feat: a platform package can differ from upstream

bundled describes the emulator's own release, and a platform package can differ both ways: Batocera's hcl package installs the binary alone and sends users to fetch the data, while its blueMSX core package copies the machines and databases beside the core. not_bundled_on and installed_on name those platforms, and the builder reads them through one rule.

schemas/emulator.schema.json

@@ -276,6 +276,16 @@
             "type": "boolean",
             "description": "Installed with the emulator for the build mode names, so packs leave the tree out."
           },
+          "installed_on": {
+            "type": "array",
+            "items": {"type": "string"},
+            "description": "Platforms (registry keys) whose own package installs this beside the emulator, so their packs leave it out."
+          },
+          "not_bundled_on": {
+            "type": "array",
+            "items": {"type": "string"},
+            "description": "Platforms (registry keys) whose package leaves out a file the emulator's release bundles, so their packs bring it."
+          },
           "source_ref": {
             "type": "string"
           }
@@ -450,6 +460,16 @@
             "type": "boolean",
             "description": "Installed with the emulator for the build mode names (all builds when mode is absent): already on the user's machine, so packs and install manifests leave it out."
           },
+          "installed_on": {
+            "type": "array",
+            "items": {"type": "string"},
+            "description": "Platforms (registry keys) whose own package installs this beside the emulator, so their packs leave it out."
+          },
+          "not_bundled_on": {
+            "type": "array",
+            "items": {"type": "string"},
+            "description": "Platforms (registry keys) whose package leaves out a file the emulator's release bundles, so their packs bring it."
+          },
           "config_key": {
             "type": "string"
           },

scripts/common.py

@@ -353,6 +353,9 @@ def load_platform_config(platform_name: str, platforms_dir: str = "platforms") -
             for sys_id, system in config.get("systems", {}).items()
         }
 
+    # The registry key, which per-platform profile fields (installed_on,
+    # not_bundled_on) name: the display name differs (MiSTer FPGA).
+    config["_key"] = platform_name
     _platform_config_cache[cache_key] = config
     return config
 

scripts/validation.py

@@ -525,7 +525,21 @@ def filter_files_by_mode(files: list[dict], standalone: bool) -> list[dict]:
     return result
 
 
-def deliverable(entries: list[dict], standalone: bool) -> list[dict]:
+def installed_with_emulator(entry: dict, platform: str = "") -> bool:
+    """Whether the emulator as this platform installs it already brings the entry.
+
+    `bundled` describes the emulator's own release. A platform package can
+    differ both ways: Batocera's hcl package installs the binary alone and
+    sends users to fetch the data (`not_bundled_on`), while its blueMSX core
+    package copies the machines and databases beside the core
+    (`installed_on`).
+    """
+    if platform and platform in (entry.get("installed_on") or []):
+        return True
+    return bool(entry.get("bundled")) and platform not in (entry.get("not_bundled_on") or [])
+
+
+def deliverable(entries: list[dict], standalone: bool, platform: str = "") -> list[dict]:
     """The entries of one build a pack has to bring.
 
     A file or tree installed with the emulator is already on the user's
@@ -533,7 +547,10 @@ def deliverable(entries: list[dict], standalone: bool) -> list[dict]:
     `mode` like any entry, so PPSSPP's assets, shipped by the standalone
     release and not by the libretro core, still reach a RetroArch pack.
     """
-    return [e for e in filter_files_by_mode(entries, standalone) if not e.get("bundled")]
+    return [
+        e for e in filter_files_by_mode(entries, standalone)
+        if not installed_with_emulator(e, platform)
+    ]
 
 
 def find_validated_variant(

scripts/verify.py

@@ -96,6 +96,7 @@ from validation import (
     check_file_validation,
     deliverable,
     filter_files_by_mode,
+    installed_with_emulator,
     outside_gap_scope,
 )
 
@@ -650,7 +651,7 @@ def find_undeclared_files(
                     "path": dest,
                     "required": f.get("required", False),
                     "hle_fallback": f.get("hle_fallback", False),
-                    "bundled": f.get("bundled", False),
+                    "bundled": installed_with_emulator(f, config.get("_key", "")),
                     "category": f.get("category", "bios"),
                     "in_repo": in_repo,
                     "note": f.get("note", ""),

tests/test_agnostic_scope.py

@@ -80,7 +80,9 @@ class BundledHashIsProof(unittest.TestCase):
         wrong = []
         for name, profile in sorted(load_emulator_profiles("emulators", skip_aliases=False).items()):
             for entry in profile.get("files") or []:
-                if not entry.get("bundled") or entry.get("unsourceable"):
+                # A bundled entry is packed only where a platform's package
+                # leaves it out: there the copy has to be the declared bytes.
+                if not entry.get("not_bundled_on") or entry.get("unsourceable"):
                     continue
                 if not any(entry.get(k) for k in ("sha1", "md5", "sha256", "crc32")):
                     continue

tests/test_audit_regressions.py

@@ -4579,6 +4579,36 @@ class BundledFilesAreNotToSource(unittest.TestCase):
                          ["ppsspp-assets"])
         self.assertEqual(deliverable(entries, True), [])
 
+    def test_a_platform_package_can_differ_both_ways(self):
+        """Batocera's hcl package installs the binary alone, its blueMSX core
+        package copies the machines beside the core."""
+        sys.path.insert(0, str(ROOT / "scripts"))
+        from validation import installed_with_emulator
+
+        hcl = {"name": "data.pak", "bundled": True, "not_bundled_on": ["batocera"]}
+        machines = {"name": "Machines/MSX/config.ini", "installed_on": ["batocera", "retropie"]}
+        self.assertTrue(installed_with_emulator(hcl, "retrobat"))
+        self.assertFalse(installed_with_emulator(hcl, "batocera"))
+        self.assertTrue(installed_with_emulator(machines, "retropie"))
+        self.assertFalse(installed_with_emulator(machines, "retroarch"))
+        self.assertTrue(installed_with_emulator(hcl))
+
+    def test_every_platform_named_is_registered(self):
+        sys.path.insert(0, str(ROOT / "scripts"))
+        from common import list_registered_platforms, load_emulator_profiles
+
+        known = set(list_registered_platforms(str(ROOT / "platforms"), include_archived=True))
+        unknown = sorted({
+            (name, platform)
+            for name, profile in load_emulator_profiles(str(ROOT / "emulators")).items()
+            for entry in (profile.get("files") or []) + (profile.get("data_directories") or [])
+            if isinstance(entry, dict)
+            for field in ("installed_on", "not_bundled_on")
+            for platform in entry.get(field) or []
+            if platform not in known
+        })
+        self.assertEqual(unknown, [])
+
     def test_the_pack_builder_skips_it(self):
         source = (ROOT / "scripts" / "packextras.py").read_text(encoding="utf-8")
         self.assertIn('if u.get("bundled"):', source)