feat: add a folder check built on the api

scripts/check_platform_folder.py

@@ -0,0 +1,61 @@
+"""Check a BIOS folder against one platform's list, from the published data only.
+
+    python check_platform_folder.py batocera path/to/bios
+"""
+from __future__ import annotations
+
+import hashlib
+import json
+import sys
+import urllib.request
+import zipfile
+from pathlib import Path
+
+API = "https://abdess.github.io/retrobios/api/v1/platforms/"
+
+
+def accepts(entry: dict, path: Path, mode: str) -> bool:
+    """Whether the platform's own check passes the file at path."""
+    raw = entry.get(mode) or ""
+    values = raw if isinstance(raw, list) else str(raw).split(",")
+    declared = [value.strip().lower() for value in values if value.strip()]
+    if mode == "existence" or not declared:
+        return path.is_file()  # nothing to compare: the platform checks presence
+    try:
+        if entry.get("zipped_file"):  # the hash of one member, its name casefolded
+            with zipfile.ZipFile(path) as archive:
+                names = {name.casefold(): name for name in archive.namelist()}
+                member = names.get(entry["zipped_file"].casefold())
+                return bool(member) and hashlib.new(mode, archive.read(member)).hexdigest() in declared
+        actual = hashlib.new(mode, path.read_bytes()).hexdigest()
+        if entry.get("zip_digest") != "composite" and any(actual.startswith(d) for d in declared):
+            return True  # a truncated md5 (Batocera) compares by its prefix
+        if mode == "md5" and path.suffix.lower() == ".zip":  # Recalbox: members in name order
+            with zipfile.ZipFile(path) as archive:
+                names = sorted(name for name in archive.namelist() if not name.endswith("/"))
+                return hashlib.md5(b"".join(map(archive.read, names))).hexdigest() in declared
+    except (OSError, zipfile.BadZipFile):
+        return False
+    return False
+
+
+def check(platform: str, root: Path, api: str = API) -> list[tuple[str, str]]:
+    """(destination, ok | wrong | missing) for each entry of the list."""
+    with urllib.request.urlopen(f"{api}{platform}.json") as response:
+        contract = json.load(response)["items"][0]["contract"]
+    mode = contract["verification_mode"]
+    results = []
+    for system in contract["systems"].values():
+        for entry in system.get("files") or []:
+            destination = entry.get("destination") or entry["name"]
+            path = root / destination
+            verdict = "ok" if accepts(entry, path, mode) else "wrong" if path.is_file() else "missing"
+            results.append((destination, verdict))
+    return results
+
+
+if __name__ == "__main__":
+    found = check(sys.argv[1], Path(sys.argv[2]))
+    for verdict in ("ok", "wrong", "missing"):
+        names = sorted({dest for dest, seen in found if seen == verdict})
+        print(f"{verdict}: {len(names)}" + "".join(f"\n  {name}" for name in names if verdict != "ok"))