"""深入诊断 SPI 数据时间分布"""
import os
from pathlib import Path
from dotenv import load_dotenv

load_dotenv(Path(__file__).resolve().parents[2] / ".env")

PG_DSN = os.environ.get("PG_DSN")
if not PG_DSN:
    raise RuntimeError("PG_DSN 未设置")

import psycopg2
import psycopg2.extras

conn = psycopg2.connect(PG_DSN)
conn.autocommit = True

def q(sql, params=None):
    with conn.cursor(cursor_factory=psycopg2.extras.RealDictCursor) as cur:
        cur.execute(sql, params or ())
        return cur.fetchall()

print("SPI 数据时间分布深入诊断")
print("=" * 60)

# 1. dwd_settlement_head 全量时间范围
rows = q("""
    SELECT 
        MIN(pay_time) AS min_pay, MAX(pay_time) AS max_pay,
        COUNT(*) AS total,
        COUNT(*) FILTER (WHERE settle_type IN (1, 3)) AS consume_cnt,
        COUNT(*) FILTER (WHERE settle_type = 2) AS recharge_cnt
    FROM dwd.dwd_settlement_head
""")
r = rows[0]
print(f"dwd_settlement_head 全量:")
print(f"  总记录: {r['total']}, 消费(1,3): {r['consume_cnt']}, 充值(2): {r['recharge_cnt']}")
print(f"  pay_time 范围: {r['min_pay']} ~ {r['max_pay']}")

# 2. settle_type IN (1,3) 的最近数据
rows2 = q("""
    SELECT pay_time::date AS pay_date, COUNT(*) AS cnt
    FROM dwd.dwd_settlement_head
    WHERE settle_type IN (1, 3)
    GROUP BY pay_time::date
    ORDER BY pay_date DESC
    LIMIT 20
""")
print(f"\n消费记录(settle_type IN 1,3) 最近 20 天:")
for r in rows2:
    print(f"  {r['pay_date']}: {r['cnt']} 条")

# 3. SPI 用 NOW() - 90 days 作为截止，检查实际截止时间
rows3 = q("SELECT NOW() AS now_ts, NOW() - INTERVAL '90 days' AS cutoff_90, NOW() - INTERVAL '30 days' AS cutoff_30")
r = rows3[0]
print(f"\n时间窗口:")
print(f"  NOW(): {r['now_ts']}")
print(f"  90天截止: {r['cutoff_90']}")
print(f"  30天截止: {r['cutoff_30']}")

# 4. 在 90 天窗口内，按月统计消费记录
rows4 = q("""
    SELECT 
        DATE_TRUNC('month', pay_time)::date AS month,
        COUNT(*) AS cnt,
        COUNT(DISTINCT member_id) AS members,
        SUM(COALESCE(pay_amount, 0)) AS total_amount
    FROM dwd.dwd_settlement_head
    WHERE settle_type IN (1, 3)
      AND pay_time >= NOW() - INTERVAL '90 days'
    GROUP BY month
    ORDER BY month
""")
print(f"\n90天窗口内消费记录按月分布:")
for r in rows4:
    print(f"  {r['month']}: {r['cnt']} 条, {r['members']} 会员, 金额 {r['total_amount']}")

# 5. 在 30 天窗口内的消费记录
rows5 = q("""
    SELECT 
        COUNT(*) AS cnt,
        COUNT(DISTINCT member_id) AS members,
        SUM(COALESCE(pay_amount, 0)) AS total_amount,
        MIN(pay_time) AS min_pay,
        MAX(pay_time) AS max_pay
    FROM dwd.dwd_settlement_head
    WHERE settle_type IN (1, 3)
      AND pay_time >= NOW() - INTERVAL '30 days'
""")
r = rows5[0]
print(f"\n30天窗口内消费记录:")
print(f"  记录数: {r['cnt']}, 会员数: {r['members']}, 总金额: {r['total_amount']}")
print(f"  时间范围: {r['min_pay']} ~ {r['max_pay']}")

# 6. 检查 ODS 层最新数据时间
rows6 = q("""
    SELECT 
        MIN(pay_time) AS min_pay, MAX(pay_time) AS max_pay, COUNT(*) AS cnt
    FROM ods.ods_settlement_records
""")
r = rows6[0]
print(f"\nods_settlement_records:")
print(f"  记录数: {r['cnt']}, pay_time 范围: {r['min_pay']} ~ {r['max_pay']}")

# 7. 检查 DWD 层是否有 2 月中旬之后的数据
rows7 = q("""
    SELECT pay_time::date AS pay_date, COUNT(*) AS cnt
    FROM dwd.dwd_settlement_head
    WHERE pay_time >= '2026-02-10'
    ORDER BY pay_date
""")
print(f"\nDWD 2026-02-10 之后的数据:")
for r in rows7:
    print(f"  {r['pay_date']}: {r['cnt']} 条")

if not rows7:
    print("  无数据!")

conn.close()
print("\n诊断完成。")