Merge pull request #4306 from YFortin/fix/azure-tts-last-word-race

fix(azure-tts): Route completion through word boundary queue to prevent last word from being missed

changelog/4306.fixed.md

@@ -0,0 +1 @@
+- Fixed Azure TTS last word being missed by observers and RTVI UI. The completion signal was racing with word timestamp processing, causing the final word's `TTSTextFrame` to arrive after `TTSStoppedFrame`. Completion is now routed through the word boundary queue to ensure all words are processed before signaling stream end.

src/pipecat/services/azure/tts.py

@@ -540,14 +540,25 @@ class AzureTTSService(TTSService, AzureBaseTTSService):
         self._last_timestamp = timestamp
 
     async def _word_processor_task_handler(self):
-        """Process word timestamps from the queue and call add_word_timestamps."""
+        """Process word timestamps from the queue and call add_word_timestamps.
+
+        Also handles a None sentinel from _handle_completed: once all pending
+        words have been drained, it signals audio stream completion via
+        _audio_queue so that run_tts exits only after the last word has been
+        processed.
+        """
         while True:
             try:
-                word, timestamp_seconds = await self._word_boundary_queue.get()
+                item = await self._word_boundary_queue.get()
-                if self._current_context_id:
+                if item is None:
-                    await self.add_word_timestamps(
+                    # All words drained — now signal audio completion.
-                        [(word, timestamp_seconds)], self._current_context_id
+                    self._audio_queue.put_nowait(None)
-                    )
+                else:
+                    word, timestamp_seconds = item
+                    if self._current_context_id:
+                        await self.add_word_timestamps(
+                            [(word, timestamp_seconds)], self._current_context_id
+                        )
                 self._word_boundary_queue.task_done()
             except asyncio.CancelledError:
                 break
@@ -569,17 +580,21 @@ class AzureTTSService(TTSService, AzureBaseTTSService):
         Args:
             evt: Completion event from Azure Speech SDK.
         """
+        # Store duration for cumulative offset calculation
+        if evt.result and evt.result.audio_duration:
+            self._current_sentence_duration = evt.result.audio_duration.total_seconds()
+
         # Flush any pending word before completing
         if self._last_word is not None:
             self._word_boundary_queue.put_nowait((self._last_word, self._last_timestamp))
             self._last_word = None
             self._last_timestamp = None
 
-        # Store duration for cumulative offset calculation
+        # Route completion through the word boundary queue so the word processor
-        if evt.result and evt.result.audio_duration:
+        # task drains all pending words before signaling audio stream completion.
-            self._current_sentence_duration = evt.result.audio_duration.total_seconds()
+        # Without this, the last word's TTSTextFrame may arrive after
-
+        # TTSStoppedFrame, causing it to be missed by observers and the UI.
-        self._audio_queue.put_nowait(None)  # Signal completion
+        self._word_boundary_queue.put_nowait(None)
 
     def _handle_canceled(self, evt):
         """Handle synthesis cancellation.