| |
| """ |
| Duplicate and relation auditor for UnsolvedAstronomy v1.1.0. |
| Verifies 0 duplicate canonical problem statements among published records and validates merged_duplicate links. |
| """ |
| import json |
| import re |
| import sys |
| from collections import Counter |
| from pathlib import Path |
|
|
| ROOT = Path(__file__).resolve().parent.parent |
| PROBLEMS_FILE = ROOT / "problems.json" |
| ARCHIVED_FILE = ROOT / "archived_problems.json" |
|
|
|
|
| def norm_text(s: str) -> str: |
| return re.sub(r"[^a-z0-9]+", " ", (s or "").lower()).strip() |
|
|
|
|
| def main(): |
| if not PROBLEMS_FILE.exists(): |
| print(f"Missing {PROBLEMS_FILE}") |
| sys.exit(1) |
| published = json.loads(PROBLEMS_FILE.read_text(encoding="utf-8")) |
| archived = json.loads(ARCHIVED_FILE.read_text(encoding="utf-8")) if ARCHIVED_FILE.exists() else [] |
| dups = [r for r in archived if r["status"] == "merged_duplicate"] |
|
|
| stmt_norms = Counter(norm_text(r["statement"]) for r in published) |
| collisions = {k: v for k, v in stmt_norms.items() if v > 1} |
| title_norms = Counter(norm_text(r["title"]) for r in published) |
| title_collisions = {k: v for k, v in title_norms.items() if v > 1} |
| pub_ids = {r["id"] for r in published} |
|
|
| bad_dup_targets = [r["id"] for r in dups if r.get("duplicate_of") not in pub_ids] |
| low_src_records = [ |
| r["id"] |
| for r in published |
| if len({s.get("source_id") for s in r.get("sources", []) if s.get("source_id")}) < 4 |
| ] |
|
|
| print(f"Checked {len(published)} published records and {len(dups)} merged_duplicate records.") |
| print(f" Statement collisions among published: {len(collisions)}") |
| print(f" Title collisions among published: {len(title_collisions)}") |
| print(f" Published records with < 4 distinct literature sources: {len(low_src_records)}") |
| print(f" Invalid duplicate_of targets: {len(bad_dup_targets)}") |
| if collisions or title_collisions or low_src_records or bad_dup_targets: |
| sys.exit(1) |
|
|
|
|
| if __name__ == "__main__": |
| main() |
|
|
|
|
