fix: look past another emulator's copy by name

xm6typeg took XM6 Utility's synthesized SCSIINROM.DAT while the X68000 dump sat under another casing; blueMSX took a 16 KB RS232.ROM once msx-emu's 8 KB copy was set aside, though data/bluemsx held its own.

scripts/common.py

@@ -1364,11 +1364,14 @@ def resolve_local_file(
                         and path not in candidates
                     ):
                         candidates.append(path)
-        if not candidates:
+        owners = _owner_tokens(file_entry)
+        if not candidates or all(_foreign_tree(p, owners) for p in candidates):
             # Case-insensitive fallback: upstream platforms disagree on
             # casing (VEC_MineStorm.vec vs VEC_Minestorm.vec). Two casings
             # can name two contents (FONT.ROM and font.rom), so this only
-            # runs when no casing matches exactly.
+            # runs when no casing matches exactly, or when only another
+            # emulator's tree does: xm6typeg got XM6 Utility's synthesized
+            # SCSIINROM.DAT while the X68000 dump sat as scsiinrom.dat.
             folded = _casefold_name_index(by_name)
             for try_name in names_to_try:
                 for match_sha1 in folded.get(try_name.casefold(), []):
@@ -1385,21 +1388,23 @@ def resolve_local_file(
             elif len(candidates) > 1 and name.casefold().endswith(".zip"):
                 candidates = _covering_archives(candidates)
             candidates = _by_affinity(candidates, file_entry, dest_hint)
+            best = candidates[0] if candidates else ""
             # A copy outside another emulator's tree comes first: the
             # affinity read pcsx2 in bios/Other/armsx2/pcsx2/ and served
             # upstream PCSX2 the patches.zip of ARMSX2.
-            owners = _owner_tokens(file_entry)
             candidates = [
                 p for p in candidates if not _foreign_tree(p, owners)
             ] or candidates
             primary = [p for p in candidates if "/.variants/" not in p]
             if primary or candidates:
                 chosen = primary[0] if primary else candidates[0]
-                # Even the best of them is another emulator's own copy: the
-                # name says nothing about this profile's file, and a data
-                # directory it reads may hold it. Dolphin got Dolphin-MMJR's
-                # codehandler.bin while data/dolphin-sys held its own.
-                if data_dir_registry and _foreign_tree(chosen, _owner_tokens(file_entry)):
+                # The best of them is another emulator's own copy, served or
+                # set aside: the name says nothing about this profile's file,
+                # and a data directory it reads may hold it. Dolphin got
+                # Dolphin-MMJR's codehandler.bin while data/dolphin-sys held
+                # its own; blueMSX got the 16 KB RS232.ROM once msx-emu's
+                # 8 KB copy was set aside, while data/bluemsx held its own.
+                if data_dir_registry and _foreign_tree(best, owners):
                     hit, _mismatch = _resolve_in_data_dirs(
                         names_to_try, data_dir_registry,
                         functools.partial(

tests/test_foreign_tree_names.py

@@ -0,0 +1,78 @@
+"""The name step does not settle on another emulator's copy.
+
+xm6typeg asked for SCSIINROM.DAT: only XM6 Utility's synthesized image
+matched the casing, under bios/Other/xm6pro68k/, and the X68000 dump stored
+as scsiinrom.dat was never a candidate. blueMSX asked for RS232.ROM: its own
+8 KB copy was msx-emu's, set aside as foreign, and the 16 KB file left was
+served while data/bluemsx held the 8 KB one.
+"""
+
+from __future__ import annotations
+
+import hashlib
+import os
+import shutil
+import sys
+import tempfile
+import unittest
+from pathlib import Path
+
+REPO_ROOT = Path(__file__).resolve().parents[1]
+sys.path.insert(0, str(REPO_ROOT / "scripts"))
+
+from common import resolve_local_file  # noqa: E402
+
+
+class ForeignTreeNames(unittest.TestCase):
+    def setUp(self):
+        (REPO_ROOT / "tmp").mkdir(exist_ok=True)
+        self.root = Path(tempfile.mkdtemp(dir=REPO_ROOT / "tmp"))
+        self.addCleanup(shutil.rmtree, self.root, True)
+        # The collection's paths are relative to the repository root.
+        self.addCleanup(os.chdir, os.getcwd())
+        os.chdir(self.root)
+        self.db = {"files": {}, "indexes": {
+            "by_md5": {}, "by_name": {}, "by_crc32": {}, "by_path_suffix": {},
+        }}
+
+    def _hold(self, rel: str, data: bytes) -> str:
+        target = self.root / rel
+        target.parent.mkdir(parents=True, exist_ok=True)
+        target.write_bytes(data)
+        sha1 = hashlib.sha1(data).hexdigest()
+        name = rel.rsplit("/", 1)[-1]
+        if "/.variants/" in rel:
+            name = name.rsplit(".", 1)[0]
+        self.db["files"][sha1] = {
+            "path": rel, "name": name, "size": len(data),
+            "md5": hashlib.md5(data).hexdigest(), "crc32": "",
+        }
+        self.db["indexes"]["by_name"].setdefault(name, []).append(sha1)
+        return rel
+
+    def test_a_casing_only_another_tree_holds(self):
+        self._hold("bios/Other/xm6pro68k/.variants/SCSIINROM.DAT.47404fbb", b"X" * 64)
+        own = self._hold("bios/Sharp/X68000/scsiinrom.dat", b"D" * 64)
+        entry = {"name": "SCSIINROM.DAT", "source_emulator": "xm6typeg"}
+        self.assertEqual(resolve_local_file(entry, self.db)[0], own)
+
+    def test_a_set_aside_copy_sends_the_name_to_the_data_directory(self):
+        self._hold("bios/Microsoft/MSX/RS232.ROM", b"L" * 128)
+        self._hold("bios/Other/msx-emu/Machines/Shared Roms/RS232.ROM", b"S" * 64)
+        cache = self.root / "data" / "bluemsx"
+        (cache / "Machines" / "Shared Roms").mkdir(parents=True)
+        (cache / "Machines" / "Shared Roms" / "RS232.ROM").write_bytes(b"S" * 64)
+        entry = {
+            "name": "RS232.ROM", "path": "Machines/Shared Roms/RS232.ROM",
+            "size": 64, "source_emulator": "bluemsx",
+        }
+        path, status = resolve_local_file(
+            entry, self.db,
+            data_dir_registry={"bluemsx": {"local_cache": str(cache)}},
+        )
+        self.assertEqual(status, "data_dir")
+        self.assertEqual(Path(path).stat().st_size, 64)
+
+
+if __name__ == "__main__":
+    unittest.main()