fix: persist media decisions across production restores

rf4_research/media_store.py

@@ -0,0 +1,59 @@
+from __future__ import annotations
+
+import json
+import shutil
+from datetime import datetime, timezone
+from pathlib import Path
+
+
+def _safe_path(root: Path, relative: str) -> Path | None:
+    target = (root / relative).resolve()
+    if not target.is_relative_to(root.resolve()):
+        return None
+    return target
+
+
+def bootstrap_media_store(source_root: Path, target_root: Path) -> dict[str, int]:
+    """Merge the immutable image baseline into the persistent media volume.
+
+    Existing decisions and files in the volume always win. New baseline assets
+    are copied in on release, so a recreated container cannot reset decisions
+    while a newer Git baseline can still add assets.
+    """
+    source_manifest_path = source_root / "manifest.json"
+    target_manifest_path = target_root / "manifest.json"
+    source = json.loads(source_manifest_path.read_text(encoding="utf-8"))
+    target = json.loads(target_manifest_path.read_text(encoding="utf-8")) if target_manifest_path.exists() else {"version": 1, "assets": []}
+    target_by_url = {item.get("asset_url"): item for item in target.get("assets", [])}
+    added = 0
+    copied = 0
+    target_root.mkdir(parents=True, exist_ok=True)
+
+    for source_item in source.get("assets", []):
+        url = source_item.get("asset_url")
+        if not url or url in target_by_url:
+            continue
+        item = dict(source_item)
+        target_by_url[url] = item
+        target.setdefault("assets", []).append(item)
+        added += 1
+
+    for item in target.get("assets", []):
+        for relative in [item.get("local_path"), *[variant.get("local_path") for variant in item.get("derivatives", [])]]:
+            if not isinstance(relative, str):
+                continue
+            source_path = _safe_path(source_root, relative)
+            target_path = _safe_path(target_root, relative)
+            if not source_path or not target_path or not source_path.is_file() or target_path.exists():
+                continue
+            target_path.parent.mkdir(parents=True, exist_ok=True)
+            shutil.copy2(source_path, target_path)
+            copied += 1
+
+    target["version"] = max(int(source.get("version", 1)), int(target.get("version", 1)))
+    if added or copied or not target_manifest_path.exists():
+        target["updated_at"] = datetime.now(timezone.utc).isoformat()
+        temporary = target_manifest_path.with_suffix(".json.tmp")
+        temporary.write_text(json.dumps(target, ensure_ascii=False, indent=2) + "\n", encoding="utf-8")
+        temporary.replace(target_manifest_path)
+    return {"added_assets": added, "copied_files": copied}