Update wehago matching logic and exclude reports

2026-07-03 09:07:04 +09:00
parent a09190ecd3
commit da073ad6ef
62 changed files with 36940 additions and 1554 deletions
@@ -0,0 +1,86 @@
+from __future__ import annotations
+
+import json
+import sys
+from datetime import datetime
+from pathlib import Path
+
+sys.path.insert(0, str(Path(__file__).resolve().parents[1]))
+
+import scripts.refresh_hanmac_wehago_ledgers as refresh
+import scripts.retry_failed_wehago_accounts_direct as direct
+import scripts.wehago_data_download_2022_work as wehago
+import scripts.wehago_ledger_api_download as api_download
+
+
+def main() -> int:
+    base = Path(r"\\wsl.localhost\Ubuntu\home\b17301\WEHAGO_DB\data_download\hanmac_refresh")
+    canonical = Path(r"\\wsl.localhost\Ubuntu\home\b17301\WEHAGO_DB\data_download\hanmac\2025")
+    run_root = base / f"retry_2025_all_direct_{datetime.now():%Y%m%d_%H%M%S}"
+    staging = run_root / "staging" / "2025"
+    staging.mkdir(parents=True, exist_ok=True)
+    wehago.CHROME_DEBUGGER_ADDRESS = "127.0.0.1:9225"
+    wehago.DOWNLOAD_DIR = run_root
+    driver = wehago.build_driver(run_root)
+    try:
+        cookies = direct.cookie_map(driver)
+    finally:
+        driver.quit()
+    accounts = wehago.discover_downloaded_accounts(canonical)
+    manifest: list[dict[str, object]] = []
+    failures: list[dict[str, str]] = []
+
+    print(f"2025 all accounts {len(accounts)} staging {staging}", flush=True)
+    for index, account in enumerate(accounts, start=1):
+        try:
+            rows = direct.fetch_ledger_rows(2025, account, cookies)
+            api_download.validate_api_rows(account, rows)
+            api_download.write_api_rows(staging / account.safe_filename, account, rows)
+            manifest.append(
+                {
+                    "account_code": account.code,
+                    "account_name": account.name,
+                    "api_request_code": f"{account.code}00",
+                    "api_response_rows": len(rows),
+                }
+            )
+            print(f"[{index}/{len(accounts)}] OK {account.code} rows={len(rows)}", flush=True)
+        except Exception as exc:
+            reason = f"{type(exc).__name__}: {exc}"
+            failures.append({"account_code": account.code, "account_name": account.name, "reason": reason})
+            print(f"[{index}/{len(accounts)}] FAIL {account.code}: {reason}", flush=True)
+
+    api_download.write_progress(
+        staging,
+        {
+            "status": "completed_with_failures" if failures else "completed",
+            "completed": len(manifest),
+            "total": len(accounts),
+            "accounts": manifest,
+            "failures": failures,
+        },
+    )
+    validation = refresh.validate_staging(staging)
+    summary = {
+        "year": 2025,
+        "requested": len(accounts),
+        "downloaded": len(manifest),
+        "failures": len(failures),
+        "failed_codes": [item["account_code"] for item in failures],
+        "staging": str(staging),
+        "validation": validation,
+    }
+    (run_root / "retry_summary.json").write_text(
+        json.dumps(summary, ensure_ascii=False, indent=2),
+        encoding="utf-8",
+    )
+    print(f"SUMMARY_PATH {run_root / 'retry_summary.json'}", flush=True)
+    print(
+        f"VALIDATION_PASSED {validation.get('passed')} DUP_GROUPS {len(validation.get('duplicate_contents', []))}",
+        flush=True,
+    )
+    return 0 if not failures else 1
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())