#!/usr/bin/env python3
"""
增量修复：只补当前 supplier_name_match 为空的包材产品
从采购订单CSV匹配供应商和成本价（取首个非空值）
"""
import csv
import pymysql
from config import DB_CONFIG

CSV_FILE = "/home/gxd/.openclaw/workspace-dingtalk-assistant/media/inbound/采购订单-20260531-1780213945661.csv"

# 后缀列表
SUFFIXES = [
    '-FM/GZ', '-TMGJ-NL', '-ALJK-BT', '-ALJK-FM/GZ',
    '-TM-BT', '-TM-PZ', '-TM-CH', '-TM-BQ', '-TM-SSM',
    '-TM-HMT', '-TM-RG', '-TM-GZ', '-TM-FM/GZ',
    '-PDD-BT', '-PDD-PZ', '-PDD-CH', '-PDD-HMT',
    '-NL', '-BT', '-PZ', '-CH', '-BQ', '-SSM', '-HMT',
    '-RG', '-GZ', '-JD', '-PDD', '-ALJK', '-TM',
    '-WPH', '-DW',
]

def strip_suffix(code):
    code = code.strip()
    for s in SUFFIXES:
        if code.endswith(s):
            base = code[:-len(s)]
            if base:
                return base
    return code


def main():
    conn = pymysql.connect(**DB_CONFIG)
    cur = conn.cursor()

    # 1. 读取CSV，聚合取首个非空供应商
    csv_best = {}
    with open(CSV_FILE, encoding="utf-8-sig") as f:
        reader = csv.DictReader(f)
        for row in reader:
            code = row["物料编码"].strip()
            supplier = row["供应商"].strip()
            price_str = row["含税单价"].strip()
            try:
                price = float(price_str) if price_str else None
            except ValueError:
                price = None

            if code not in csv_best:
                csv_best[code] = (supplier, price)
            else:
                old_sup, old_price = csv_best[code]
                # 优先取非空供应商
                if supplier and not old_sup:
                    csv_best[code] = (supplier, old_price if old_price is not None else price)
                # 如果旧记录没有价格，用新价格
                if old_price is None and price is not None:
                    csv_best[code] = (old_sup if old_sup else supplier, price)

    print(f"CSV 唯一物料编码: {len(csv_best)}")

    # 2. 构建 CSV 索引：精确匹配 + 后缀剥离匹配
    csv_exact = {}   # code -> (supplier, price)
    csv_base = {}    # base_code -> (supplier, price)
    for code, (sup, price) in csv_best.items():
        csv_exact[code] = (sup, price)
        base = strip_suffix(code)
        # 只当 base != code 且 base 不在精确索引中时才加
        if base != code and base not in csv_exact:
            if base not in csv_base:
                csv_base[base] = (sup, price)
            else:
                # 优先取非空供应商
                old_sup, old_price = csv_base[base]
                if sup and not old_sup:
                    csv_base[base] = (sup, old_price if old_price is not None else price)

    # 3. 找出需要修复的包材产品——supplier_name_match 为空
    keywords = ['瓶子','泵头','彩盒','海绵头','瓶贴','标签','收缩膜','瓶体','瓶标','软管','盖子','膜袋','标贴','唇膏管','内料','海绵','瓶子+内塞','泵头','瓶标','海绵头']
    clauses = " OR ".join(["product_name LIKE '%%{}%%'".format(k) for k in keywords])
    cur.execute("""
        SELECT id, product_code, product_name, cost_price 
        FROM products 
        WHERE ({}) AND (supplier_name_match IS NULL OR supplier_name_match = '')
    """.format(clauses))
    need_fix = cur.fetchall()
    print(f"需要修复的包材: {len(need_fix)}")

    # 4. 逐个匹配
    updates = []
    exact_hits = 0
    base_hits = 0
    no_hits = 0

    for pid, product_code, product_name, old_price in need_fix:
        code = product_code.strip()
        sup = None
        price = None

        # 4a. 精确匹配
        if code in csv_exact:
            sup_s, price_s = csv_exact[code]
            if sup_s:
                sup = sup_s
            if price_s is not None:
                price = price_s
            exact_hits += 1
        # 4b. 后缀剥离匹配
        elif code in csv_base:
            sup_s, price_s = csv_base[code]
            if sup_s:
                sup = sup_s
            if price_s is not None:
                price = price_s
            base_hits += 1
        else:
            no_hits += 1

        if sup or price is not None:
            # 保留原有价格当新价格不可用时
            final_price = price if price is not None else (float(old_price) if old_price else None)
            updates.append((sup or '', final_price, pid))

    print(f"\n匹配结果: 精确={exact_hits}, 后缀={base_hits}, 未匹配={no_hits}")
    print(f"待更新: {len(updates)}")

    # 5. 批量更新
    if updates:
        update_sql = "UPDATE products SET supplier_name_match=%s, cost_price=%s WHERE id=%s"
        updated = 0
        batch_size = 500
        for i in range(0, len(updates), batch_size):
            batch = updates[i:i+batch_size]
            cur.executemany(update_sql, batch)
            conn.commit()
            updated += len(batch)
            print(f"  已更新 {updated}/{len(updates)}")

    # 6. 最终统计
    cur.execute("SELECT COUNT(*) FROM products WHERE ({}) AND (supplier_name_match IS NULL OR supplier_name_match = '')".format(clauses))
    still_empty = cur.fetchone()[0]
    print(f"\n修复后，包材仍然为空: {still_empty}")

    cur.execute("SELECT COUNT(*) FROM products WHERE supplier_name_match != ''")
    total_sup = cur.fetchone()[0]
    cur.execute("SELECT COUNT(*) FROM products WHERE cost_price > 0")
    total_price = cur.fetchone()[0]
    print(f"全部产品: 有供应商={total_sup}, 有成本价={total_price}")

    conn.close()
    print("\n✅ 完成！")


if __name__ == "__main__":
    main()