | None:
-    compact_text = _compact(text)
-    compact_wake = _compact(wake_word)
-    index = compact_text.find(compact_wake)
-    if index < 0:
-        return None
-    end = index + len(compact_wake)
-    compact_remainder = compact_text[end:].strip("，,。.!！？? ")
-    if not compact_remainder:
-        return ""
-    original = text.replace(" ", "")
-    return original[-len(compact_remainder) :]
-
-
-def _compact(text: str) -> str:
-    return "".join(ch for ch in text.strip() if not ch.isspace())

tests/test_live_runtime.py

@@ -10,6 +10,7 @@ from owner_voice_pet.runtime import LiveVoiceRuntime
 from owner_voice_pet.transport import MemoryAudioTransport
 from owner_voice_pet.tts import SineTtsProvider
 from owner_voice_pet.vad import EnergyVadProvider, VadRecorder
+from owner_voice_pet.wakeword import KeywordWakeWordProvider
 
 
 def segment_frames(start_id: int, start_ms: int) -> list[AudioFrame]:
@@ -21,6 +22,17 @@ def segment_frames(start_id: int, start_ms: int) -> list[AudioFrame]:
     ]
 
 
+def wake_frame(frame_id: int, timestamp_ms: int) -> AudioFrame:
+    return AudioFrame(
+        b"\xff\x7f",
+        16000,
+        1,
+        timestamp_ms,
+        frame_id,
+        {"duration_ms": 20, "wake_word": "小杰小杰", "wake_confidence": 0.95},
+    )
+
+
 class QueueSttProvider:
     def __init__(self, texts: list[str]) -> None:
         self.texts = list(texts)
@@ -39,10 +51,17 @@ class QueueSttProvider:
 class RecordingReporter:
     def __init__(self) -> None:
         self.statuses: list[str] = []
+        self.transcripts: list[str] = []
         self.errors: list[str] = []
+        self.events: list[str] = []
 
     def status(self, state: str, message: str, *, turn_id: int | None = None) -> None:
         self.statuses.append(message)
+        self.events.append(f"status:{message}")
+
+    def transcript(self, text: str, *, final: bool, turn_id: int | None = None) -> None:
+        self.transcripts.append(text)
+        self.events.append(f"transcript:{text}")
 
     def error(self, stage: str, code: str, message: str, *, turn_id: int | None = None) -> None:
         self.errors.append(f"{stage}:{code}:{message}")
@@ -50,8 +69,11 @@ class RecordingReporter:
 
 def make_runtime(texts: list[str], context: ConversationContext | None = None) -> tuple[LiveVoiceRuntime, QueueSttProvider, MockLlmProvider, MemoryAudioTransport, RecordingReporter]:
     frames = []
-    for idx in range(4):
+    for idx, _text in enumerate(texts):
-        frames.extend(segment_frames(idx * 4, idx * 80))
+        base_id = idx * 5
+        base_ms = idx * 120
+        frames.append(wake_frame(base_id, base_ms))
+        frames.extend(segment_frames(base_id + 1, base_ms + 20))
     transport = MemoryAudioTransport(frames)
     stt = QueueSttProvider(texts)
     llm = MockLlmProvider(["这是答复。"])
@@ -60,6 +82,7 @@ def make_runtime(texts: list[str], context: ConversationContext | None = None) -
     runtime = LiveVoiceRuntime(
         config=AppConfig(llm_api_key="secret", speech_provider="cloud"),
         transport=transport,
+        wakeword=KeywordWakeWordProvider(),
         vad_recorder=VadRecorder(EnergyVadProvider(), min_duration_ms=40, end_silence_ms=40),
         stt=stt,
         llm=llm,
@@ -72,19 +95,18 @@ def make_runtime(texts: list[str], context: ConversationContext | None = None) -
 
 class LiveRuntimeTests(unittest.TestCase):
     def test_repeated_runtime_runs_two_turns_and_returns_to_standby(self) -> None:
-        runtime, stt, llm, transport, reporter = make_runtime(
+        runtime, stt, llm, transport, reporter = make_runtime(["第一问", "第二问"])
-            ["小杰小杰", "第一问", "小杰小杰", "第二问"]
-        )
         summary = runtime.run(max_turns=2)
 
         self.assertEqual(summary.completed_turns, 2)
-        self.assertEqual(len(stt.calls), 4)
+        self.assertEqual(len(stt.calls), 2)
         self.assertEqual(len(llm.calls), 2)
         self.assertEqual(len(transport.played_segments), 2)
+        self.assertEqual(reporter.transcripts, ["第一问", "第二问"])
         self.assertIn("恢复待机：可继续唤醒", reporter.statuses[-1])
 
     def test_temporary_context_is_sent_to_second_llm_call(self) -> None:
-        runtime, _, llm, _, _ = make_runtime(["小杰小杰", "第一问", "小杰小杰", "第二问"])
+        runtime, _, llm, _, _ = make_runtime(["第一问", "第二问"])
         runtime.run(max_turns=2)
 
         second_call_text = [message.content for message in llm.calls[1]]
@@ -94,14 +116,31 @@ class LiveRuntimeTests(unittest.TestCase):
 
     def test_new_runtime_context_starts_empty(self) -> None:
         first_context = ConversationContext()
-        first_runtime, _, _, _, _ = make_runtime(["小杰小杰", "第一问"], context=first_context)
+        first_runtime, _, _, _, _ = make_runtime(["第一问"], context=first_context)
         first_runtime.run(max_turns=1)
         self.assertGreater(len(first_context.messages()), 0)
 
         second_context = ConversationContext()
-        make_runtime(["小杰小杰", "第二问"], context=second_context)
+        make_runtime(["第二问"], context=second_context)
         self.assertEqual(second_context.messages(), ())
 
+    def test_transcript_is_reported_before_llm_thinking(self) -> None:
+        runtime, _, _, _, reporter = make_runtime(["第一问"])
+        runtime.run(max_turns=1)
+
+        transcript_index = reporter.events.index("transcript:第一问")
+        thinking_index = next(
+            index for index, event in enumerate(reporter.events) if event == "status:思考中：正在生成回复"
+        )
+        self.assertLess(transcript_index, thinking_index)
+
+    def test_wake_keyword_does_not_pollute_llm_user_message(self) -> None:
+        runtime, _, llm, _, _ = make_runtime(["第一问"])
+        runtime.run(max_turns=1)
+
+        self.assertEqual(llm.calls[0][-1].content, "第一问")
+        self.assertNotIn("小杰小杰", llm.calls[0][-1].content)
+
 
 if __name__ == "__main__":
     unittest.main()