sync-v2 step 6: memory + connection bounds (§7, §4)

- concurrency.loop_semaphore: per-event-loop semaphores (safe across pytest loops
  and process restarts via a WeakKeyDictionary keyed by the running loop).
- provider: holds a per-account (≤4) and per-host (PER_HOST_CONN_CAP) slot for the
  whole lifetime of an IMAP socket, released on close; SMTP send takes a per-host
  slot for its duration. A worker can no longer open hundreds of sockets to one
  host/account.
- indexer: parsing goes through a PARSE_CONCURRENCY semaphore, and a message over
  STREAM_PARSE_MB is parsed from its blob file (parse_message_file) instead of an
  in-memory tree, so peak RAM ~= PARSE_CONCURRENCY x STREAM_PARSE_MB regardless of
  mailbox sizes. Backfill parallelism stays bounded by the bulk queue only.
- config: PER_HOST_CONN_CAP, PARSE_CONCURRENCY, STREAM_PARSE_MB.
- tests: per-account cap blocks the 5th connection / distinct accounts independent;
  file-vs-bytes parse equivalence + large-message streaming routing.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

backend/tests/providers/test_conn_limits.py

@@ -0,0 +1,58 @@
+"""Per-account / per-host connection caps (refactor §4).
+
+The provider holds a per-account (≤4) and per-host semaphore slot for the whole
+lifetime of an open socket, so a worker can never open hundreds of simultaneous
+connections to one account or host.
+"""
+
+from __future__ import annotations
+
+import asyncio
+
+import pytest
+
+from app.providers.imap_smtp.provider import _PER_ACCOUNT_CONN_CAP, ImapSmtpProvider, ProviderConfig
+
+
+def _cfg(username: str, host: str) -> ProviderConfig:
+    return ProviderConfig(
+        imap_host=host,
+        imap_port=993,
+        imap_tls="ssl",
+        smtp_host=host,
+        smtp_port=465,
+        smtp_tls="ssl",
+        username=username,
+        password="pw",
+    )
+
+
+async def test_per_account_connection_cap() -> None:
+    cfg = _cfg("user@example.com", "imap.example.com")
+    providers = [ImapSmtpProvider(cfg) for _ in range(_PER_ACCOUNT_CONN_CAP + 1)]
+
+    # Fill the account's slots (host cap defaults to 50, so account is the limiter).
+    for provider in providers[:_PER_ACCOUNT_CONN_CAP]:
+        await provider._acquire_conn_slots()
+
+    # The next acquire blocks until a slot frees.
+    with pytest.raises(TimeoutError):
+        await asyncio.wait_for(providers[-1]._acquire_conn_slots(), timeout=0.2)
+
+    providers[0]._release_conn_slots()
+    await asyncio.wait_for(providers[-1]._acquire_conn_slots(), timeout=0.5)  # now succeeds
+
+    for provider in providers[1:]:
+        provider._release_conn_slots()
+
+
+async def test_distinct_accounts_do_not_share_slots() -> None:
+    # Saturating account A must not block account B (independent semaphores).
+    a = [ImapSmtpProvider(_cfg("a@example.com", "imap.example.com")) for _ in range(4)]
+    for provider in a:
+        await provider._acquire_conn_slots()
+    b = ImapSmtpProvider(_cfg("b@example.com", "imap.example.com"))
+    await asyncio.wait_for(b._acquire_conn_slots(), timeout=0.5)
+    b._release_conn_slots()
+    for provider in a:
+        provider._release_conn_slots()