change to hex and update DBs (#1284)

scripts/migrate_replay_ids.py

@@ -0,0 +1,127 @@
+#!/usr/bin/env python3
+"""
+migrate_replay_ids.py
+
+One-shot migration: renames decimal replay directories to hex and updates
+the _id field inside each replay_data.json.gz to match.
+
+Run on the server:
+    python3 migrate_replay_ids.py
+    python3 migrate_replay_ids.py --dry-run   # preview without touching anything
+
+Reads STORAGE_VOL_PATH from environment or TSSBOT/.env.
+"""
+from __future__ import annotations
+
+import argparse
+import gzip
+import json
+import os
+import sys
+from pathlib import Path
+
+_HERE = Path(__file__).resolve().parent
+
+# Try loading .env if python-dotenv is available
+try:
+    from dotenv import load_dotenv
+    load_dotenv(dotenv_path=_HERE / ".env")
+except ImportError:
+    pass
+
+STORAGE_VOL_PATH = os.environ.get("STORAGE_VOL_PATH", "").strip()
+if not STORAGE_VOL_PATH:
+    print("ERROR: STORAGE_VOL_PATH not set", file=sys.stderr)
+    sys.exit(1)
+
+REPLAYS_DIR = Path(STORAGE_VOL_PATH) / "REPLAYS" / "TSS"
+
+
+def is_decimal_id(name: str) -> bool:
+    """True if the directory name looks like a plain decimal integer (not hex)."""
+    try:
+        int(name)
+        return True
+    except ValueError:
+        return False
+
+
+def to_hex(decimal_str: str) -> str:
+    return hex(int(decimal_str))[2:].lower()
+
+
+def migrate_one(src_dir: Path, *, dry_run: bool) -> tuple[str, str] | None:
+    """
+    Rename src_dir from decimal to hex, update _id inside replay_data.json.gz.
+    Returns (old_name, new_name) on success, None if already hex or skipped.
+    """
+    name = src_dir.name
+    if not is_decimal_id(name):
+        return None
+
+    hex_name = to_hex(name)
+    dst_dir  = src_dir.parent / hex_name
+
+    if dst_dir.exists():
+        print(f"  SKIP {name} → {hex_name} (destination already exists)")
+        return None
+
+    gz_path = src_dir / "replay_data.json.gz"
+
+    if gz_path.exists():
+        try:
+            with gzip.open(gz_path, "rb") as fh:
+                data = json.loads(fh.read().decode("utf-8"))
+        except Exception as exc:
+            print(f"  ERROR reading {gz_path}: {exc}")
+            return None
+
+        data["_id"] = hex_name
+
+        if not dry_run:
+            with gzip.open(gz_path, "wb") as fh:
+                fh.write(json.dumps(data, ensure_ascii=False).encode("utf-8"))
+
+    if not dry_run:
+        src_dir.rename(dst_dir)
+
+    return name, hex_name
+
+
+def main() -> None:
+    parser = argparse.ArgumentParser(description=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter)
+    parser.add_argument("--dry-run", action="store_true", help="Print what would happen without changing anything")
+    args = parser.parse_args()
+
+    if not REPLAYS_DIR.is_dir():
+        print(f"ERROR: {REPLAYS_DIR} does not exist", file=sys.stderr)
+        sys.exit(1)
+
+    entries = sorted(p for p in REPLAYS_DIR.iterdir() if p.is_dir())
+    candidates = [p for p in entries if is_decimal_id(p.name)]
+
+    if not candidates:
+        print("No decimal-named replay directories found. Nothing to do.")
+        return
+
+    mode = "DRY RUN — " if args.dry_run else ""
+    print(f"{mode}Found {len(candidates)} decimal director(ies) to migrate in {REPLAYS_DIR}\n")
+
+    ok = skipped = errors = 0
+    for src_dir in candidates:
+        result = migrate_one(src_dir, dry_run=args.dry_run)
+        if result is None:
+            skipped += 1
+        else:
+            old, new = result
+            tag = "[dry-run] " if args.dry_run else ""
+            print(f"  {tag}{old}  →  {new}")
+            ok += 1
+
+    print(f"\nDone. migrated={ok}  skipped={skipped}  errors={errors}")
+    if args.dry_run:
+        print("(dry run — nothing was changed)")
+
+
+if __name__ == "__main__":
+    main()