From e97c1fbbcc00a8e1dd8430ed468d18835fddb107 Mon Sep 17 00:00:00 2001
From: Abdessamad Derraz <3028866+Abdess@users.noreply.github.com>
Date: Tue, 11 Aug 2026 14:40:33 +0200
Subject: [PATCH] fix: keep unsourceable files from matching homonyms

---
 scripts/common.py | 15 ++++++++++++---
 tests/test_e2e.py | 26 ++++++++++++++++++++++++++
 2 files changed, 38 insertions(+), 3 deletions(-)

diff --git a/scripts/common.py b/scripts/common.py
index 26502c7e..32405737 100644
--- a/scripts/common.py
+++ b/scripts/common.py
@@ -680,7 +680,16 @@ def resolve_local_file(
             file_entry, files_db.get(match_sha1, {}).get("size")
         )
 
-    if not has_strong_hash:
+    # An entry the profile documents as unsourceable is one the collection
+    # cannot hold: a font that only exists inside a paid package, cheat codes
+    # written per title. A file answering to its name is therefore a homonym
+    # from another system, not a stale dump of the same file - kanji.rom names
+    # an FPseNG font, an openMSX one and a 3DO ROM. Content still decides: the
+    # hash and path steps above run first, so collecting the real bytes makes
+    # the entry resolve.
+    unsourceable = bool(file_entry.get("unsourceable"))
+
+    if not has_strong_hash and not unsourceable:
         candidates = []
         for try_name in names_to_try:
             for match_sha1 in by_name.get(try_name, []):
@@ -751,7 +760,7 @@ def resolve_local_file(
                 primary = [p for p, _ in valid if "/.variants/" not in p]
                 return (primary[0] if primary else valid[0][0]), "hash_mismatch"
             # No candidate contains the zipped_file -fall through to step 5
-        else:
+        elif not unsourceable:
             primary = [p for p, _ in candidates if "/.variants/" not in p]
             return (primary[0] if primary else candidates[0][0]), "hash_mismatch"
 
@@ -853,7 +862,7 @@ def resolve_local_file(
                             return candidate, status
                         data_dir_mismatch = data_dir_mismatch or candidate
 
-    if data_dir_mismatch:
+    if data_dir_mismatch and not unsourceable:
         return data_dir_mismatch, "hash_mismatch"
 
     # Agnostic fallback: for filename-agnostic files, find any DB file
diff --git a/tests/test_e2e.py b/tests/test_e2e.py
index caa6efaf..85d23e64 100644
--- a/tests/test_e2e.py
+++ b/tests/test_e2e.py
@@ -793,6 +793,32 @@ class TestE2E(unittest.TestCase):
         path, status = resolve_local_file(entry, self.db)
         self.assertEqual(status, "hash_mismatch")
 
+    def test_07b_resolve_unsourceable_ignores_homonym(self):
+        """An unsourceable entry takes content, never a file of the same name."""
+        entry = {
+            "name": "wrong_hash.bin",
+            "md5": "ffffffffffffffffffffffffffffffff",
+            "unsourceable": "ships inside a paid application package",
+        }
+        path, status = resolve_local_file(entry, self.db)
+        self.assertIsNone(path)
+        self.assertEqual(status, "not_found")
+
+        entry = {"name": "no_md5.bin", "unsourceable": "written per title"}
+        path, status = resolve_local_file(entry, self.db)
+        self.assertIsNone(path)
+        self.assertEqual(status, "not_found")
+
+        # Collecting the real bytes still resolves it.
+        entry = {
+            "name": "present_req.bin",
+            "sha1": self.files["present_req.bin"]["sha1"],
+            "unsourceable": "ships inside a paid application package",
+        }
+        path, status = resolve_local_file(entry, self.db)
+        self.assertEqual(status, "sha1_exact")
+        self.assertIn("present_req.bin", path)
+
     def test_08_resolve_variants_deprioritized(self):
         entry = {"name": "present_req.bin"}
         path, status = resolve_local_file(entry, self.db)