feat: add generate_truth.py CLI

scripts/common.py

@@ -1171,7 +1171,8 @@ def _enrich_hashes(entry: dict, db: dict) -> None:
         record = db["files"].get(sha1)
     if record is None and md5:
         by_md5 = db.get("by_md5", {})
-        ref_sha1 = by_md5.get(md5.lower())
+        md5_str = md5 if isinstance(md5, str) else md5[0] if md5 else ""
+        ref_sha1 = by_md5.get(md5_str.lower()) if md5_str else None
         if ref_sha1 and db.get("files"):
             record = db["files"].get(ref_sha1)
     if record is None:
@@ -1197,9 +1198,12 @@ def _merge_file_into_system(
 
     if existing is not None:
         existing["_cores"] = existing.get("_cores", set()) | {emu_name}
-        existing["_source_refs"] = existing.get("_source_refs", set()) | (
-            {file_entry["source_ref"]} if file_entry.get("source_ref") else set()
-        )
+        sr = file_entry.get("source_ref")
+        if sr is not None:
+            sr_key = str(sr) if not isinstance(sr, str) else sr
+            existing["_source_refs"] = existing.get("_source_refs", set()) | {sr_key}
+        else:
+            existing.setdefault("_source_refs", set())
         if file_entry.get("required") and not existing.get("required"):
             existing["required"] = True
         for h in ("sha1", "md5", "sha256", "crc32"):
@@ -1226,8 +1230,10 @@ def _merge_file_into_system(
         if val is not None:
             entry[field] = val
     entry["_cores"] = {emu_name}
-    if file_entry.get("source_ref"):
-        entry["_source_refs"] = {file_entry["source_ref"]}
+    sr = file_entry.get("source_ref")
+    if sr is not None:
+        sr_key = str(sr) if not isinstance(sr, str) else sr
+        entry["_source_refs"] = {sr_key}
     else:
         entry["_source_refs"] = set()
 

scripts/generate_truth.py

@@ -0,0 +1,123 @@
+#!/usr/bin/env python3
+"""Generate ground-truth YAML files per platform from emulator profiles.
+
+Usage:
+    python scripts/generate_truth.py --platform retroarch --output-dir dist/truth
+    python scripts/generate_truth.py --all --output-dir dist/truth
+"""
+
+from __future__ import annotations
+
+import argparse
+import os
+import sys
+
+sys.path.insert(0, os.path.dirname(__file__))
+from common import (
+    generate_platform_truth,
+    list_registered_platforms,
+    load_database,
+    load_emulator_profiles,
+    load_target_config,
+)
+
+try:
+    import yaml
+except ImportError:
+    print("Error: PyYAML required (pip install pyyaml)", file=sys.stderr)
+    sys.exit(1)
+
+DEFAULT_OUTPUT_DIR = "dist/truth"
+DEFAULT_PLATFORMS_DIR = "platforms"
+DEFAULT_EMULATORS_DIR = "emulators"
+DEFAULT_DB_FILE = "database.json"
+
+
+def parse_args(argv: list[str] | None = None) -> argparse.Namespace:
+    parser = argparse.ArgumentParser(
+        description="Generate ground-truth YAML from emulator profiles",
+    )
+    group = parser.add_mutually_exclusive_group(required=True)
+    group.add_argument("--all", action="store_true", help="all registered platforms")
+    group.add_argument("--platform", help="single platform name")
+    parser.add_argument(
+        "--output-dir", default=DEFAULT_OUTPUT_DIR, help="output directory",
+    )
+    parser.add_argument(
+        "--target", "-t", default=None, help="hardware target filter",
+    )
+    parser.add_argument(
+        "--include-archived", action="store_true",
+        help="include archived platforms with --all",
+    )
+    parser.add_argument(
+        "--platforms-dir", default=DEFAULT_PLATFORMS_DIR,
+    )
+    parser.add_argument(
+        "--emulators-dir", default=DEFAULT_EMULATORS_DIR,
+    )
+    parser.add_argument("--db", default=DEFAULT_DB_FILE, help="database.json path")
+    return parser.parse_args(argv)
+
+
+def main(argv: list[str] | None = None) -> None:
+    args = parse_args(argv)
+
+    # Load registry
+    registry_path = os.path.join(args.platforms_dir, "_registry.yml")
+    with open(registry_path) as f:
+        registry = (yaml.safe_load(f) or {}).get("platforms", {})
+
+    # Load emulator profiles
+    profiles = load_emulator_profiles(args.emulators_dir)
+
+    # Load database (optional)
+    db: dict | None = None
+    if os.path.exists(args.db):
+        db = load_database(args.db)
+
+    # Determine platforms
+    if args.all:
+        platforms = list_registered_platforms(
+            args.platforms_dir, include_archived=args.include_archived,
+        )
+    else:
+        platforms = [args.platform]
+
+    os.makedirs(args.output_dir, exist_ok=True)
+
+    for name in platforms:
+        # Resolve target cores
+        target_cores: set[str] | None = None
+        if args.target:
+            try:
+                target_cores = load_target_config(
+                    name, args.target, args.platforms_dir,
+                )
+            except FileNotFoundError:
+                print(f"  {name}: no target config, skipped")
+                continue
+
+        result = generate_platform_truth(
+            name, registry, profiles, db=db, target_cores=target_cores,
+        )
+
+        out_path = os.path.join(args.output_dir, f"{name}.yml")
+        with open(out_path, "w") as f:
+            yaml.dump(
+                result, f,
+                default_flow_style=False, sort_keys=False, allow_unicode=True,
+            )
+
+        n_systems = len(result.get("systems", {}))
+        n_files = sum(
+            len(sys_data.get("files", {}))
+            for sys_data in result.get("systems", {}).values()
+        )
+        print(f"  {name}: {n_systems} systems, {n_files} files -> {out_path}")
+
+    print("Done.")
+
+
+if __name__ == "__main__":
+    main()