TTS to only use stream entity method when streaming request comes in (#145167)

Co-authored-by: Franck Nijhof <git@frenck.dev>

homeassistant/components/tts/__init__.py

@@ -852,12 +852,9 @@ class SpeechManager:
         else:
             _LOGGER.debug("Generating audio for %s", message[0:32])
 
-            async def message_stream() -> AsyncGenerator[str]:
-                yield message
-
             extension = options.get(ATTR_PREFERRED_FORMAT, _DEFAULT_FORMAT)
             data_gen = self._async_generate_tts_audio(
-                engine_instance, message_stream(), language, options
+                engine_instance, message, language, options
             )
 
         cache = TTSCache(
@@ -931,7 +928,7 @@ class SpeechManager:
     async def _async_generate_tts_audio(
         self,
         engine_instance: TextToSpeechEntity | Provider,
-        message_stream: AsyncGenerator[str],
+        message_or_stream: str | AsyncGenerator[str],
         language: str,
         options: dict[str, Any],
     ) -> AsyncGenerator[bytes]:
@@ -979,9 +976,12 @@ class SpeechManager:
         if engine_instance.name is None or engine_instance.name is UNDEFINED:
             raise HomeAssistantError("TTS engine name is not set.")
 
-        if isinstance(engine_instance, Provider):
+        if isinstance(engine_instance, Provider) or isinstance(message_or_stream, str):
-            message = "".join([chunk async for chunk in message_stream])
+            if isinstance(message_or_stream, str):
-            extension, data = await engine_instance.async_get_tts_audio(
+                message = message_or_stream
+            else:
+                message = "".join([chunk async for chunk in message_or_stream])
+            extension, data = await engine_instance.async_internal_get_tts_audio(
                 message, language, options
             )
 
@@ -997,7 +997,7 @@ class SpeechManager:
 
         else:
             tts_result = await engine_instance.internal_async_stream_tts_audio(
-                TTSAudioRequest(language, options, message_stream)
+                TTSAudioRequest(language, options, message_or_stream)
             )
             extension = tts_result.extension
             data_gen = tts_result.data_gen

homeassistant/components/tts/entity.py

@@ -165,6 +165,18 @@ class TextToSpeechEntity(RestoreEntity, cached_properties=CACHED_PROPERTIES_WITH
         self.async_write_ha_state()
         return await self.async_stream_tts_audio(request)
 
+    @final
+    async def async_internal_get_tts_audio(
+        self, message: str, language: str, options: dict[str, Any]
+    ) -> TtsAudioType:
+        """Load tts audio file from the engine and update state.
+
+        Return a tuple of file extension and data as bytes.
+        """
+        self.__last_tts_loaded = dt_util.utcnow().isoformat()
+        self.async_write_ha_state()
+        return await self.async_get_tts_audio(message, language, options=options)
+
     async def async_stream_tts_audio(
         self, request: TTSAudioRequest
     ) -> TTSAudioResponse:

homeassistant/components/tts/legacy.py

@@ -7,7 +7,7 @@ from collections.abc import Coroutine, Mapping
 from functools import partial
 import logging
 from pathlib import Path
-from typing import TYPE_CHECKING, Any
+from typing import TYPE_CHECKING, Any, final
 
 import voluptuous as vol
 
@@ -252,3 +252,15 @@ class Provider:
         return await self.hass.async_add_executor_job(
             partial(self.get_tts_audio, message, language, options=options)
         )
+
+    @final
+    async def async_internal_get_tts_audio(
+        self, message: str, language: str, options: dict[str, Any]
+    ) -> TtsAudioType:
+        """Load tts audio file from provider.
+
+        Proxies request to mimic the entity interface.
+
+        Return a tuple of file extension and data as bytes.
+        """
+        return await self.async_get_tts_audio(message, language, options)

tests/components/assist_pipeline/test_pipeline.py

@@ -1627,25 +1627,15 @@ async def test_chat_log_tts_streaming(
         ),
     )
 
-    received_tts = []
+    async def async_get_tts_audio(
-
+        message: str,
-    async def async_stream_tts_audio(
+        language: str,
-        request: tts.TTSAudioRequest,
+        options: dict[str, Any] | None = None,
     ) -> tts.TTSAudioResponse:
-        """Mock stream TTS audio."""
+        """Mock get TTS audio."""
+        return ("mp3", b"".join([chunk.encode() for chunk in to_stream_tts]))
 
-        async def gen_data():
+    mock_tts_entity.async_get_tts_audio = async_get_tts_audio
-            async for msg in request.message_gen:
-                received_tts.append(msg)
-                yield msg.encode()
-
-        return tts.TTSAudioResponse(
-            extension="mp3",
-            data_gen=gen_data(),
-        )
-
-    mock_tts_entity.async_stream_tts_audio = async_stream_tts_audio
-    mock_tts_entity.async_supports_streaming_input = Mock(return_value=True)
 
     with patch(
         "homeassistant.components.assist_pipeline.pipeline.conversation.async_get_agent_info",
@@ -1717,7 +1707,6 @@ async def test_chat_log_tts_streaming(
 
     streamed_text = "".join(to_stream_tts)
     assert tts_result == streamed_text
-    assert len(received_tts) == expected_chunks
+    assert expected_chunks == 1
-    assert "".join(received_tts) == streamed_text
 
     assert process_events(events) == snapshot