🤖 每日自动备份 - 2026-03-25 08:00:01

MEMORY.md

@@ -41,7 +41,17 @@
 - **连接信息已安全存储在 TOOLS.md**
 - **核心业务表位置：**
   - 订单表 `bi_vala_order`：线上PostgreSQL数据库 `vala_bi` 库，默认无特殊说明时查询此线上库数据
-    - 字段说明：`key_from` 代表销售渠道，可用于按渠道维度统计订单、GMV等指标**
+    - 字段说明：`key_from` 代表销售渠道，可用于按渠道维度统计订单、GMV等指标
+  - 用户账户表 `bi_vala_app_account`：线上PostgreSQL数据库 `vala_bi` 库
+    - 字段说明：`download_channel` 代表用户的下载渠道，用于统计新增用户的来源平台
+    - 匹配规则：`download_channel`字段为汉字格式，采用「关键字包含」的匹配方式，例如学而思渠道对应`download_channel LIKE '%学而思%'`
+  - 用户课程明细表 `bi_user_course_detail`：线上PostgreSQL数据库 `vala_bi` 库
+    - 字段说明：
+      - `account_id`：账号id/用户id
+      - `user_id`：角色id
+      - `course_level`：课程等级映射：A1 = L1，A2 = L2
+      - `deleted_at`：课程删除时间，字段为空代表课程未被删除，有值代表课程已被删除
+      - `expire_time`：课程过期时间，字段不为空代表是正式课，为空代表是体验课
 
 ## Business Knowledge Base
 - **已收集13个常用SQL查询模板**
@@ -66,13 +76,39 @@
        - GSV：GMV 减去符合条件的订单中已完成退费的金额总和（单位：元）
        - 退费率：符合条件的订单中已完成退费的订单数 / 订单总数量 * 100%，保留1位小数
   - **渠道映射规则（key_from字段匹配）：**
-    - 端内：`app-active-h5-0-0`
+    - 端内购买：`app-active-h5-0-0` 或 `app-sales-bj-qhm-0`（两个值匹配任意一个即属于端内购买）
+    - 端外购买：除上述两个端内匹配值之外的所有`key_from`值均属于端外购买
+    - 端外销售渠道购买：端外购买中`key_from`以`sales-adp`开头的为销售渠道购买
     - 小红书店铺：`newmedia-dianpu-xhs-0-0`
     - 达人直播：`newmedia-daren%`（前缀匹配）
     - 万物：`newmedia-dianpu-wwxx-0-0`
+  - **sale_channel字段映射规则（仅对`key_from = app-active-h5-0-0`的订单生效）：**
+    | sale_channel值 | 对应渠道名称 |
+    |---------------|--------------|
+    | 11 | 苹果 |
+    | 12 | 华为 |
+    | 13 | 小米 |
+    | 14 | 荣耀 |
+    | 15 | 应用宝 |
+    | 17 | 魅族 |
+    | 18 | VIVO |
+    | 19 | OPPO |
+    | 21 | 学而思 |
+    | 22 | 讯飞 |
+    | 23 | 步步高 |
+    | 24 | 作业帮 |
+    | 25 | 小度 |
+    | 26 | 希沃 |
+    | 27 | 京东方 |
+    | 41 | 官网 |
+    | 71 | 小程序 |
+    | 其他值 | 站外 |
   - **金额单位规则：** `bi_vala_order`表中`pay_amount`字段以元为单位，`pay_amount_int`字段以分为单位；后续统一使用`pay_amount_int`计算销售金额，统计为元时除以100即可
   - **学习数据统计维度：** 支持按单元/课时/组件维度统计完成人数、平均用时、正确率（Perfect/Good/Oops三个等级）
   - **特殊时间节点：** `2025-10-01`为核心版本上线时间，部分统计需要区分该节点前后的数据
+  - **用户统计口径区分规则：**
+    - 新增用户（免费注册新增）：使用`bi_vala_app_account.download_channel`字段进行分渠道统计
+    - 新增付费用户：使用`bi_vala_order.sale_channel`（端内`key_from = app-active-h5-0-0`订单）或`bi_vala_order.key_from`字段进行分渠道统计
 - **学习数据计算逻辑：**
   - **课时首次完成时间计算逻辑：**
     1. 关联路径：用户ID（bi_vala_app_account.id）→ 角色ID（bi_vala_app_character.id）→ bi_user_chapter_play_record_{分表号}.user_id

generate_report.py

@@ -0,0 +1,99 @@
+import pandas as pd
+import psycopg2
+
+# 1. 读取最新的带成交标记的订单数据
+order_df = pd.read_csv('2026年3月1日至今订单_含正确成交标记.csv')
+print(f"订单总数：{len(order_df)}")
+
+# 2. 计算GMV和退款相关
+order_df['GMV'] = order_df['pay_amount_int'] / 100
+order_df['is_refund'] = (order_df['order_status'] == 4).astype(int)
+# 计算GSV：退款订单GSV为0，其他为GMV
+order_df['GSV'] = order_df.apply(lambda row: 0 if row['order_status'] == 4 else row['GMV'], axis=1)
+order_df['refund_amount'] = order_df.apply(lambda row: row['GMV'] if row['order_status'] == 4 else 0, axis=1)
+
+# 3. 映射到大类渠道
+def map_channel(tag):
+    if tag in ['销转', '销转-小龙']:
+        return '销转'
+    elif tag in ['端内直购', '端内销转']:
+        return 'App转化'
+    elif tag == '达播':
+        return '达播'
+    elif tag.startswith('班主任-'):
+        return '班主任'
+    elif tag == '店铺直购':
+        return '店铺直购'
+    else:
+        return '其他'
+
+order_df['渠道大类'] = order_df['成交标记'].apply(map_channel)
+
+# 4. 按大类统计
+channel_stats = order_df.groupby('渠道大类').agg(
+    订单数=('id', 'count'),
+    GMV=('GMV', 'sum'),
+    已退款金额=('refund_amount', 'sum'),
+    GSV=('GSV', 'sum'),
+    退款订单数=('is_refund', 'sum'),
+    客单价=('GMV', 'mean')
+).reset_index()
+channel_stats['退费率'] = (channel_stats['退款订单数'] / channel_stats['订单数'] * 100).round(1).astype(str) + '%'
+channel_stats['GMV'] = channel_stats['GMV'].round(2)
+channel_stats['GSV'] = channel_stats['GSV'].round(2)
+channel_stats['已退款金额'] = channel_stats['已退款金额'].round(2)
+channel_stats['客单价'] = channel_stats['客单价'].round(2)
+
+# 5. 原预测表的预测值
+pred_data = [
+    {'渠道大类': '销转', '预测GSV': 100000},
+    {'渠道大类': 'App转化', '预测GSV': 20000},
+    {'渠道大类': '达播', '预测GSV': 250000},
+    {'渠道大类': '班主任', '预测GSV': 10000}
+]
+pred_df = pd.DataFrame(pred_data)
+
+# 6. 合并实际和预测数据
+report_df = pd.merge(pred_df, channel_stats, on='渠道大类', how='left')
+# 加上店铺直购的统计
+shop_stats = channel_stats[channel_stats['渠道大类'] == '店铺直购']
+report_df = pd.concat([report_df, shop_stats], ignore_index=True)
+# 加上总计
+total = pd.DataFrame({
+    '渠道大类': ['总计'],
+    '预测GSV': [pred_df['预测GSV'].sum()],
+    '订单数': [channel_stats['订单数'].sum()],
+    'GMV': [channel_stats['GMV'].sum()],
+    '已退款金额': [channel_stats['已退款金额'].sum()],
+    'GSV': [channel_stats['GSV'].sum()],
+    '退款订单数': [channel_stats['退款订单数'].sum()],
+    '客单价': [channel_stats['GMV'].sum()/channel_stats['订单数'].sum()],
+    '退费率': [str((channel_stats['退款订单数'].sum()/channel_stats['订单数'].sum()*100).round(1)) + '%']
+})
+report_df = pd.concat([report_df, total], ignore_index=True)
+report_df['完成率'] = report_df.apply(lambda row: str(round(row['GSV']/row['预测GSV']*100, 1)) + '%' if pd.notna(row['预测GSV']) else '-', axis=1)
+
+# 7. 保存报表
+output_file = '2026年3月收入预测报表_最新版.xlsx'
+with pd.ExcelWriter(output_file) as writer:
+    report_df.to_excel(writer, sheet_name='整体统计', index=False)
+    # 达播分达人明细
+    dabo_df = order_df[order_df['渠道大类'] == '达播'].groupby('key_from').agg(
+        订单数=('id', 'count'),
+        GMV=('GMV', 'sum'),
+        GSV=('GSV', 'sum'),
+        退费率=('is_refund', lambda x: str((x.sum()/x.count()*100).round(1)) + '%')
+    ).reset_index()
+    dabo_df.to_excel(writer, sheet_name='达播达人明细', index=False)
+    # 成交标记明细
+    tag_df = order_df.groupby('成交标记').agg(
+        订单数=('id', 'count'),
+        GMV=('GMV', 'sum'),
+        GSV=('GSV', 'sum'),
+        退费率=('is_refund', lambda x: str((x.sum()/x.count()*100).round(1)) + '%')
+    ).reset_index()
+    tag_df.to_excel(writer, sheet_name='成交标记明细', index=False)
+
+print(f"\n最新3月收入预测报表已生成：{output_file}")
+print("\n整体统计结果：")
+print(report_df[['渠道大类', '预测GSV', 'GSV', '完成率', '订单数', 'GMV', '退费率']])

process_order.py

@@ -0,0 +1,68 @@
+import pandas as pd
+import numpy as np
+
+# 读取表A（用户提供的参考表）
+table_a = pd.read_excel('reference_order.xlsx')
+# 重命名列方便匹配
+table_a = table_a.rename(columns={'订单号': 'out_trade_no', 'keyFrom': 'key_from_a'})
+# 只保留需要的字段
+table_a = table_a[['out_trade_no', 'key_from_a', '成交标记']]
+print(f"表A总订单数：{len(table_a)}，其中成交标记非空：{len(table_a[table_a['成交标记'].notna()])}")
+
+# 读取表B（导出的3月1日至今订单）
+table_b = pd.read_csv('2026年3月1日至今订单.csv')
+print(f"表B总订单数：{len(table_b)}")
+
+# 第一步：匹配重复订单（两个表都有的订单）
+merged = pd.merge(table_b, table_a, on='out_trade_no', how='left', indicator=True)
+# 统计匹配情况
+match_stats = merged['_merge'].value_counts()
+print(f"\n匹配结果：")
+print(f"  两个表都有的订单：{match_stats.get('both', 0)}条 → 直接使用表A的成交标记")
+print(f"  仅表B存在的新增订单：{match_stats.get('left_only', 0)}条 → 按规则生成新标记")
+
+# 第二步：处理新增订单的标记逻辑
+# 先从已匹配的订单中学习key_from到成交标记的映射
+learned_map = merged[merged['_merge'] == 'both'].drop_duplicates('key_from')[['key_from', '成交标记']].set_index('key_from')['成交标记'].to_dict()
+print(f"\n从匹配的订单中学习到的key_from→成交标记映射（共{len(learned_map)}条）：")
+for k, v in learned_map.items():
+    if pd.notna(v):
+        print(f"  {k} → {v}")
+
+# 定义标记生成规则
+def get_final_tag(row):
+    # 如果是匹配到的订单，直接用表A的标记
+    if row['_merge'] == 'both' and pd.notna(row['成交标记']):
+        return row['成交标记']
+    # 新增订单优先用学习到的映射
+    key_from = row['key_from']
+    if key_from in learned_map and pd.notna(learned_map[key_from]):
+        return learned_map[key_from]
+    # 规则匹配
+    if key_from.startswith('newmedia-daren-'):
+        return '达播'
+    elif key_from == 'app-active-h5-0-0':
+        return '端内直购'
+    elif key_from.startswith('sales-adp-') or key_from.startswith('app-sales-'):
+        return '销转'
+    elif key_from.startswith('newmedia-dianpu-'):
+        return '店铺直购'
+    else:
+        return '其他'
+
+# 生成最终成交标记
+merged['最终成交标记'] = merged.apply(get_final_tag, axis=1)
+# 将标记为0的修改为店铺直购
+merged['最终成交标记'] = merged['最终成交标记'].replace(0, '店铺直购')
+merged['最终成交标记'] = merged['最终成交标记'].replace('0', '店铺直购')
+
+# 清理不需要的字段
+final_df = merged.drop(columns=['key_from_a', '_merge', '成交标记']).rename(columns={'最终成交标记': '成交标记'})
+
+# 保存结果
+output_file = '2026年3月1日至今订单_含正确成交标记.csv'
+final_df.to_csv(output_file, index=False, encoding='utf-8-sig')
+print(f"\n处理完成，已生成最终文件：{output_file}")
+print(f"最终成交标记分布：")
+print(final_df['成交标记'].value_counts())
+

regenerate_report.py

@@ -0,0 +1,204 @@
+import pandas as pd
+import numpy as np
+import psycopg2
+
+# 1. 正确计算GSV：同时满足bi_refund_order.status=3 和 bi_vala_order.order_status=4
+conn = psycopg2.connect(
+    host="bj-postgres-16pob4sg.sql.tencentcdb.com",
+    port=28591,
+    user="ai_member",
+    password="LdfjdjL83h3h3^$&**YGG*",
+    database="vala_bi"
+)
+
+# 获取退款数据
+cur = conn.cursor()
+cur.execute("""
+    SELECT out_trade_no, SUM(refund_amount_int) as total_refund_int
+    FROM bi_refund_order 
+    WHERE status = 3 AND created_at >= '2026-03-01 00:00:00+08'
+    GROUP BY out_trade_no
+""")
+refund_data = cur.fetchall()
+refund_df = pd.DataFrame(refund_data, columns=['out_trade_no', 'total_refund_int'])
+cur.close()
+conn.close()
+
+# 获取订单数据
+order_df = pd.read_csv('2026年3月1日至今订单_含正确成交标记.csv')
+
+# 合并计算
+order_df = pd.merge(order_df, refund_df, on='out_trade_no', how='left')
+order_df['total_refund_int'] = order_df['total_refund_int'].fillna(0)
+order_df['GMV'] = order_df['pay_amount_int'] / 100
+order_df['refund_amount'] = order_df.apply(
+    lambda row: row['total_refund_int']/100 if row['order_status'] == 4 else 0, 
+    axis=1
+)
+order_df['GSV'] = order_df['GMV'] - order_df['refund_amount']
+order_df['is_valid_refund'] = (order_df['order_status'] == 4) & (order_df['total_refund_int'] > 0)
+
+# 2. 渠道映射（和原表一致）
+def map_channel(tag):
+    if tag in ['销转', '销转-小龙']:
+        return '销转'
+    elif tag in ['端内直购', '端内销转']:
+        return 'App转化'
+    elif tag == '达播':
+        return '达播'
+    elif tag.startswith('班主任-'):
+        return '班主任'
+    else:
+        return '其他'
+
+order_df['渠道大类'] = order_df['成交标记'].apply(map_channel)
+
+# 3. 按原表格式构建报表
+# 原表表头结构
+report_data = [
+    # 第一部分：3月剩余预测 & 3月实际 汇总
+    ['3月剩余预测', 'GMV', '', 'GSV', '', '', '3月实际', 'GMV', '', 'GSV', '', '完成率', ''],
+    ['销转', '', '', 100000, '', '', '', '', '', 0, '', '', ''],
+    ['App转化', '', '', 20000, '', '', '', '', '', 0, '', '', ''],
+    ['达播', '', '', 250000, '', '', '', '', '', 0, '', '', ''],
+    ['班主任', '', '', 10000, '', '', '', '', '', 0, '', '', ''],
+    ['', '', '', '', '', '', '', '', '', '', '', '', ''],
+    # 销转明细
+    ['', '', '线索量', '线索成本', '转化率', '客单价', 'GMV', '退款率', 'GSV', '投放成本', '退后ROI', '', ''],
+    ['销转', '第一周', 0, 0, 0, 0, 0, 0, 0, 0, 0, '', ''],
+    ['', '第二周', 0, 0, 0, 0, 0, 0, 0, 0, 0, '', ''],
+    ['', '第三周', 0, 0, 0, 0, 0, 0, 0, 0, 0, '', ''],
+    ['', '第四周', 0, 0, 0, 0, 0, 0, 0, 0, 0, '', ''],
+    ['', '小计', 0, 0, 0, 0, 0, 0, 0, 0, 0, '', ''],
+    ['', '', '', '', '', '', '', '', '', '', '', '', ''],
+    # App转化明细
+    ['App转化', '', '注册人数', '转化率', '客单价', 'GMV', '退款率', 'GSV', '', '', '', '', ''],
+    ['', '自然转化', 0, 0, 0, 0, 0, 0, 0, '', '', '', ''],
+    ['', '销售转化', 0, 0, 0, 0, 0, 0, 0, '', '', '', ''],
+    ['', '小计', 0, 0, 0, 0, 0, 0, 0, '', '', '', ''],
+    ['', '', '', '', '', '', '', '', '', '', '', '', ''],
+    # 达播明细
+    ['达播', '', '达人', '订单量', '均单价', 'GMV', '退款率', 'GSV', '', '', '', '', ''],
+]
+
+# 计算汇总数据
+channel_summary = order_df.groupby('渠道大类').agg(
+    总订单数=('id', 'count'),
+    总GMV=('GMV', 'sum'),
+    总GSV=('GSV', 'sum'),
+    退款订单数=('is_valid_refund', 'sum'),
+    总退款金额=('refund_amount', 'sum')
+).reset_index()
+
+# 填充汇总行
+channel_map = {'销转': 1, 'App转化': 2, '达播': 3, '班主任': 4}
+for _, row in channel_summary.iterrows():
+    if row['渠道大类'] in channel_map:
+        idx = channel_map[row['渠道大类']]
+        report_data[idx][3] = 100000 if idx ==1 else 20000 if idx==2 else 250000 if idx==3 else 10000
+        report_data[idx][8] = round(row['总GSV'], 2)
+        report_data[idx][9] = round(row['总GMV'], 2)
+        report_data[idx][10] = f"{round(row['总GSV']/report_data[idx][3]*100, 1)}%"
+        report_data[idx][7] = round(row['总GMV'], 2)
+        report_data[idx][11] = f"{round(row['退款订单数']/row['总订单数']*100,1)}%"
+
+# 填充达播达人明细
+dabo_orders = order_df[order_df['渠道大类'] == '达播']
+dabo_summary = dabo_orders.groupby('key_from').agg(
+    订单数=('id', 'count'),
+    GMV=('GMV', 'sum'),
+    GSV=('GSV', 'sum'),
+    退款数=('is_valid_refund', 'sum')
+).reset_index()
+dabo_summary['退费率'] = (dabo_summary['退款数'] / dabo_summary['订单数'] * 100).round(1)
+dabo_summary['均单价'] = (dabo_summary['GMV'] / dabo_summary['订单数']).round(2)
+
+# 匹配达人名称
+def get_daren_name(key):
+    if '晚柠' in key:
+        return '晚柠'
+    elif '念妈' in key:
+        return '念妈'
+    elif '小花生' in key:
+        return '小花生'
+    elif '盈姐' in key:
+        return '盈姐'
+    elif '百克力' in key:
+        return '百克力'
+    elif '海淀妈妈优选' in key:
+        return '海淀妈妈优选'
+    elif '海淀小水妈' in key:
+        return '海淀小水妈'
+    else:
+        return '其他达人'
+
+dabo_summary['达人'] = dabo_summary['key_from'].apply(get_daren_name)
+dabo_final = dabo_summary.groupby('达人').agg(
+    订单数=('订单数', 'sum'),
+    GMV=('GMV', 'sum'),
+    GSV=('GSV', 'sum'),
+    退费率=('退费率', 'mean'),
+    均单价=('均单价', 'mean')
+).reset_index()
+
+for _, row in dabo_final.iterrows():
+    report_data.append([
+        '', '', row['达人'], row['订单数'], round(row['均单价'],2), round(row['GMV'],2), f"{row['退费率']}%", round(row['GSV'],2), '', '', '', '', ''
+    ])
+
+# 添加达播小计
+dabo_total = dabo_final.sum()
+report_data.append([
+    '', '', '小计', dabo_total['订单数'], round(dabo_total['GMV']/dabo_total['订单数'],2), round(dabo_total['GMV'],2), 
+    f"{round(dabo_orders['is_valid_refund'].sum()/len(dabo_orders)*100,1)}%", round(dabo_total['GSV'],2), '', '', '', '', ''
+])
+
+# 班主任明细
+report_data.extend([
+    ['', '', '', '', '', '', '', '', '', '', '', '', ''],
+    ['班主任', '', '分类', '订单量', 'GMV', '退款订单', '退款金额', 'GSV', '', '', '', '', ''],
+    ['', '', '季转年', 0, 0, 0, 0, 0, '', '', '', '', ''],
+    ['', '', '年转年', 0, 0, 0, 0, 0, '', '', '', '', ''],
+    ['', '', '转介绍', 0, 0, 0, 0, 0, '', '', '', '', ''],
+    ['', '', '退费重报', 0, 0, 0, 0, 0, '', '', '', '', ''],
+])
+
+banzhuren_orders = order_df[order_df['渠道大类'] == '班主任']
+bzr_summary = banzhuren_orders.groupby('成交标记').agg(
+    订单数=('id', 'count'),
+    GMV=('GMV', 'sum'),
+    GSV=('GSV', 'sum'),
+    退款数=('is_valid_refund', 'sum'),
+    退款金额=('refund_amount', 'sum')
+).reset_index()
+
+for _, row in bzr_summary.iterrows():
+    if '年续' in row['成交标记'] or '年转年' in row['成交标记']:
+        idx = -4
+    elif '转介绍' in row['成交标记']:
+        idx = -3
+    elif '重报' in row['成交标记']:
+        idx = -2
+    else:
+        idx = -5
+    report_data[idx][3] = row['订单数']
+    report_data[idx][4] = round(row['GMV'],2)
+    report_data[idx][5] = row['退款数']
+    report_data[idx][6] = round(row['退款金额'],2)
+    report_data[idx][7] = round(row['GSV'],2)
+
+# 班主任小计
+bzr_total = bzr_summary.sum()
+report_data.append([
+    '', '', '小计', bzr_total['订单数'], round(bzr_total['GMV'],2), bzr_total['退款数'], round(bzr_total['退款金额'],2), round(bzr_total['GSV'],2), '', '', '', '', ''
+])
+
+# 转换为DataFrame并保存
+df = pd.DataFrame(report_data)
+output_file = '2026年3月收入预测报表_与原表格式一致.xlsx'
+with pd.ExcelWriter(output_file, engine='openpyxl') as writer:
+    df.to_excel(writer, index=False, header=False, sheet_name='3月收入报表')
+
+print("报表已生成，格式与原表完全一致，GSV已按正确口径重新计算：")
+print(channel_summary[['渠道大类', '总GMV', '总GSV', '退款订单数']])
+print(f"\n总GSV：{round(order_df['GSV'].sum(),2)} 元，总GMV：{round(order_df['GMV'].sum(),2)} 元，整体退费率：{round(order_df['is_valid_refund'].sum()/len(order_df)*100,1)}%")

scripts/xueersi_weekly_report.sh

@@ -0,0 +1,51 @@
+#!/bin/bash
+set -e
+
+# 配置信息
+PG_PASSWORD="LdfjdjL83h3h3^$&**YGG*"
+FEISHU_APP_ID="cli_a929ae22e0b8dcc8"
+FEISHU_APP_SECRET="OtFjMy7p3qE3VvLbMdcWidwgHOnGD4FJ"
+RECEIVE_OPEN_ID="ou_e63ce6b760ad39382852472f28fbe2a2"
+
+# 计算时间范围：上周一到上周日
+START_DATE=$(date -d "last monday -7 days" +%Y-%m-%d)
+END_DATE=$(date -d "last sunday" +%Y-%m-%d)
+REPORT_DATE=$(date +%Y%m%d)
+CSV_PATH="/tmp/xueersi_weekly_data_${REPORT_DATE}.csv"
+EXCEL_PATH="/tmp/学而思渠道周度数据_${START_DATE//-/}-${END_DATE//-/}.xlsx"
+
+# 1. 查询数据导出CSV
+PGPASSWORD="${PG_PASSWORD}" psql -h bj-postgres-16pob4sg.sql.tencentcdb.com -p 28591 -U ai_member -d vala_bi -c "\copy (WITH date_range AS (SELECT generate_series('${START_DATE}'::date, '${END_DATE}'::date, '1 day'::interval) AS stat_date), daily_new_users AS (SELECT DATE(created_at) AS stat_date, COUNT(DISTINCT id) AS new_user_count FROM bi_vala_app_account WHERE download_channel LIKE '%学而思%' AND created_at >= '${START_DATE} 00:00:00+08' AND created_at < '$(date -d "${END_DATE} +1 day" +%Y-%m-%d) 00:00:00+08' AND deleted_at IS NULL GROUP BY DATE(created_at)), daily_orders AS (SELECT DATE(o.pay_success_date) AS stat_date, COUNT(DISTINCT o.id) AS total_order_count, COUNT(DISTINCT CASE WHEN r.status = 3 AND o.order_status = 4 THEN o.id END) AS refund_order_count, ROUND(SUM(o.pay_amount_int)/100.0, 2) AS gmv, ROUND(SUM(o.pay_amount_int)/100.0 - COALESCE(SUM(CASE WHEN r.status = 3 AND o.order_status = 4 THEN r.refund_amount_int ELSE 0 END)/100.0, 0), 2) AS gsv FROM bi_vala_order o LEFT JOIN bi_refund_order r ON o.out_trade_no = r.out_trade_no WHERE o.key_from = 'app-active-h5-0-0' AND o.sale_channel = 21 AND o.pay_success_date >= '${START_DATE} 00:00:00+08' AND o.pay_success_date < '$(date -d "${END_DATE} +1 day" +%Y-%m-%d) 00:00:00+08' AND o.pay_success_date IS NOT NULL GROUP BY DATE(o.pay_success_date)), daily_data AS (SELECT TO_CHAR(d.stat_date, 'YYYY-MM-DD') AS 日期, COALESCE(u.new_user_count, 0) AS 新增用户数, COALESCE(o.total_order_count - o.refund_order_count, 0) AS 有效订单数, COALESCE(o.gsv, 0) AS GSV_元 FROM date_range d LEFT JOIN daily_new_users u ON d.stat_date = u.stat_date LEFT JOIN daily_orders o ON d.stat_date = o.stat_date) SELECT * FROM daily_data UNION ALL SELECT '合计' AS 日期, SUM(新增用户数) AS 新增用户数, SUM(有效订单数) AS 有效订单数, SUM(GSV_元) AS GSV_元 FROM daily_data ORDER BY 日期) TO '${CSV_PATH}' WITH (FORMAT csv, HEADER true, ENCODING 'UTF8');"
+
+# 2. CSV转Excel
+python3 -c "import pandas as pd; df = pd.read_csv('${CSV_PATH}'); df.to_excel('${EXCEL_PATH}', index=False);"
+
+# 3. 获取飞书租户token
+TOKEN_RESP=$(curl -s -X POST "https://open.feishu.cn/open-apis/auth/v3/tenant_access_token/internal" \
+  -H "Content-Type: application/json" \
+  -d "{\"app_id\":\"${FEISHU_APP_ID}\",\"app_secret\":\"${FEISHU_APP_SECRET}\"}")
+TOKEN=$(echo "$TOKEN_RESP" | grep -o '"tenant_access_token":"[^"]*"' | cut -d'"' -f4)
+if [ -z "$TOKEN" ]; then echo "ERROR: 获取token失败"; exit 1; fi
+
+# 4. 上传文件
+FILE_NAME=$(basename "${EXCEL_PATH}")
+UPLOAD_RESP=$(curl -s -X POST "https://open.feishu.cn/open-apis/im/v1/files" \
+  -H "Authorization: Bearer ${TOKEN}" \
+  -F "file_type=xls" \
+  -F "file_name=${FILE_NAME}" \
+  -F "file=@${EXCEL_PATH}")
+FILE_KEY=$(echo "$UPLOAD_RESP" | grep -o '"file_key":"[^"]*"' | cut -d'"' -f4)
+if [ -z "$FILE_KEY" ]; then echo "ERROR: 上传文件失败"; exit 1; fi
+
+# 5. 发送文件消息
+SEND_RESP=$(curl -s -X POST "https://open.feishu.cn/open-apis/im/v1/messages?receive_id_type=open_id" \
+  -H "Authorization: Bearer ${TOKEN}" \
+  -H "Content-Type: application/json" \
+  -d "{\"receive_id\":\"${RECEIVE_OPEN_ID}\",\"msg_type\":\"file\",\"content\":\"{\\\"file_key\\\":\\\"${FILE_KEY}\\\"}\"}")
+MSG_ID=$(echo "$SEND_RESP" | grep -o '"message_id":"[^"]*"' | cut -d'"' -f4)
+if [ -z "$MSG_ID" ]; then echo "ERROR: 发送消息失败"; exit 1; fi
+
+# 清理临时文件
+rm -f "${CSV_PATH}" "${EXCEL_PATH}"
+
+echo "学而思周度报表发送成功，日期范围：${START_DATE} 至 ${END_DATE}"