Initial local translation API with persistent projects and batch jobs

examples/extract_tomie_ui.py

@@ -0,0 +1,69 @@
+"""Extract Naninovel DefaultUI TextAsset from Tomie into this workspace.
+
+Only reads the game. Writes a JSON translation input and a source manifest.
+Requires: python -m pip install -e ".[unity]"
+"""
+import argparse
+import hashlib
+import json
+from pathlib import Path
+
+
+def parse_dictionary(text):
+    result = []
+    seen = set()
+    for line_number, line in enumerate(text.splitlines(), 1):
+        if not line.strip() or line.lstrip().startswith((";", "#", "//")):
+            continue
+        key, separator, value = line.partition(":")
+        if not separator or not key.strip() or not value.strip():
+            continue
+        key = key.strip()
+        if key in seen:
+            raise ValueError(f"Duplicate dictionary key at line {line_number}: {key}")
+        seen.add(key)
+        result.append({"id": key, "text": value.lstrip(), "line": line_number})
+    return result
+
+
+def main():
+    parser = argparse.ArgumentParser(description=__doc__)
+    parser.add_argument("game", type=Path)
+    parser.add_argument("--output", type=Path, default=Path("data/tomie/input.json"))
+    parser.add_argument("--limit", type=int, default=15, help="0 extracts the full DefaultUI dictionary")
+    args = parser.parse_args()
+    if args.limit < 0:
+        parser.error("limit must be nonnegative")
+    try:
+        import UnityPy
+    except ImportError:
+        parser.error('Install optional extractor: python -m pip install -e ".[unity]"')
+    asset_path = args.game.resolve() / "TomieWGM_Data" / "data.unity3d"
+    output = args.output.resolve()
+    if output.is_relative_to(args.game.resolve()):
+        parser.error("Choose an output outside the game directory")
+    environment = UnityPy.load(str(asset_path))
+    assets = [obj for obj in environment.objects if obj.type.name == "TextAsset" and obj.read().m_Name == "DefaultUI"]
+    if len(assets) != 1:
+        parser.error(f"Expected exactly one DefaultUI TextAsset, found {len(assets)}")
+    obj = assets[0]
+    script = obj.read().m_Script
+    if isinstance(script, bytes):
+        script = script.decode("utf-8-sig")
+    entries = parse_dictionary(script)
+    selected = entries[:args.limit] if args.limit else entries
+    output.parent.mkdir(parents=True, exist_ok=True)
+    output.write_text(json.dumps([{"id": row["id"], "text": row["text"]} for row in selected], ensure_ascii=False, indent=2), encoding="utf-8")
+    manifest = {
+        "game": str(args.game.resolve()), "source_file": str(asset_path), "asset_name": "DefaultUI",
+        "path_id": obj.path_id, "serialized_file": obj.assets_file.name,
+        "script_sha256": hashlib.sha256(script.encode("utf-8")).hexdigest(),
+        "total_entries": len(entries), "selected_entries": selected,
+        "note": "Read-only extraction. No patching, font replacement or changes to game files.",
+    }
+    output.with_suffix(".manifest.json").write_text(json.dumps(manifest, ensure_ascii=False, indent=2), encoding="utf-8")
+    print(f"Extracted {len(selected)}/{len(entries)} UI strings to {output}")
+
+
+if __name__ == "__main__":
+    main()