# -*- coding: utf-8 -*-
"""比对 BD_Manual 文档中的字段列表与数据库实际列，输出差异报告。

用法：python scripts/ops/_verify_bd_manual_fields.py
输出：stdout（差异报告）
"""
import os
import sys
import re

# 加载 .env
from dotenv import load_dotenv
load_dotenv(os.path.join(os.path.dirname(__file__), '..', '..', '.env'))

import psycopg2

DSN = os.environ.get("TEST_DB_DSN")
if not DSN:
    print("ERROR: TEST_DB_DSN 未设置", file=sys.stderr)
    sys.exit(1)

# 要验证的表（BD_Manual 文件 → 表列表）
TABLES_TO_CHECK = [
    # assistant_service_records
    "dwd.dwd_assistant_service_log",
    "dwd.dwd_assistant_service_log_ex",
    # recharge_settlements
    "dwd.dwd_recharge_order",
    "dwd.dwd_recharge_order_ex",
    # store_goods_master
    "dwd.dim_store_goods",
    "dwd.dim_store_goods_ex",
    # site_tables_master
    "dwd.dim_table",
    "dwd.dim_table_ex",
    # goods_stock_movements
    "dwd.dwd_goods_stock_movement",
    # goods_stock_summary
    "dwd.dwd_goods_stock_summary",
    # member_balance_changes
    "dwd.dwd_member_balance_change",
    "dwd.dwd_member_balance_change_ex",
    # store_goods_sales_records
    "dwd.dwd_store_goods_sale",
    "dwd.dwd_store_goods_sale_ex",
    # DWS
    "dws.dws_goods_stock_daily_summary",
    "dws.dws_goods_stock_monthly_summary",
]

# BD_Manual 文件 → 文档中列出的列名
BD_MANUAL_COLS: dict[str, list[str]] = {}

def parse_md_table_cols(filepath: str) -> dict[str, list[str]]:
    """从 BD_Manual markdown 文件中提取每个表的列名列表。"""
    result = {}
    current_table = None
    in_table = False

    with open(filepath, 'r', encoding='utf-8') as f:
        lines = f.readlines()

    for line in lines:
        line = line.strip()
        # 检测表名（如 "## 1. dwd_assistant_service_log（主表）"）
        m = re.match(r'^##\s+\d+\.\s+(\w+)', line)
        if m:
            current_table = m.group(1)
            in_table = False
            continue

        # 检测 markdown 表格行
        if current_table and '|' in line:
            cells = [c.strip() for c in line.split('|')]
            cells = [c for c in cells if c]
            if len(cells) >= 2:
                first = cells[0]
                # 跳过表头分隔行
                if first.startswith('---') or first.startswith(':---'):
                    continue
                # 跳过表头行
                if first in ('DWD 列名', 'DWS 列名', 'ODS 字段', '日期'):
                    in_table = True
                    continue
                if in_table:
                    # 提取列名（去掉 backtick）
                    col = first.strip('`').strip()
                    if col and not col.startswith('~~') and col != 'scd2_*':
                        if current_table not in result:
                            result[current_table] = []
                        result[current_table].append(col)

    return result


# 解析所有 BD_Manual 文件
BD_FILES = [
    "docs/database/BD_Manual_assistant_service_records.md",
    "docs/database/BD_Manual_recharge_settlements.md",
    "docs/database/BD_Manual_store_goods_master.md",
    "docs/database/BD_Manual_site_tables_master.md",
    "docs/database/BD_Manual_goods_stock_movements.md",
    "docs/database/BD_Manual_goods_stock_summary.md",
    "docs/database/BD_Manual_member_balance_changes.md",
    "docs/database/BD_Manual_store_goods_sales_records.md",
    "docs/database/BD_Manual_dws_goods_stock_summary.md",
]

all_doc_cols: dict[str, list[str]] = {}
for f in BD_FILES:
    parsed = parse_md_table_cols(f)
    for table, cols in parsed.items():
        all_doc_cols[table] = cols

# 查询数据库实际列
SCD2_COLS = {"scd2_start_time", "scd2_end_time", "scd2_is_current", "scd2_version"}

conn = psycopg2.connect(DSN)
try:
    cur = conn.cursor()
    for full_table in TABLES_TO_CHECK:
        schema, table = full_table.split('.')
        cur.execute("""
            SELECT column_name
            FROM information_schema.columns
            WHERE table_schema = %s AND table_name = %s
            ORDER BY ordinal_position
        """, (schema, table))
        db_cols = [row[0] for row in cur.fetchall()]
        db_cols_no_scd2 = [c for c in db_cols if c not in SCD2_COLS]

        doc_cols = all_doc_cols.get(table, [])

        if not doc_cols:
            print(f"\n⚠️  {full_table}: 文档中未找到列定义（表名 '{table}' 未匹配）")
            print(f"   DB 列 ({len(db_cols)}): {db_cols}")
            continue

        doc_set = set(doc_cols)
        db_set = set(db_cols_no_scd2)

        in_doc_not_db = doc_set - db_set
        in_db_not_doc = db_set - doc_set

        status = "✅" if not in_doc_not_db and not in_db_not_doc else "❌"
        print(f"\n{status} {full_table}: 文档 {len(doc_cols)} 列, DB {len(db_cols_no_scd2)} 列 (不含 SCD2)")

        if in_doc_not_db:
            print(f"   📄 文档有但 DB 无: {sorted(in_doc_not_db)}")
        if in_db_not_doc:
            print(f"   🗄️  DB 有但文档无: {sorted(in_db_not_doc)}")
finally:
    conn.close()