Add X-VC voice conversion mode

platform/docker/profile-controller/profile_controller.py

@@ -34,6 +34,8 @@ SEPARATOR_LABEL_KEY = "com.mike-ai.stem-separator"
 SEPARATOR_WORKER = os.environ.get("SEPARATOR_WORKER", "").strip()
 VOICE_LABEL_KEY = "com.mike-ai.voice-worker"
 VOICE_WORKER = os.environ.get("VOICE_WORKER", "").strip()
+VOICE_CHANGE_LABEL_KEY = "com.mike-ai.voice-change-worker"
+VOICE_CHANGE_WORKER = os.environ.get("VOICE_CHANGE_WORKER", "").strip()
 LOCK = threading.Lock()
 log = logging.getLogger("profile-controller")
 
@@ -135,6 +137,17 @@ def voice_container() -> dict:
     return matches[0]
 
 
+def voice_change_container() -> dict:
+    if not VOICE_CHANGE_WORKER:
+        raise RuntimeError("voice-change worker is not configured")
+    matches = [item for item in labelled_containers(VOICE_CHANGE_LABEL_KEY)
+               if item.get("Labels", {}).get(VOICE_CHANGE_LABEL_KEY) == VOICE_CHANGE_WORKER]
+    if len(matches) != 1:
+        raise RuntimeError(
+            f"expected exactly one voice-change worker {VOICE_CHANGE_WORKER!r}, found {len(matches)}")
+    return matches[0]
+
+
 def stop_music_if_configured() -> None:
     if MUSIC_WORKER:
         stop_container(music_container(), timeout=30)
@@ -150,6 +163,11 @@ def stop_voice_if_configured() -> None:
         stop_container(voice_container(), timeout=30)
 
 
+def stop_voice_change_if_configured() -> None:
+    if VOICE_CHANGE_WORKER:
+        stop_container(voice_change_container(), timeout=30)
+
+
 def stop_container(item: dict, timeout: int = 120) -> None:
     if item.get("State") != "running":
         return
@@ -190,6 +208,7 @@ def set_image_worker(running: bool, kind: str = IMAGE_WORKER) -> dict:
             stop_music_if_configured()
             stop_separator_if_configured()
             stop_voice_if_configured()
+            stop_voice_change_if_configured()
             for other in image_containers():
                 if other["Id"] != item["Id"]:
                     stop_container(other, timeout=20)
@@ -217,6 +236,7 @@ def set_music_worker(running: bool) -> dict:
             stop_container(tts_container(), timeout=30)
             stop_separator_if_configured()
             stop_voice_if_configured()
+            stop_voice_change_if_configured()
             start_container(item)
         else:
             stop_container(item, timeout=30)
@@ -236,6 +256,7 @@ def set_separator_worker(running: bool) -> dict:
             stop_container(tts_container(), timeout=30)
             stop_music_if_configured()
             stop_voice_if_configured()
+            stop_voice_change_if_configured()
             start_container(item)
         else:
             stop_container(item, timeout=30)
@@ -244,7 +265,7 @@ def set_separator_worker(running: bool) -> dict:
 
 
 def set_voice_worker(running: bool) -> dict:
-    """Start Vevo2 exclusively, or stop it before LLM restoration."""
+    """Start OmniVoice exclusively, or stop it before LLM restoration."""
     with LOCK:
         item = voice_container()
         if running:
@@ -255,6 +276,7 @@ def set_voice_worker(running: bool) -> dict:
             stop_container(tts_container(), timeout=30)
             stop_music_if_configured()
             stop_separator_if_configured()
+            stop_voice_change_if_configured()
             start_container(item)
         else:
             stop_container(item, timeout=30)
@@ -262,6 +284,26 @@ def set_voice_worker(running: bool) -> dict:
                 "state": "running" if running else "stopped"}
 
 
+def set_voice_change_worker(running: bool) -> dict:
+    """Start X-VC exclusively, or stop it before another mode is loaded."""
+    with LOCK:
+        item = voice_change_container()
+        if running:
+            for profile_item in containers().values():
+                stop_container(profile_item)
+            for worker in image_containers():
+                stop_container(worker, timeout=20)
+            stop_container(tts_container(), timeout=30)
+            stop_music_if_configured()
+            stop_separator_if_configured()
+            stop_voice_if_configured()
+            start_container(item)
+        else:
+            stop_container(item, timeout=30)
+        return {"voice_change_worker": VOICE_CHANGE_WORKER,
+                "state": "running" if running else "stopped"}
+
+
 def active_profile(items: dict[str, dict] | None = None) -> str | None:
     items = items or containers()
     active = [name for name, item in items.items() if item.get("State") == "running"]
@@ -280,6 +322,7 @@ def activate(profile: str) -> dict:
         stop_music_if_configured()
         stop_separator_if_configured()
         stop_voice_if_configured()
+        stop_voice_change_if_configured()
         start_container(tts_container())
         items = containers()
         missing = [name for name in ALLOWED if name not in items]
@@ -365,6 +408,13 @@ class Handler(BaseHTTPRequestHandler):
                             "unhealthy" if "(unhealthy)" in voice_status else
                             "starting" if voice.get("State") == "running" else
                             "stopped")
+            voice_change = voice_change_container() if VOICE_CHANGE_WORKER else {}
+            voice_change_status = voice_change.get("Status", "")
+            voice_change_health = ("disabled" if not VOICE_CHANGE_WORKER else
+                                   "healthy" if "(healthy)" in voice_change_status else
+                                   "unhealthy" if "(unhealthy)" in voice_change_status else
+                                   "starting" if voice_change.get("State") == "running" else
+                                   "stopped")
             self.reply(200, {"active_profile": active_profile(items),
                              "music_worker": music.get("State", "disabled"),
                              "music_health": music_health,
@@ -372,6 +422,8 @@ class Handler(BaseHTTPRequestHandler):
                              "separator_health": separator_health,
                              "voice_worker": voice.get("State", "disabled"),
                              "voice_health": voice_health,
+                             "voice_change_worker": voice_change.get("State", "disabled"),
+                             "voice_change_health": voice_change_health,
                              "profiles": {name: items.get(name, {}).get(
                                  "State", "missing") for name in ALLOWED}})
         except Exception as exc:
@@ -410,6 +462,13 @@ class Handler(BaseHTTPRequestHandler):
                 log.exception("voice worker transition failed")
                 self.reply(503, {"error": str(exc)})
             return
+        if self.path in {"/workers/voice-change/start", "/workers/voice-change/stop"}:
+            try:
+                self.reply(200, set_voice_change_worker(self.path.endswith("/start")))
+            except Exception as exc:
+                log.exception("voice-change worker transition failed")
+                self.reply(503, {"error": str(exc)})
+            return
         worker_paths = {
             "/workers/image/start": (IMAGE_WORKER, True),
             "/workers/image/stop": (IMAGE_WORKER, False),