autool-dispatcher/scripts/reset_catalog_update_state.py

#!/usr/bin/env python3

import argparse
import json
import os
import sqlite3
import sys
import time
from typing import Any, Dict


PROJECT_ROOT = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
if PROJECT_ROOT not in sys.path:
    sys.path.insert(0, PROJECT_ROOT)

from config import MONITORING_DB_PATH


def _safe_loads(value: Any) -> Dict[str, Any]:
    if not value:
        return {}
    try:
        payload = json.loads(value)
    except (TypeError, ValueError):
        return {}
    return payload if isinstance(payload, dict) else {}


def _clear_payload_last_updated(raw_payload: Any) -> str:
    payload = _safe_loads(raw_payload)
    if not payload:
        return str(raw_payload or "")
    payload["last_updated"] = ""
    original_row = payload.get("original_row")
    if isinstance(original_row, dict):
        original_row = dict(original_row)
        original_row["last_updated"] = ""
        original_row["最后更新"] = ""
        original_row["更新时间"] = ""
        payload["original_row"] = original_row
    return json.dumps(payload, ensure_ascii=False)


def reset_catalog_update_state(args) -> int:
    db_path = str(args.db_path or "").strip()
    apply_changes = bool(args.apply)
    clear_payload = bool(args.clear_payload)
    now_iso = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime())
    with sqlite3.connect(db_path) as connection:
        connection.row_factory = sqlite3.Row
        table_row = connection.execute(
            """
            SELECT name
              FROM sqlite_master
             WHERE type = 'table'
               AND name = 'app_catalog'
            """
        ).fetchone()
        if not table_row:
            raise SystemExit(f"app_catalog not found in {db_path}")

        counts = connection.execute(
            """
            SELECT
                COUNT(*) AS total,
                COALESCE(SUM(CASE WHEN COALESCE(last_updated, '') != '' THEN 1 ELSE 0 END), 0) AS last_updated_rows,
                COALESCE(SUM(CASE WHEN COALESCE(last_update_interval_days, 0) != 0 THEN 1 ELSE 0 END), 0) AS interval_rows
              FROM app_catalog
            """
        ).fetchone()
        task_counts = connection.execute(
            """
            SELECT
                COALESCE(SUM(CASE WHEN COALESCE(is_new_app, 1) = 0 THEN 1 ELSE 0 END), 0) AS app_update_rows,
                COALESCE(
                    SUM(
                        CASE
                            WHEN task_status = 'pending'
                                 AND COALESCE(is_new_app, 1) = 0
                                 AND COALESCE(error_reason, '') LIKE 'catalog_version_update:%'
                            THEN 1
                            ELSE 0
                        END
                    ),
                    0
                ) AS version_pending_rows
              FROM collection_task
            """
        ).fetchone()

        if not apply_changes:
            print(
                "dry_run=1 db={db} total={total} clear_last_updated={last_updated_rows} "
                "reset_interval_rows={interval_rows} app_update_tasks={app_update_rows} "
                "delete_version_pending={version_pending_rows}".format(
                    db=db_path,
                    total=int(counts["total"] or 0),
                    last_updated_rows=int(counts["last_updated_rows"] or 0),
                    interval_rows=int(counts["interval_rows"] or 0),
                    app_update_rows=int(task_counts["app_update_rows"] or 0),
                    version_pending_rows=int(task_counts["version_pending_rows"] or 0),
                )
            )
            return 0

        cursor = connection.execute(
            """
            DELETE FROM collection_task
             WHERE task_status = 'pending'
               AND COALESCE(is_new_app, 1) = 0
               AND COALESCE(error_reason, '') LIKE 'catalog_version_update:%'
            """
        )
        deleted_version_pending = int(cursor.rowcount or 0)

        cursor = connection.execute(
            """
            UPDATE app_catalog
               SET last_update_interval_days = 0,
                   updated_at = ?
             WHERE COALESCE(last_update_interval_days, 0) != 0
            """,
            (now_iso,),
        )
        reset_interval_rows = int(cursor.rowcount or 0)

        if clear_payload:
            payload_rows = connection.execute(
                """
                SELECT package_name, task_payload_json
                  FROM app_catalog
                 WHERE COALESCE(task_payload_json, '') != ''
                """
            ).fetchall()
            for row in payload_rows:
                connection.execute(
                    """
                    UPDATE app_catalog
                       SET task_payload_json = ?,
                           updated_at = ?
                     WHERE package_name = ?
                    """,
                    (_clear_payload_last_updated(row["task_payload_json"]), now_iso, row["package_name"]),
                )

        cursor = connection.execute(
            """
            UPDATE app_catalog
               SET last_updated = '',
                   updated_at = ?
             WHERE COALESCE(last_updated, '') != ''
            """,
            (now_iso,),
        )
        cleared_last_updated = int(cursor.rowcount or 0)
        print(
            "dry_run=0 db={db} cleared_last_updated={cleared_last_updated} "
            "reset_interval_rows={reset_interval_rows} deleted_version_pending={deleted_version_pending} "
            "clear_payload={clear_payload}".format(
                db=db_path,
                cleared_last_updated=cleared_last_updated,
                reset_interval_rows=reset_interval_rows,
                deleted_version_pending=deleted_version_pending,
                clear_payload=int(clear_payload),
            )
        )
        return 0


def build_parser() -> argparse.ArgumentParser:
    parser = argparse.ArgumentParser(
        description="Clear app_catalog last_updated and delete pending catalog_version_update collection_task rows."
    )
    parser.add_argument("--db-path", default=MONITORING_DB_PATH, help="Target monitoring sqlite path.")
    parser.add_argument("--apply", action="store_true", help="Apply changes. Without this, only prints counts.")
    parser.add_argument(
        "--clear-payload",
        action="store_true",
        help="Also clear last_updated inside task_payload_json/original_row.",
    )
    return parser


def main() -> int:
    return reset_catalog_update_state(build_parser().parse_args())


if __name__ == "__main__":
    raise SystemExit(main())