Add multilingual audit CI pipeline + extract mandelblog_content_guard

scripts/multilingual_audit_ci.py

@@ -0,0 +1,99 @@
+#!/usr/bin/env python3
+from __future__ import annotations
+
+import argparse
+import json
+from pathlib import Path
+
+
+def load_json(path: Path) -> dict:
+    return json.loads(path.read_text())
+
+
+def locale_rows(payload: dict) -> list[tuple[str, dict]]:
+    summary = payload.get("summary", {})
+    return [(locale, data) for locale, data in summary.items() if locale != "snippets"]
+
+
+def print_error(payload: dict) -> int:
+    error = payload.get("error")
+    if error:
+        print(f"AUDIT ERROR: {error}")
+        return 2
+    return 0
+
+
+def print_summary(payload: dict) -> tuple[int, int]:
+    total_block = 0
+    total_warn = 0
+    for locale, data in locale_rows(payload):
+        sev = data.get("by_severity", {})
+        block = int(sev.get("block", 0) or 0)
+        warn = int(sev.get("warn", 0) or 0)
+        log = int(sev.get("log", 0) or 0)
+        total_block += block
+        total_warn += warn
+        print(
+            f"LOCALE {locale}: issues_found={data.get('issues_found', 0)} "
+            f"issues_remaining={data.get('remaining_issues', 0)} "
+            f"block={block} warn={warn} log={log}"
+        )
+    return total_block, total_warn
+
+
+def print_regressions(current: dict, previous: dict) -> None:
+    prev_summary = {locale: data for locale, data in locale_rows(previous)}
+    regressions = []
+    for locale, data in locale_rows(current):
+        prev = prev_summary.get(locale, {})
+        cur_remaining = int(data.get("remaining_issues", 0) or 0)
+        prev_remaining = int(prev.get("remaining_issues", 0) or 0)
+        cur_sev = data.get("by_severity", {})
+        prev_sev = prev.get("by_severity", {})
+        delta = {
+            "remaining": cur_remaining - prev_remaining,
+            "block": int(cur_sev.get("block", 0) or 0) - int(prev_sev.get("block", 0) or 0),
+            "warn": int(cur_sev.get("warn", 0) or 0) - int(prev_sev.get("warn", 0) or 0),
+            "log": int(cur_sev.get("log", 0) or 0) - int(prev_sev.get("log", 0) or 0),
+        }
+        if any(value > 0 for value in delta.values()):
+            regressions.append((locale, delta))
+    if regressions:
+        print("REGRESSIONS:")
+        for locale, delta in regressions:
+            print(
+                f"- {locale}: remaining={delta['remaining']:+d} block={delta['block']:+d} "
+                f"warn={delta['warn']:+d} log={delta['log']:+d}"
+            )
+    else:
+        print("REGRESSIONS: none")
+
+
+def main() -> int:
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--json", required=True, help="Current multilingual audit JSON file")
+    parser.add_argument("--previous-json", help="Optional previous audit JSON file for regression comparison")
+    args = parser.parse_args()
+
+    current = load_json(Path(args.json))
+    error_status = print_error(current)
+    if error_status:
+        return error_status
+    total_block, total_warn = print_summary(current)
+
+    if args.previous_json:
+        prev_path = Path(args.previous_json)
+        if prev_path.exists():
+            print_regressions(current, load_json(prev_path))
+        else:
+            print("REGRESSIONS: previous artifact not found")
+
+    if total_block > 0:
+        return 2
+    if total_warn > 0:
+        return 1
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())