refactor: address PR #851 review round 1 (ingest transport)

- Add IngestTransport.active_consumer_count (0 by default; LocalTransport stores
  the started count) so app.py logs the worker count without re-checking
  INGEST_QUEUE — the last backend-knowledge leak in the lifespan is gone.
- Document that DistributedTransport is postgres/procrastinate-specific by design
  (aclose() calls ProcrastinateTaskProducer.drain()); other distributed backends
  would be separate IngestTransport subclasses.
- Clarify the _wire_vector_sync_state log line (writes app.state + singleton, not
  only the singleton).
- Strengthen the LocalTransport test: assert active_consumer_count transitions
  0→N and that each worker receives a distinct cloned receive stream.

Refs: Deck #196

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

nextcloud_mcp_server/app.py

@@ -377,7 +377,7 @@ def _wire_vector_sync_state(
     # app.state (Starlette) + the module singleton share the same attribute names.
     _apply(app.state)
     _apply(_vector_sync_state)
-    logger.info("Vector sync state stored in module singleton")
+    logger.info("Vector sync state published (app.state + module singleton)")
 
     # Also share with the mounted /app browser sub-app, if present.
     for route in app.routes:
@@ -1763,9 +1763,7 @@ def get_app(transport: str = "streamable-http", enabled_apps: list[str] | None =
 
                 logger.info(
                     "Background sync tasks started: 1 scanner + %s processors (queue=%s)",
-                    0
-                    if settings.ingest_queue == "postgres"
-                    else settings.vector_sync_processor_workers,
+                    transport.active_consumer_count,
                     settings.ingest_queue,
                 )
 
@@ -1969,9 +1967,7 @@ def get_app(transport: str = "streamable-http", enabled_apps: list[str] | None =
 
                     logger.info(
                         "Background sync tasks started: 1 user manager + %s processors (queue=%s)",
-                        0
-                        if settings.ingest_queue == "postgres"
-                        else settings.vector_sync_processor_workers,
+                        transport.active_consumer_count,
                         settings.ingest_queue,
                     )
 

nextcloud_mcp_server/vector/queue/transport.py

@@ -104,6 +104,17 @@ class IngestTransport(abc.ABC):
         reads procrastinate job counts via the producer instead)."""
         return None
 
+    @property
+    def active_consumer_count(self) -> int:
+        """In-process consumers started by :meth:`run_consumers` for this process.
+
+        ``0`` by default — distributed backends run their consumers as a separate
+        ``worker`` process. Lets the lifespan log the worker count without
+        re-inspecting ``INGEST_QUEUE`` (keeping the backend choice inside the
+        transport).
+        """
+        return 0
+
     async def run_consumers(
         self, task_group: TaskGroup, spawn_worker: SpawnWorker, count: int
     ) -> None:
@@ -139,6 +150,7 @@ class LocalTransport(IngestTransport):
         self._send_stream = send_stream
         self._receive_stream = receive_stream
         self._producer = MemoryTaskProducer(send_stream)
+        self._active_consumer_count = 0
 
     @property
     def producer(self) -> TaskProducer:
@@ -152,6 +164,10 @@ class LocalTransport(IngestTransport):
     def receive_stream(self) -> MemoryObjectReceiveStream[DocumentTask]:
         return self._receive_stream
 
+    @property
+    def active_consumer_count(self) -> int:
+        return self._active_consumer_count
+
     async def run_consumers(
         self, task_group: TaskGroup, spawn_worker: SpawnWorker, count: int
     ) -> None:
@@ -162,6 +178,7 @@ class LocalTransport(IngestTransport):
         # the prior inline lifespan behaviour.
         for i in range(count):
             await task_group.start(spawn_worker, i, self._receive_stream.clone())
+        self._active_consumer_count = count
 
 
 class DistributedTransport(IngestTransport):
@@ -171,6 +188,12 @@ class DistributedTransport(IngestTransport):
     by the lifespan; :meth:`run_consumers` is the inherited no-op (the
     ``nextcloud-mcp-server worker`` process drains the queue) and :meth:`aclose`
     closes the pool once on shutdown.
+
+    This adapter is postgres/procrastinate-specific by design: :meth:`aclose`
+    calls ``ProcrastinateTaskProducer.drain()`` (the narrow ``_producer`` type
+    confirms it). A different distributed backend (Redis/NATS/SQS) with its own
+    shutdown semantics would be a separate :class:`IngestTransport` subclass, not
+    a reconfiguration of this one.
     """
 
     def __init__(self, producer: ProcrastinateTaskProducer):