From ccf1a1573adac625d1d326cfb16896acaabd1ca2 Mon Sep 17 00:00:00 2001
From: Abdessamad Derraz <3028866+Abdess@users.noreply.github.com>
Date: Tue, 6 Oct 2026 01:47:48 +0200
Subject: [PATCH] fix: read platform hashes as verify does in PR checks

---
 scripts/validate_pr.py           | 34 ++++++++++++++++++----------
 tests/test_validate_pr_hashes.py | 39 ++++++++++++++++++++++++++++++++
 2 files changed, 61 insertions(+), 12 deletions(-)
 create mode 100644 tests/test_validate_pr_hashes.py

diff --git a/scripts/validate_pr.py b/scripts/validate_pr.py
index 25a6a8e8..462dbdd6 100644
--- a/scripts/validate_pr.py
+++ b/scripts/validate_pr.py
@@ -25,7 +25,13 @@ import sys
 from pathlib import Path
 
 sys.path.insert(0, os.path.dirname(__file__))
-from common import compute_hashes, list_registered_platforms, load_database, yaml_load
+from common import (
+    compute_hashes,
+    list_registered_platforms,
+    load_database,
+    load_platform_config,
+    parse_md5_list,
+)
 
 try:
     import yaml
@@ -109,20 +115,24 @@ def load_platform_hashes(platforms_dir: str) -> dict:
     if not os.path.isdir(platforms_dir) or yaml is None:
         return known
 
+    # Read the way verify reads them: inheritance and shared groups resolved,
+    # and a Recalbox md5 field split into the revisions it lists. The raw
+    # file kept "a,b,c" as one value, so a PR adding the seventh accepted
+    # falcon.img was told its hash differs.
     for name in list_registered_platforms(platforms_dir, include_archived=True):
-        f = Path(platforms_dir) / f"{name}.yml"
-        with open(f) as fh:
-            try:
-                config = yaml_load(fh) or {}
-            except yaml.YAMLError:
-                continue
+        try:
+            config = load_platform_config(name, platforms_dir)
+        except (OSError, yaml.YAMLError) as exc:
+            print(f"WARNING: {name}: {exc}", file=sys.stderr)
+            continue
 
-        for sys_id, system in config.get("systems", {}).items():
+        for system in config.get("systems", {}).values():
             for file_entry in system.get("files", []):
-                if "sha1" in file_entry:
-                    known["sha1"].add(file_entry["sha1"])
-                if "md5" in file_entry:
-                    known["md5"].add(file_entry["md5"])
+                sha1 = file_entry.get("sha1")
+                for value in sha1 if isinstance(sha1, list) else [sha1]:
+                    if value:
+                        known["sha1"].add(str(value).lower())
+                known["md5"].update(parse_md5_list(file_entry.get("md5")))
                 if "name" in file_entry:
                     known["names"].add(file_entry["name"])
 
diff --git a/tests/test_validate_pr_hashes.py b/tests/test_validate_pr_hashes.py
new file mode 100644
index 00000000..28075668
--- /dev/null
+++ b/tests/test_validate_pr_hashes.py
@@ -0,0 +1,39 @@
+"""validate_pr knows every revision a platform accepts.
+
+It read the platform YAMLs raw: a Recalbox md5 field "a,b,c" stayed one
+value, so a pull request adding the seventh falcon.img Recalbox lists was
+answered "hash differs, may be a variant". Inheritance and shared groups
+were lost the same way.
+"""
+
+from __future__ import annotations
+
+import sys
+import unittest
+from pathlib import Path
+
+import yaml
+
+REPO_ROOT = Path(__file__).resolve().parents[1]
+sys.path.insert(0, str(REPO_ROOT / "scripts"))
+
+
+class EveryAcceptedRevisionIsKnown(unittest.TestCase):
+    def test_multi_hash_members_are_known(self):
+        import validate_pr
+
+        known = validate_pr.load_platform_hashes(str(REPO_ROOT / "platforms"))
+        raw = yaml.safe_load((REPO_ROOT / "platforms" / "recalbox.yml").read_text())
+        listed = [
+            value.strip().lower()
+            for system in raw["systems"].values()
+            for entry in system.get("files", [])
+            if "," in str(entry.get("md5", ""))
+            for value in str(entry["md5"]).split(",")
+        ]
+        self.assertTrue(listed)
+        self.assertEqual([v for v in listed if v not in known["md5"]], [])
+
+
+if __name__ == "__main__":
+    unittest.main()