sorter: drop non-warhammer game systems (default: Middle-Earth)

The collection is warhammer-only; everything else stays in _unsorted
even when names match the catalog. The catalog snapshot keeps all
systems - the filter is a sorter default (DROPPED_SYSTEMS, applied in
load_catalog and after keyword classification), adjustable with
--drop-system/--keep-system.

dlphelper/sorter.py

@@ -32,9 +32,15 @@ UNSORTED = "_unsorted"
 UNCLASSIFIED_CLASS = "Unclassified"
 # A file whose system the keyword tables could not identify only gets a
 # catalog unit when the fuzzy hit is strong - otherwise it would wander
-# into other games' sections (Middle-earth, Blood Bowl, ...) on a guess.
+# into other games' sections (Blood Bowl, ...) on a guess.
 CROSS_SYSTEM_MIN = 0.75
 
+# Game systems whose sections the sorter must never place into. The
+# collection is warhammer-only: everything else (Middle-earth proxies,
+# ...) stays in _unsorted even when a name matches the catalog. The
+# catalog snapshot itself keeps all systems; the filter lives here.
+DROPPED_SYSTEMS = frozenset({"Middle-Earth"})
+
 # Weak acceptance tier: when nothing reached the normal threshold, a hit
 # at least this strong is accepted but flagged low confidence ("[w]" in
 # the plan, confidence field in the audit, listed in
@@ -178,9 +184,14 @@ def classify_item(item: SortItem) -> None:
 # faction group layer (GW naming, shared with warhammer_data)
 
 
-def load_catalog(path: Path) -> list[dict]:
+def load_catalog(path: Path,
+                 dropped: frozenset[str] | None = DROPPED_SYSTEMS) -> list[dict]:
+    """Load catalog entries, dropping non-collected game systems by default
+    (pass dropped=frozenset() to keep everything)."""
     data = json.loads(path.read_text(encoding="utf-8"))
     entries = data["entries"] if isinstance(data, dict) else data
+    if dropped:
+        entries = [e for e in entries if e.get("system") not in dropped]
     for e in entries:
         e["_clean"] = _clean_name(e["name"])
         e["_clean_aliases"] = [_clean_name(a) for a in e.get("aliases", [])]
@@ -672,6 +683,17 @@ def _display_name(unit: str) -> str:
 # plan / apply
 # ---------------------------------------------------------------------------
 
+def _drop_dropped_system(item: SortItem) -> None:
+    """Dropped systems (see DROPPED_SYSTEMS) count as unclassified: the
+    collection does not cover them, so their names must not place files
+    into those sections. Called after keyword classification, before
+    matching - the item can still be re-homed by other signals."""
+    if item.system in DROPPED_SYSTEMS:
+        item.system = "Unknown"
+        item.group = item.faction = UNSORTED
+        item.subfaction = None
+
+
 def sort_plan(
     files: list[Path],
     library: Path,
@@ -716,6 +738,7 @@ def sort_plan(
         if unit_name is None:
             item = inspect(f)
             classify_item(item)
+            _drop_dropped_system(item)
             match_catalog(item, catalog, threshold)
             if root is not None:
                 # manual trees: let the real ancestor dir names compete
@@ -741,6 +764,7 @@ def sort_plan(
                     clean=_clean_name(unit_name), inner_names=listing,
                 )
                 classify_item(probe)
+                _drop_dropped_system(probe)
                 match_catalog(probe, catalog, threshold)
                 fields = {k: getattr(probe, k) for k in (
                     "system", "group", "faction", "subfaction", "klass",