From 8d81aee2352a4516107e880e871ee6440d7f34d3 Mon Sep 17 00:00:00 2001
From: Abdessamad Derraz <3028866+Abdess@users.noreply.github.com>
Date: Tue, 17 Mar 2026 15:35:30 +0100
Subject: [PATCH] refactor: quality audit fixes, honest verification reporting

- batocera_scraper: fix OrderedDict parsing for ast.literal_eval
- auto_fetch: fix TypeError when sha1/md5 is None
- verify: filter non-ZIP files for zipped_file entries (F2)
- verify: distinguish ZIP read errors from hash mismatches (F5)
- generate_pack: track seen_destinations with source hash (F7)

Batocera ep64/ep128.zip now correctly reported as MISSING
instead of false UNTESTED (resolved to .rom instead of .zip)
---
 scripts/auto_fetch.py               |  4 ++--
 scripts/generate_pack.py            |  4 ++--
 scripts/scraper/batocera_scraper.py |  4 +++-
 scripts/verify.py                   | 23 +++++++++++++++--------
 4 files changed, 22 insertions(+), 13 deletions(-)

diff --git a/scripts/auto_fetch.py b/scripts/auto_fetch.py
index 949d759f..065dabfe 100644
--- a/scripts/auto_fetch.py
+++ b/scripts/auto_fetch.py
@@ -345,8 +345,8 @@ def generate_issue_body(missing: list[dict], platform: str) -> str:
     ]
 
     for entry in missing:
-        sha1 = entry.get("sha1", "N/A")
-        md5 = entry.get("md5", "N/A")
+        sha1 = entry.get("sha1") or "N/A"
+        md5 = entry.get("md5") or "N/A"
         lines.append(f"| `{entry['name']}` | {entry['system']} | `{sha1[:12]}...` | `{md5[:12]}...` |")
 
     lines.extend([
diff --git a/scripts/generate_pack.py b/scripts/generate_pack.py
index 714770c6..0de565f9 100644
--- a/scripts/generate_pack.py
+++ b/scripts/generate_pack.py
@@ -263,7 +263,7 @@ def generate_pack(
     missing_files = []
     untested_files = []
     user_provided = []
-    seen_destinations = set()
+    seen_destinations = {}
 
     with zipfile.ZipFile(zip_path, "w", zipfile.ZIP_DEFLATED) as zf:
         for sys_id, system in sorted(config.get("systems", {}).items()):
@@ -279,7 +279,7 @@ def generate_pack(
                 dedup_key = full_dest
                 if dedup_key in seen_destinations:
                     continue
-                seen_destinations.add(dedup_key)
+                seen_destinations[dedup_key] = file_entry.get("sha1") or file_entry.get("md5") or ""
 
                 storage = file_entry.get("storage", "embedded")
 
diff --git a/scripts/scraper/batocera_scraper.py b/scripts/scraper/batocera_scraper.py
index 2d56f000..aa67acc0 100644
--- a/scripts/scraper/batocera_scraper.py
+++ b/scripts/scraper/batocera_scraper.py
@@ -150,7 +150,9 @@ class Scraper(BaseScraper):
 
         clean_dict_str = "\n".join(lines)
 
-        clean_dict_str = clean_dict_str.replace("OrderedDict(", "dict(")
+        # OrderedDict({...}) -> just the inner dict literal
+        clean_dict_str = re.sub(r'OrderedDict\(\s*\{', '{', clean_dict_str)
+        clean_dict_str = re.sub(r'\}\s*\)', '}', clean_dict_str)
 
         try:
             return ast.literal_eval(clean_dict_str)
diff --git a/scripts/verify.py b/scripts/verify.py
index 02a029b1..cf60ddff 100644
--- a/scripts/verify.py
+++ b/scripts/verify.py
@@ -120,13 +120,16 @@ def resolve_to_local_path(file_entry: dict, db: dict) -> str | None:
                 if os.path.exists(path):
                     candidates.append((path, entry.get("md5", "")))
         if candidates:
+            if has_zipped_file:
+                candidates = [(p, m) for p, m in candidates if p.endswith(".zip")]
             if md5 and not has_zipped_file:
                 md5_lower = md5.lower()
                 for path, db_md5 in candidates:
                     if db_md5.lower() == md5_lower:
                         return path
-            primary = [p for p, _ in candidates if "/.variants/" not in p]
-            return primary[0] if primary else candidates[0][0]
+            if candidates:
+                primary = [p for p, _ in candidates if "/.variants/" not in p]
+                return primary[0] if primary else candidates[0][0]
 
     return None
 
@@ -156,17 +159,21 @@ def verify_entry_md5(file_entry: dict, local_path: str | None) -> dict:
 
     if zipped_file:
         found_in_zip = False
+        had_error = False
         for md5_candidate in md5_list or [""]:
             result = check_inside_zip(local_path, zipped_file, md5_candidate)
             if result == Status.OK:
                 return {"name": name, "status": Status.OK, "path": local_path}
-            if result != "not_in_zip":
+            if result == "error":
+                had_error = True
+            elif result != "not_in_zip":
                 found_in_zip = True
-        reason = (
-            f"{zipped_file} not found inside ZIP"
-            if not found_in_zip
-            else f"{zipped_file} MD5 mismatch inside ZIP"
-        )
+        if had_error and not found_in_zip:
+            reason = f"{local_path} is not a valid ZIP or read error"
+        elif not found_in_zip:
+            reason = f"{zipped_file} not found inside ZIP"
+        else:
+            reason = f"{zipped_file} MD5 mismatch inside ZIP"
         return {
             "name": name, "status": Status.UNTESTED, "path": local_path,
             "reason": reason,