From 7d2c393118687562d6acd9564e4253d5166f92ea Mon Sep 17 00:00:00 2001
From: Abdessamad Derraz <3028866+Abdess@users.noreply.github.com>
Date: Fri, 6 Nov 2026 03:53:26 +0100
Subject: [PATCH] fix: provenance gaps read held names by index

profile_needs decided held-ness from each record's own name. np2pi's KANJI2.ROM, held as the PC-88 n88knj2.rom, was listed missing and paired with a Fujitsu dump. It now reads by_name as verify does, and an unsourceable entry is not held by a homonym, which lists xm7's KANJI.ROM.
---
 scripts/provenance_report.py | 10 +++++++++-
 tests/test_provenance.py     | 23 +++++++++++++++++++++++
 2 files changed, 32 insertions(+), 1 deletion(-)

diff --git a/scripts/provenance_report.py b/scripts/provenance_report.py
index 77d217143..a54c1b0a6 100644
--- a/scripts/provenance_report.py
+++ b/scripts/provenance_report.py
@@ -80,11 +80,17 @@ def profile_needs(profiles: dict, db: dict) -> dict:
         (_crc_key(str(f.get("crc32", "")).lower()), f.get("size"))
         for f in files.values() if f.get("crc32")
     }
+    # Held under its own name or under an alias the database indexes, as
+    # the resolver and verify read it: np2pi's KANJI2.ROM is the PC-88
+    # n88knj2.rom, and its own name alone listed it missing.
     sizes_by_name: dict[str, set[int]] = {}
     for record in files.values():
         sizes_by_name.setdefault(str(record.get("name", "")).casefold(), set()).add(
             record.get("size")
         )
+    for alias, sha1s in indexes.get("by_name", {}).items():
+        held = sizes_by_name.setdefault(str(alias).casefold(), set())
+        held.update(files[sha1].get("size") for sha1 in sha1s if sha1 in files)
 
     by_hash: dict[tuple, list[dict]] = {}
     by_name: dict[str, list[tuple[dict, dict]]] = {}
@@ -117,7 +123,9 @@ def profile_needs(profiles: dict, db: dict) -> dict:
             if all(entry.get(k) is None for k in ("size", "min_size", "max_size")):
                 continue
             names = [name, *(entry.get("aliases") or [])]
-            if any(
+            # An unsourceable entry is not held by a homonym: xm7's
+            # KANJI2.ROM is the Fujitsu font, the held one the PC-88's.
+            if not entry.get("unsourceable") and any(
                 _fits(held, entry)
                 for alias in names
                 for held in sizes_by_name.get(str(alias).casefold(), ())
diff --git a/tests/test_provenance.py b/tests/test_provenance.py
index a1ed98574..d76d3a9b4 100644
--- a/tests/test_provenance.py
+++ b/tests/test_provenance.py
@@ -417,6 +417,29 @@ class TestProvenanceReport(unittest.TestCase):
         )
 
 
+class HeldUnderAnAlias(unittest.TestCase):
+    """profile_needs read held names from each record's own name: np2pi's
+    KANJI2.ROM, which the database holds as the PC-88 n88knj2.rom, was
+    listed missing and paired with the Fujitsu FM77AV40SX dump."""
+
+    DB = {
+        "files": {"a" * 40: {"name": "n88knj2.rom", "size": 131072}},
+        "indexes": {"by_name": {"KANJI2.ROM": ["a" * 40], "n88knj2.rom": ["a" * 40]}},
+    }
+
+    def _needs(self, **fields):
+        from scripts.provenance_report import profile_needs
+
+        entry = {"name": "KANJI2.ROM", "size": 131072, **fields}
+        return profile_needs({"emu": {"files": [entry]}}, self.DB)["by_name"]
+
+    def test_an_alias_the_database_indexes_is_held(self):
+        self.assertEqual(self._needs(), {})
+
+    def test_an_unsourceable_entry_is_not_held_by_a_homonym(self):
+        self.assertIn("kanji2.rom", self._needs(unsourceable="another machine's font"))
+
+
 class TestProvenancePage(unittest.TestCase):
     def _page(self, extra_entries=None, provenance=True):
         db = {