ETL 完成

README.md

@@ -1,57 +1,593 @@
 # 飞球 ETL 系统（ODS → DWD）
 
-面向门店业务的 ETL：拉取/或离线灌入上游 JSON，先落地 ODS，再清洗装载 DWD（含 SCD2 维度、事实增量），并提供质量校验报表。
+面向门店业务的 ETL：从上游 API 或离线 JSON 采集订单、支付、会员、库存等数据，先落地 **ODS**，再清洗装载 **DWD**（含 SCD2 维度、事实增量），并输出质量校验报表。
 
 ## 快速运行（离线示例 JSON）
-1) 环境：Python 3.10+、PostgreSQL；`.env` 关键项：`PG_DSN=postgresql://local-Python:Neo-local-1991125@100.64.0.4:5432/LLZQ-test`，`INGEST_SOURCE_DIR=C:\dev\LLTQ\export\test-json-doc`。  
-2) 安装依赖：  
+> 以下命令默认在 `etl_billiards/` 目录执行（项目会从 `etl_billiards/.env` 读取配置；也可直接设置环境变量）。
+
+1) 环境：Python 3.10+、PostgreSQL。
+2) 配置：编辑 `etl_billiards/.env`（或设环境变量），至少包含：
+   ```env
+   STORE_ID=123
+   PG_DSN=postgresql://<user>:<password>@<host>:<port>/<db>
+   # 示例：使用仓库内置的最小样例（仅 1 个 JSON）
+   INGEST_SOURCE_DIR=../tmp/single_ingest
+   ```
+3) 安装依赖：
    ```bash
    cd etl_billiards
    pip install -r requirements.txt
-   ```  
-3) 一键 ODS→DWD→质检：  
+   ```
+4) 回放入库（ODS）→ 装载 DWD → 质检（可用 `--ingest-source` 覆盖 `INGEST_SOURCE_DIR`）：
    ```bash
-   python -m etl_billiards.cli.main --tasks INIT_ODS_SCHEMA,INIT_DWD_SCHEMA --pipeline-flow INGEST_ONLY
-   python -m etl_billiards.cli.main --tasks MANUAL_INGEST --pipeline-flow INGEST_ONLY --ingest-source "C:\dev\LLTQ\export\test-json-doc"
-   python -m etl_billiards.cli.main --tasks DWD_LOAD_FROM_ODS --pipeline-flow INGEST_ONLY
-   python -m etl_billiards.cli.main --tasks DWD_QUALITY_CHECK --pipeline-flow INGEST_ONLY
-   # 报表：etl_billiards/reports/dwd_quality_report.json
+   python -m cli.main --pipeline-flow INGEST_ONLY --tasks INIT_ODS_SCHEMA,INIT_DWD_SCHEMA
+   python -m cli.main --pipeline-flow INGEST_ONLY --tasks MANUAL_INGEST --ingest-source "../tmp/single_ingest"
+   python -m cli.main --pipeline-flow INGEST_ONLY --tasks DWD_LOAD_FROM_ODS
+   python -m cli.main --pipeline-flow INGEST_ONLY --tasks DWD_QUALITY_CHECK
+   # 报表：reports/dwd_quality_report.json
    ```
 
-## 目录与文件作用
-- 根目录：`etl_billiards/` 主代码；`requirements.txt` 依赖；`run_etl.sh/.bat` 启动脚本；`.env/.env.example` 配置；`tmp/` 存放草稿/调试/备份。  
-- etl_billiards/ 主线目录  
-  - `config/`：`defaults.py` 默认值，`env_parser.py` 解析 .env，`settings.py` 统一配置加载。  
-  - `api/`：`client.py` HTTP 请求、重试与分页。  
-  - `database/`：`connection.py` 连接封装，`operations.py` 批量 upsert，DDL：`schema_ODS_doc.sql`、`schema_dwd_doc.sql`。  
-  - `tasks/`：业务任务  
-    - `init_schema_task.py`：INIT_ODS_SCHEMA / INIT_DWD_SCHEMA。  
-    - `manual_ingest_task.py`：示例 JSON → ODS。  
-    - `dwd_load_task.py`：ODS → DWD（映射、SCD2/事实增量）。  
-    - 其他任务按需扩展。  
-  - `loaders/`：ODS/DWD/SCD2 Loader 实现。  
-  - `scd/`：`scd2_handler.py` 处理维度 SCD2 历史。  
-  - `quality/`：质量检查器（行数/金额对照）。  
-  - `orchestration/`：`scheduler.py` 调度；`task_registry.py` 任务注册；`run_tracker.py` 运行记录。  
-  - `scripts/`：重建/测试/探活工具。  
-  - `docs/`：`ods_to_dwd_mapping.md` 映射说明，`ods_sample_json.md` 示例 JSON 说明，`dwd_quality_check.md` 质检说明。  
-  - `reports/`：质检输出（如 `dwd_quality_report.json`）。  
-  - `tests/`：单元/集成测试；`utils/`：通用工具。  
-  - `backups/`（若存在）：关键文件备份。
+> 可按需单独运行：  
+> - 仅建表：`python -m cli.main --tasks INIT_ODS_SCHEMA`  
+> - 仅 ODS 灌入：`python -m cli.main --tasks MANUAL_INGEST`  
+> - 仅 DWD 装载：`python -m cli.main --tasks INIT_DWD_SCHEMA,DWD_LOAD_FROM_ODS`
 
-## 业务流程与文件关系
-1) 调度入口：`cli/main.py` 解析 CLI → `orchestration/scheduler.py` 依 `task_registry.py` 创建任务 → 初始化 DB/API/Config 上下文。  
-2) ODS：`init_schema_task.py` 执行 `schema_ODS_doc.sql` 建表；`manual_ingest_task.py` 从 `INGEST_SOURCE_DIR` 读 JSON，批量 upsert ODS。  
-3) DWD：`init_schema_task.py` 执行 `schema_dwd_doc.sql` 建表；`dwd_load_task.py` 依据 `TABLE_MAP/FACT_MAPPINGS` 从 ODS 清洗写入 DWD，维度走 SCD2（`scd/scd2_handler.py`），事实按时间/水位增量。  
-4) 质检：质量任务读取 ODS/DWD，统计行数/金额，输出 `reports/dwd_quality_report.json`。  
-5) 配置：`config/defaults.py` + `.env` + CLI 参数叠加；HTTP（如启用在线）走 `api/client.py`；DB 访问走 `database/connection.py`。  
-6) 文档：`docs/ods_to_dwd_mapping.md` 记录字段映射；`docs/ods_sample_json.md` 描述示例数据结构，便于对照调试。
+> Windows：可用 `etl_billiards/run_ods.bat` 一键执行 ODS 建表 + 灌入示例 JSON（`INIT_ODS_SCHEMA` + `MANUAL_INGEST`）。
+
+## 正式环境（在线抓取 → 更新 ODS → 更新 DWD）
+
+**核心入口 CLI（推荐在 `etl_billiards/` 目录执行）**
+- `python -m cli.main`
+
+### 必备配置（建议通过环境变量或 `.env`）
+- 数据库：`PG_DSN`、`STORE_ID`
+- 在线抓取：`API_TOKEN`（可选 `API_BASE`、`API_TIMEOUT`、`API_PAGE_SIZE`、`API_RETRY_MAX`）
+- 输出目录（可选）：`EXPORT_ROOT`、`LOG_ROOT`、`FETCH_ROOT`/`JSON_FETCH_ROOT`
+
+**安全提示**：建议将数据库/Token 等凭证保存在 `.env` 或受控秘钥管理中，生产环境使用最小权限账号。
+
+### 推荐定时方式 A（两段定时，更清晰）
+1) **更新 ODS（在线抓取 + 入库，FULL）**
+   ```bash
+   python -m cli.main \
+     --pipeline-flow FULL \
+     --tasks PRODUCTS,TABLES,MEMBERS,ASSISTANTS,PACKAGES_DEF,ORDERS,PAYMENTS,REFUNDS,COUPON_USAGE,INVENTORY_CHANGE,TOPUPS,TABLE_DISCOUNT,ASSISTANT_ABOLISH,LEDGER \
+     --pg-dsn "$PG_DSN" --store-id "$STORE_ID" \
+     --api-token "$API_TOKEN"
+   ```
+2) **ODS → DWD（将新增/变更同步到 DWD）**
+   ```bash
+   python -m cli.main \
+     --pipeline-flow INGEST_ONLY \
+     --tasks DWD_LOAD_FROM_ODS \
+     --pg-dsn "$PG_DSN" --store-id "$STORE_ID"
+   ```
+
+### 推荐定时方式 B（一条命令串起来）
+同一条命令先跑在线抓取/入库任务，再跑 DWD 装载任务：
+```bash
+python -m cli.main \
+  --pipeline-flow FULL \
+  --tasks PRODUCTS,TABLES,MEMBERS,ASSISTANTS,PACKAGES_DEF,ORDERS,PAYMENTS,REFUNDS,COUPON_USAGE,INVENTORY_CHANGE,TOPUPS,TABLE_DISCOUNT,ASSISTANT_ABOLISH,LEDGER,DWD_LOAD_FROM_ODS \
+  --pg-dsn "$PG_DSN" --store-id "$STORE_ID" \
+  --api-token "$API_TOKEN"
+```
+
+### `pipeline-flow` 说明
+- `FULL`：在线抓取落盘 + 本地清洗入库（ODS 任务会走抓取；`DWD_LOAD_FROM_ODS` 仅走入库阶段）
+- `FETCH_ONLY`：仅在线抓取落盘，不入库
+- `INGEST_ONLY`：仅从本地 JSON 回放入库（适合离线回放/补跑）
+
+## 目录结构与关键文件
+- 仓库根目录：`etl_billiards/` 主代码；`app/` 示例 runner；`开发笔记/` 项目笔记；`tmp/` 草稿/调试归档；`requirements.txt`（仓库根）依赖；`run_etl.sh/.bat` 启动脚本。  
+  - 注意：根目录的 `run_etl.sh/.bat` 运行时要求当前目录为 `etl_billiards/`（因为入口是 `python -m cli.main`）。
+- `etl_billiards/`（主代码目录）
+  - `.env`：本地配置文件（可选，用环境变量也可）
+  - `cli/`：CLI 入口（`cli/main.py`）
+  - `config/`：`defaults.py` 默认值；`env_parser.py` 解析 `.env`/环境变量；`settings.py` AppConfig 加载/校验
+  - `api/`：`client.py` HTTP 请求、重试、分页与落盘
+  - `database/`：`connection.py` 连接封装；`operations.py` 批量 upsert；DDL：`schema_ODS_doc.sql`、`schema_dwd_doc.sql`
+  - `tasks/`：业务任务（ODS 抓取/回放、DWD 装载、质检等）
+  - `loaders/`：ODS/DWD/SCD2 Loader 实现
+  - `scd/`：`scd2_handler.py`（维度 SCD2 历史）
+  - `quality/`：质量检查器（行数/金额对照）
+  - `orchestration/`：`scheduler.py` 调度；`task_registry.py` 注册；`cursor_manager.py` 水位管理；`run_tracker.py` 运行记录
+  - `scripts/`：重建/测试/探活工具
+  - `docs/`：映射/样例/质检说明文档
+  - `fetch-test/`：接口联调/规则验证的一次性脚本与报告（不影响主流程）
+  - `reports/`：质检输出（如 `dwd_quality_report.json`）
+  - `tests/`：单元/集成测试
+
+## 项目文件索引（维护/AI 快速定位）
+> 说明：用于维护/AI 快速定位文件路径与用途；默认不列出 `.git/`、`__pycache__/`、`.pytest_cache/`、`*.pyc` 等自动生成内容。
+
+### /
+- `.gitignore`：Git 忽略规则。
+- `.gitkeep`：占位文件（用于保留空目录）。
+- `README.md`：项目总览与使用说明（本文档）。
+- `requirements.txt`：仓库根依赖清单（不含版本约束，建议优先用 `etl_billiards/requirements.txt`）。
+- `run_etl.bat`：Windows 启动脚本（需先 `cd etl_billiards`；入口为 `python -m cli.main`）。
+- `run_etl.sh`：Linux/macOS 启动脚本（需先 `cd etl_billiards`；会加载当前目录 `.env`）。
+
+### app/
+- `app/etl_busy.py`：忙时 ETL 示例函数（TODO，占位）。
+- `app/etl_idle.py`：闲时 ETL 示例函数（TODO，占位）。
+- `app/runner.py`：简易 Runner：按 `--mode busy/idle` 调用 `app/etl_busy.py` 或 `app/etl_idle.py`（示例/未接入主 ETL）。
+
+### etl_billiards/
+- `etl_billiards/.env`：本地运行环境变量（含敏感信息，勿提交/勿外传）。
+- `etl_billiards/__init__.py`：Python 包标记文件。
+- `etl_billiards/ods_row_report.json`：ODS 行数对照报告（source json vs ODS 表）。
+- `etl_billiards/requirements.txt`：ETL 运行依赖（带最低版本约束）。
+- `etl_billiards/run_ods.bat`：Windows 一键脚本：重建 ODS 并灌入示例 JSON。
+- `etl_billiards/setup.py`：打包/安装脚本（当前项目主要按“`cd etl_billiards; python -m cli.main`”方式运行）。
+
+### etl_billiards/api/
+- `etl_billiards/api/__init__.py`：Python 包标记文件。
+- `etl_billiards/api/client.py`：API客户端：统一封装 POST/重试/分页与列表提取逻辑。
+- `etl_billiards/api/endpoint_routing.py`：“近期记录 / 历史记录(Former)”接口路由规则。
+- `etl_billiards/api/local_json_client.py`：本地 JSON 客户端，模拟 APIClient 的分页接口，从落盘的 JSON 回放数据。
+- `etl_billiards/api/recording_client.py`：包装 APIClient，将分页响应落盘便于后续本地清洗。
+
+### etl_billiards/cli/
+- `etl_billiards/cli/__init__.py`：Python 包标记文件。
+- `etl_billiards/cli/main.py`：CLI主入口
+
+### etl_billiards/config/
+- `etl_billiards/config/__init__.py`：Python 包标记文件。
+- `etl_billiards/config/defaults.py`：配置默认值定义
+- `etl_billiards/config/env_parser.py`：环境变量解析
+- `etl_billiards/config/settings.py`：配置管理主类
+
+### etl_billiards/database/
+- `etl_billiards/database/__init__.py`：Python 包标记文件。
+- `etl_billiards/database/base.py`：数据库操作（批量、RETURNING支持）
+- `etl_billiards/database/connection.py`：Database connection manager with capped connect_timeout.
+- `etl_billiards/database/operations.py`：数据库批量操作
+- `etl_billiards/database/schema_dwd_doc.sql`：DWD Schema DDL（含字段/注释/口径说明）。
+- `etl_billiards/database/schema_etl_admin.sql`：etl_admin 元数据 Schema DDL（任务/水位/运行记录等）。
+- `etl_billiards/database/schema_ODS_doc.sql`：ODS Schema DDL（含字段/注释/口径说明）。
+- `etl_billiards/database/seed_ods_tasks.sql`：SQL 种子脚本：初始化/注册 ODS 任务。
+- `etl_billiards/database/seed_scheduler_tasks.sql`：SQL 种子脚本：初始化调度任务配置。
+
+### etl_billiards/database/Deleded & backup/
+- （本目录无直接文件）
+
+### etl_billiards/docs/
+- `etl_billiards/docs/dwd_main_tables_dictionary.md`：DWD 主表（非 Ex）表格说明书
+- `etl_billiards/docs/在线抓取，更新ODS ，然后将更新的ODS内容，对应到DWD的更新。.md`：在线抓取，更新ODS ，然后将更新的ODS内容，对应到DWD的更新。
+
+### etl_billiards/fetch-test/
+- `etl_billiards/fetch-test/compare_recent_former_endpoints.py`：对比“近期记录”与“历史记录(Former)”接口：
+- `etl_billiards/fetch-test/README.md`：fetch-test
+- `etl_billiards/fetch-test/recent_vs_former_report.json`：报告/比对输出（JSON）。
+- `etl_billiards/fetch-test/recent_vs_former_report.md`：近期记录 vs 历史记录(Former) 接口对比报告
+
+### etl_billiards/loaders/
+- `etl_billiards/loaders/__init__.py`：Python 包标记文件。
+- `etl_billiards/loaders/base_loader.py`：数据加载器基类
+
+### etl_billiards/loaders/dimensions/
+- `etl_billiards/loaders/dimensions/__init__.py`：Python 包标记文件。
+- `etl_billiards/loaders/dimensions/assistant.py`：助教维度加载器
+- `etl_billiards/loaders/dimensions/member.py`：会员维度表加载器
+- `etl_billiards/loaders/dimensions/package.py`：团购/套餐定义加载器
+- `etl_billiards/loaders/dimensions/product.py`：商品维度 + 价格SCD2 加载器
+- `etl_billiards/loaders/dimensions/table.py`：台桌维度加载器
+
+### etl_billiards/loaders/facts/
+- `etl_billiards/loaders/facts/__init__.py`：Python 包标记文件。
+- `etl_billiards/loaders/facts/assistant_abolish.py`：助教作废事实表
+- `etl_billiards/loaders/facts/assistant_ledger.py`：助教流水事实表
+- `etl_billiards/loaders/facts/coupon_usage.py`：券核销事实表
+- `etl_billiards/loaders/facts/inventory_change.py`：库存变动事实表
+- `etl_billiards/loaders/facts/order.py`：订单事实表加载器
+- `etl_billiards/loaders/facts/payment.py`：支付事实表加载器
+- `etl_billiards/loaders/facts/refund.py`：退款事实表加载器
+- `etl_billiards/loaders/facts/table_discount.py`：台费打折事实表
+- `etl_billiards/loaders/facts/ticket.py`：小票详情加载器
+- `etl_billiards/loaders/facts/topup.py`：充值记录事实表
+
+### etl_billiards/loaders/ods/
+- `etl_billiards/loaders/ods/__init__.py`：Python 包标记文件。
+- `etl_billiards/loaders/ods/generic.py`：Generic ODS loader that keeps raw payload + primary keys.
+
+### etl_billiards/models/
+- `etl_billiards/models/__init__.py`：Python 包标记文件。
+- `etl_billiards/models/parsers.py`：数据类型解析器
+- `etl_billiards/models/validators.py`：数据验证器
+
+### etl_billiards/orchestration/
+- `etl_billiards/orchestration/__init__.py`：Python 包标记文件。
+- `etl_billiards/orchestration/cursor_manager.py`：游标管理器
+- `etl_billiards/orchestration/run_tracker.py`：运行记录追踪器
+- `etl_billiards/orchestration/scheduler.py`：ETL 调度：支持在线抓取、离线清洗入库、全流程三种模式。
+- `etl_billiards/orchestration/task_registry.py`：任务注册表
+
+### etl_billiards/quality/
+- `etl_billiards/quality/__init__.py`：Python 包标记文件。
+- `etl_billiards/quality/balance_checker.py`：余额一致性检查器
+- `etl_billiards/quality/base_checker.py`：数据质量检查器基类
+
+### etl_billiards/reports/
+- `etl_billiards/reports/dwd_quality_report.json`：DWD 质量核对输出（行数/金额对照）。
+
+### etl_billiards/scd/
+- `etl_billiards/scd/__init__.py`：Python 包标记文件。
+- `etl_billiards/scd/scd2_handler.py`：SCD2 (Slowly Changing Dimension Type 2) 处理逻辑
+
+### etl_billiards/scripts/
+- `etl_billiards/scripts/bootstrap_schema.py`：Apply the PRD-aligned warehouse schema (ODS/DWD/DWS) to PostgreSQL.
+- `etl_billiards/scripts/build_dwd_from_ods.py`：Populate PRD DWD tables from ODS payload snapshots.
+- `etl_billiards/scripts/build_dws_order_summary.py`：Recompute billiards_dws.dws_order_summary from DWD fact tables.
+- `etl_billiards/scripts/check_ods_json_vs_table.py`：ODS JSON 字段核对脚本：对照当前数据库中的 ODS 表字段，检查示例 JSON（默认目录 C:\dev\LLTQ\export\test-json-doc）
+- `etl_billiards/scripts/check_ods_gaps.py`：ODS 缺失校验脚本：API 主键 vs ODS 主键逐条比对，输出缺失明细样例。
+- `etl_billiards/scripts/reload_ods_windowed.py`：ODS 窗口化补跑脚本：按时间切片重跑 ODS 任务，并可配置窗口粒度与延时。
+- `etl_billiards/scripts/rebuild_db_and_run_ods_to_dwd.py`：一键重建 ETL 相关 Schema，并执行 ODS → DWD。
+- `etl_billiards/scripts/rebuild_ods_from_json.py`：从本地 JSON 示例目录重建 billiards_ods.* 表，并导入样例数据。
+- `etl_billiards/scripts/run_tests.py`：灵活的测试执行脚本，可像搭积木一样组合不同参数或预置命令（模式/数据库/归档路径等），
+- `etl_billiards/scripts/Temp1.py`：空 Python 文件（占位/临时）。
+- `etl_billiards/scripts/test_db_connection.py`：Quick utility for validating PostgreSQL connectivity (ASCII-only output).
+- `etl_billiards/scripts/test_presets.py`：测试命令仓库：集中维护 run_tests.py 的常用组合，支持一键执行。
+
+### etl_billiards/scripts/Deleded & backup/
+- （本目录无直接文件）
+
+### etl_billiards/tasks/
+- `etl_billiards/tasks/__init__.py`：Python 包标记文件。
+- `etl_billiards/tasks/assistant_abolish_task.py`：助教作废任务
+- `etl_billiards/tasks/assistants_task.py`：助教账号任务
+- `etl_billiards/tasks/base_dwd_task.py`：DWD任务基类
+- `etl_billiards/tasks/base_task.py`：ETL任务基类（引入 Extract/Transform/Load 模板方法）
+- `etl_billiards/tasks/coupon_usage_task.py`：平台券核销任务
+- `etl_billiards/tasks/dwd_load_task.py`：DWD 装载任务：从 ODS 增量写入 DWD（维度 SCD2，事实按时间增量）。
+- `etl_billiards/tasks/dwd_quality_task.py`：DWD 质量核对任务：按 dwd_quality_check.md 输出行数/金额对照报表。
+- `etl_billiards/tasks/init_dwd_schema_task.py`：初始化 DWD Schema：执行 schema_dwd_doc.sql，可选先 DROP SCHEMA。
+- `etl_billiards/tasks/init_schema_task.py`：任务：初始化运行环境，执行 ODS 与 etl_admin 的 DDL，并准备日志/导出目录。
+- `etl_billiards/tasks/inventory_change_task.py`：库存变更任务
+- `etl_billiards/tasks/ledger_task.py`：助教流水任务
+- `etl_billiards/tasks/manual_ingest_task.py`：手工示例数据灌入：按 schema_ODS_doc.sql 的表结构写入 ODS。
+- `etl_billiards/tasks/members_dwd_task.py`：DWD Task：Process Member Records from ODS to Dimension Table.
+- `etl_billiards/tasks/members_task.py`：会员ETL任务
+- `etl_billiards/tasks/ods_json_archive_task.py`：在线抓取 ODS 相关接口并落盘为 JSON（用于后续离线回放/入库）。
+- `etl_billiards/tasks/ods_tasks.py`：ODS ingestion tasks.
+- `etl_billiards/tasks/orders_task.py`：订单ETL任务
+- `etl_billiards/tasks/packages_task.py`：团购/套餐定义任务
+- `etl_billiards/tasks/payments_dwd_task.py`：DWD Task：Process Payment Records from ODS to Fact Table.
+- `etl_billiards/tasks/payments_task.py`：支付记录ETL任务
+- `etl_billiards/tasks/products_task.py`：商品档案（PRODUCTS）ETL任务
+- `etl_billiards/tasks/refunds_task.py`：退款记录任务
+- `etl_billiards/tasks/table_discount_task.py`：台费折扣任务
+- `etl_billiards/tasks/tables_task.py`：台桌档案任务
+- `etl_billiards/tasks/ticket_dwd_task.py`：DWD Task：Process Ticket Details from ODS to DWD fact tables.
+- `etl_billiards/tasks/topups_task.py`：充值记录任务
+
+### etl_billiards/tasks/dwd/
+- （本目录无直接文件）
+
+### etl_billiards/tests/
+- `etl_billiards/tests/__init__.py`：Python 包标记文件。
+
+### etl_billiards/tests/integration/
+- `etl_billiards/tests/integration/__init__.py`：Python 包标记文件。
+- `etl_billiards/tests/integration/test_database.py`：数据库集成测试
+
+### etl_billiards/tests/unit/
+- `etl_billiards/tests/unit/__init__.py`：Python 包标记文件。
+- `etl_billiards/tests/unit/task_test_utils.py`：ETL 任务测试的共用辅助模块，涵盖在线/离线模式所需的伪造数据、客户端与配置等工具函数。
+- `etl_billiards/tests/unit/test_config.py`：配置管理测试
+- `etl_billiards/tests/unit/test_endpoint_routing.py`：Unit tests for recent/former endpoint routing.
+- `etl_billiards/tests/unit/test_etl_tasks_offline.py`：离线模式任务测试，通过回放归档 JSON 来验证 T+L 链路可用。
+- `etl_billiards/tests/unit/test_etl_tasks_online.py`：在线模式下的端到端任务测试，验证所有任务在模拟 API 下能顺利执行。
+- `etl_billiards/tests/unit/test_etl_tasks_stages.py`：验证 14 个任务的 E/T/L 分阶段调用（FakeDB/FakeAPI，不访问真实接口或数据库）。
+- `etl_billiards/tests/unit/test_ods_tasks.py`：Unit tests for the new ODS ingestion tasks.
+- `etl_billiards/tests/unit/test_parsers.py`：解析器测试
+- `etl_billiards/tests/unit/test_reporting.py`：汇总与报告工具的单测。
+
+### etl_billiards/utils/
+- `etl_billiards/utils/__init__.py`：Python 包标记文件。
+- `etl_billiards/utils/helpers.py`：通用工具函数
+- `etl_billiards/utils/json_store.py`：JSON 归档/读取的通用工具。
+- `etl_billiards/utils/reporting.py`：简单的任务结果汇总与格式化工具。
+
+### tmp/
+- `tmp/20251121-task.txt`：历史任务/计划记录（可能存在编码问题）。
+- `tmp/doc_extracted.txt`：从 DWD 文档抽取的正文（大文本）。
+- `tmp/doc_lines.txt`：DWD 文档按行抽取/对照（文本）。
+- `tmp/dwd_tables.json`：DWD 表清单（JSON）。
+- `tmp/dwd_tables_full.json`：DWD 表清单（完整版 JSON）。
+- `tmp/hebing.py`：临时脚本：按“同名 key”合并目录内 md+json 输出 merged_output.txt。
+- `tmp/README_FULL.md`：历史/草稿：README 详细版（已合并进根 README）。
+- `tmp/rebuild_run_20251214-042115.log`：运行日志/调试输出（临时文件）。
+- `tmp/rewrite_schema_dwd_doc_comments.py`：临时脚本：批量重写 DWD DDL 注释（归档/草稿）。
+- `tmp/rewrite_schema_ods_doc_comments.py`：临时脚本：批量重写 ODS DDL 注释（归档/草稿）。
+- `tmp/schema_dwd.sql`：DWD schema 草稿/导出（归档）。
+- `tmp/schema_dwd_doc.sql`：DWD schema doc 版本（归档）。
+- `tmp/schema_ODS_doc copy.sql`：ODS schema doc 备份（归档）。
+- `tmp/schema_ODS_doc.sql`：ODS schema doc 版本（归档）。
+- `tmp/temp_chinese.txt`：编码/文本对照测试。
+- `tmp/tmp_debug_sql.py`：临时脚本：调试 SQL/映射（归档）。
+- `tmp/tmp_drop_dwd.py`：临时脚本：DROP SCHEMA billiards_dwd（危险，勿在生产执行）。
+- `tmp/tmp_dwd_tasks.py`：临时脚本：调试 DWD 相关任务（归档）。
+- `tmp/tmp_problems.py`：临时脚本：问题排查记录/复现（归档）。
+- `tmp/tmp_run_sql.py`：临时脚本：拼接/执行一条 INSERT...SELECT 验证映射（需 PG_DSN）。
+- `tmp/非球接口API.md`：上游接口笔记/汇总（草稿/归档）。
+
+### tmp/a/
+- （本目录无直接文件）
+
+### tmp/b/
+- （本目录无直接文件）
+
+### tmp/etl_billiards_misc/
+- `tmp/etl_billiards_misc/0.py`：Simple PostgreSQL connectivity smoke-checker.
+- `tmp/etl_billiards_misc/feiqiu-ETL.code-workspace`：VS Code workspace 文件（归档）。
+- `tmp/etl_billiards_misc/草稿.txt`：草稿/说明（归档）。
+
+### tmp/etl_billiards_misc/backups/
+- `tmp/etl_billiards_misc/backups/manual_ingest_task.py`：历史版本备份（归档）。
+- `tmp/etl_billiards_misc/backups/manual_ingest_task.py.bak_20251209`：历史版本备份（归档）。
+- `tmp/etl_billiards_misc/backups/schema_ODS_doc.sql`：历史版本备份（归档）。
+- `tmp/etl_billiards_misc/backups/schema_ODS_doc.sql.bak_20251209`：历史版本备份（归档）。
+
+### tmp/etl_billiards_misc/tmp & Delete/
+- `tmp/etl_billiards_misc/tmp & Delete/.env.example`：旧示例配置（归档）。
+- `tmp/etl_billiards_misc/tmp & Delete/dwd_schema_columns.txt`：DWD 字段提取/对照文本（归档）。
+- `tmp/etl_billiards_misc/tmp & Delete/DWD层设计建议.docx`：DWD 设计建议文档（归档）。
+- `tmp/etl_billiards_misc/tmp & Delete/DWD层设计草稿.md`：DWD 设计草稿（归档）。
+- `tmp/etl_billiards_misc/tmp & Delete/schema_dwd_doc.sql.bak`：schema 备份（归档）。
+- `tmp/etl_billiards_misc/tmp & Delete/schema_ODS_doc.sql.bak`：schema 备份（归档）。
+- `tmp/etl_billiards_misc/tmp & Delete/schema_ODS_doc.sql.rewrite2.bak`：schema 重写过程备份（归档）。
+- `tmp/etl_billiards_misc/tmp & Delete/schema_v2.sql`：schema v2 草稿（归档）。
+
+### tmp/recharge_only/
+- `tmp/recharge_only/recharge_settlements.json`：离线样例 JSON（仅充值结算）。
+
+### tmp/single_ingest/
+- `tmp/single_ingest/goods_stock_movements.json`：离线最小样例 JSON（单文件）。
+
+### 开发笔记/
+- `开发笔记/记录.md`：开发/迁移过程的备忘与待办（归档）。
+
+## 架构与流程
+执行链路（控制流）：
+1) CLI（`cli/main.py`）解析参数 → 生成 AppConfig → 初始化日志/DB/API；  
+2) 调度层（`orchestration/scheduler.py`）按 `task_registry.py` 实例化任务，设置 run_uuid、cursor（水位）、上下文；  
+3) 任务执行模板：获取时间窗口/水位（`cursor_manager.py`）→ Extract（API 分页/重试或离线读 JSON）→ Transform（解析/校验）→ Load（Loader 批量 upsert/SCD2/增量写入，底层 `database/operations.py`）→（可选）质量检查 → 更新水位与运行记录（`run_tracker.py`），提交/回滚事务。
+
+数据流与依赖：
+- 配置：`config/defaults.py` + `.env`/环境变量 + CLI 参数叠加
+- 在线：`api/client.py` 支撑分页/重试；可落盘 JSON（`pipeline.fetch_root`）
+- 离线：`manual_ingest_task.py` 从 `INGEST_SOURCE_DIR` 回放入库
+- DWD：`dwd_load_task.py` 依据 `TABLE_MAP/FACT_MAPPINGS` 映射装载，维度走 SCD2，事实走增量
+- 质检：`dwd_quality_task.py` 输出 `reports/dwd_quality_report.json`
+
+## ODS → DWD 策略与建模要点
+1) ODS 留底：保留源主键、payload、时间/来源信息，便于回溯。  
+2) DWD 清洗：维度走 SCD2，事实按时间/水位增量；字段类型、单位、枚举标准化，同时保留溯源字段。  
+3) 颗粒一致：一张 DWD 表只承载一种业务事件/颗粒，避免混颗粒。  
+4) 业务键统一：site_id、member_id、table_id、order_settle_id、order_trade_no 等统一命名。  
+5) 不过度汇总：DWD 只做明细/轻度清洗，聚合留到 DWS/报表。  
+6) 去嵌套：数组展开为子表/子行，重复 profile 提炼为维度。  
+7) 长期演进：优先加列/加表，减少对已有表结构的破坏。
+
+## 常用 CLI
+```bash
+cd etl_billiards
+
+# 运行 defaults.py 中的默认任务列表（在线 FULL 流程）
+python -m cli.main --pg-dsn "$PG_DSN" --store-id "$STORE_ID" --api-token "$API_TOKEN"
+
+# 运行指定任务
+python -m cli.main --tasks INIT_ODS_SCHEMA,MANUAL_INGEST --pipeline-flow INGEST_ONLY --ingest-source "../tmp/single_ingest"
+
+# 覆盖 DSN / API / 输出目录
+python -m cli.main --pg-dsn "postgresql://user:pwd@host:5432/db" --store-id 123 --api-token "..." --fetch-root "./json_fetch"
+
+# 试运行（不写库）
+python -m cli.main --dry-run --tasks DWD_LOAD_FROM_ODS
+```
+
+## 测试
+说明：仓库未固定 pytest 版本（运行测试需自行安装 `pytest`）。
+
+```bash
+cd etl_billiards
+pip install pytest
+
+# 单元测试（模拟 API + FakeDB）
+pytest tests/unit
+
+# 集成测试（需要设置 TEST_DB_DSN）
+TEST_DB_DSN="postgresql://user:pwd@host:5432/db" pytest tests/integration/test_database.py
+
+# 便捷测试执行器（可选）
+python scripts/run_tests.py --suite online -k ORDERS
+python scripts/test_db_connection.py --dsn "postgresql://user:pwd@host:5432/db" --query "SELECT 1"
+```
+
+## 开发与扩展
+- 新任务：在 `tasks/` 继承 BaseTask，实现 `get_task_code/execute`，并在 `orchestration/task_registry.py` 注册。  
+- 新 Loader/Checker：参考 `loaders/`、`quality/`，复用批量 upsert/质检接口。  
+- 新配置项：在 `config/defaults.py` 增加默认值，并在 `config/env_parser.py` 增加环境变量映射（如需要）。
+
+## ODS 任务上线指引
+- 元数据/任务注册脚本：
+  - `etl_billiards/database/seed_ods_tasks.sql`
+  - `etl_billiards/database/seed_scheduler_tasks.sql`
+- 确认 `etl_admin.etl_task` 中已启用所需任务（不同环境需替换 store_id / schema）。  
+- 离线回放/重建 ODS（开发/运维）：  
+  ```bash
+  cd etl_billiards
+  python scripts/rebuild_ods_from_json.py --dsn "$PG_DSN" --json-dir "C:\\path\\to\\json-doc"
+  ```
+
+## ODS 表概览（数据路径）
+| ODS 表名                           | 接口 Path                                         | 数据列表路径                  |
+| ---------------------------------- | ------------------------------------------------- | ----------------------------- |
+| assistant_accounts_master          | /PersonnelManagement/SearchAssistantInfo          | data.assistantInfos           |
+| assistant_service_records          | /AssistantPerformance/GetOrderAssistantDetails    | data.orderAssistantDetails    |
+| assistant_cancellation_records     | /AssistantPerformance/GetAbolitionAssistant       | data.abolitionAssistants      |
+| goods_stock_movements              | /GoodsStockManage/QueryGoodsOutboundReceipt       | data.queryDeliveryRecordsList |
+| goods_stock_summary                | /TenantGoods/GetGoodsStockReport                  | data                          |
+| group_buy_packages                 | /PackageCoupon/QueryPackageCouponList             | data.packageCouponList        |
+| group_buy_redemption_records       | /Site/GetSiteTableUseDetails                      | data.siteTableUseDetailsList  |
+| member_profiles                    | /MemberProfile/GetTenantMemberList                | data.tenantMemberInfos        |
+| member_balance_changes             | /MemberProfile/GetMemberCardBalanceChange         | data.tenantMemberCardLogs     |
+| member_stored_value_cards          | /MemberProfile/GetTenantMemberCardList            | data.tenantMemberCards        |
+| payment_transactions               | /PayLog/GetPayLogListPage                         | data                          |
+| platform_coupon_redemption_records | /Promotion/GetOfflineCouponConsumePageList        | data                          |
+| recharge_settlements               | /Site/GetRechargeSettleList                       | data.settleList               |
+| refund_transactions                | /Order/GetRefundPayLogList                        | data                          |
+| settlement_records                 | /Site/GetAllOrderSettleList                       | data.settleList               |
+| settlement_ticket_details          | /Order/GetOrderSettleTicketNew                    | 完整 JSON                     |
+| site_tables_master                 | /Table/GetSiteTables                              | data.siteTables               |
+| stock_goods_category_tree          | /TenantGoodsCategory/QueryPrimarySecondaryCategory| data.goodsCategoryList        |
+| store_goods_master                 | /TenantGoods/GetGoodsInventoryList                | data.orderGoodsList           |
+| store_goods_sales_records          | /TenantGoods/GetGoodsSalesList                    | data.orderGoodsLedgers        |
+| table_fee_discount_records         | /Site/GetTaiFeeAdjustList                         | data.taiFeeAdjustInfos        |
+| table_fee_transactions             | /Site/GetSiteTableOrderDetails                    | data.siteTableUseDetailsList  |
+| tenant_goods_master                | /TenantGoods/QueryTenantGoods                     | data.tenantGoodsList          |
+
+> 完整字段级映射见 `etl_billiards/docs/` 与 ODS/DWD DDL。
 
 ## 当前状态（2025-12-09）
-- 示例 JSON 全量灌入；DWD 行数与 ODS 对齐。  
+- 示例 JSON 已全量灌入，DWD 行数与 ODS 对齐。  
 - 分类维度已展平大类+子类：`dim_goods_category` 26 行（category_level/leaf 已赋值）。  
-- 剩余空值多因源数据为空；补值需先确认上游是否提供。
+- 部分空字段源数据即为空，如需补值请先确认上游。
 
 ## 可精简/归档
-- `tmp/`、`tmp/etl_billiards_misc/` 中的草稿、旧备份、调试脚本仅供参考，运行不依赖。  
-- 根级保留必要文件（README、requirements、run_etl.*、.env/.env.example），其余临时文件已移至 tmp。
+- `tmp/`、`tmp/etl_billiards_misc/` 中草稿、旧备份、调试脚本仅供参考，不影响运行。  
+- 根级保留必要文件（README、requirements、run_etl.*），其余临时文件按需归档至 `tmp/`。
+
+## 一键更新（推荐）
+
+日常需要把数据从 ODS 更新到最新，并同步刷新 DWD/DWS 时，直接运行一键脚本：
+
+```bash
+cd etl_billiards
+python run_update.py
+```
+
+常用参数：
+- `--overlap-seconds 3600`：冗余抓取窗口（默认 3600 秒）
+- `--dws-rebuild-days 1`：DWS 回算冗余天数（默认 1 天）
+- `--dws-start YYYY-MM-DD --dws-end YYYY-MM-DD`：手工指定 DWS 回算日期范围
+- `--skip-ods`：跳过 ODS 在线抓取（仅跑 DWD/DWS）
+- `--ods-tasks ODS_PAYMENT,ODS_TABLE_USE,...`：只跑指定 ODS 任务
+- `--check-ods-gaps`：在 ODS 更新完成后执行缺失校验（API 主键 vs ODS 主键）
+- `--check-ods-overlap-hours 24`：缺失校验时，从 ODS 最新截止时间回溯的小时数（默认 24）
+- `--check-ods-window-days 1`：缺失校验 API 窗口粒度（默认 1 天）
+- `--check-ods-page-size 200`：缺失校验 API 每页大小（默认 200）
+- `--check-ods-timeout-sec 1800`：缺失校验步骤超时秒数（默认 1800）
+- `--check-ods-task-codes ODS_PAYMENT,ODS_TABLE_USE,...`：仅校验指定 ODS 任务
+
+### ODS 缺失校验（API vs ODS）
+
+说明：
+- 校验口径为 ODS 表 `MAX(fetched_at)` 的最小值，视为“最新一致截止时间”。  
+- `--from-cutoff` 会从该截止时间回溯 N 小时（默认 24 小时）到当前，便于日常增量校验。  
+
+全量校验（从 2025-07 至今）：
+```bash
+cd etl_billiards
+python scripts/check_ods_gaps.py --start 2025-07-01
+```
+
+更新时校验（从 ODS 最新截止时间回溯 24h）：
+```bash
+cd etl_billiards
+python run_update.py --check-ods-gaps
+```
+
+## FAQ
+- 字段空值：若映射已存在且源列非空仍为空，再检查上游 JSON；维度 SCD2 按全量合并。  
+- DSN/路径：确认 `PG_DSN`、`STORE_ID`、`INGEST_SOURCE_DIR` 与本地一致。  
+- 新增任务：在 `tasks/` 实现并注册到 `task_registry.py`，必要时同步更新 DDL 与映射。  
+- 权限/运行：检查网络、账号权限；脚本需执行权限（如 `chmod +x run_etl.sh`）。
+
+---
+
+## Cutoff（截止时间）检查
+
+当你需要“上次数据截止到什么时候”“现在应该从哪里开始补跑”时，使用任务 `CHECK_CUTOFF`：
+
+```bash
+cd etl_billiards
+python -m cli.main --pipeline-flow INGEST_ONLY --tasks CHECK_CUTOFF
+```
+
+它会输出：
+- `etl_admin.etl_cursor`：每个任务的 `last_start/last_end/last_run_id`（调度游标）
+- ODS：对 `DWD_LOAD_FROM_ODS` 依赖的各个 `billiards_ods.*` 表做 `MAX(fetched_at)`（真实已入库 ODS 的截止）
+- DWD/DWS：输出若干关键表的最大业务时间/最大更新时刻，便于快速核对
+
+> 如果 `etl_cursor.last_end` 很新，但 ODS 的 `MAX(fetched_at)` 很旧，通常表示在线抓取没跑通（最常见是 `API_TOKEN` 过期导致 401）。
+
+## 冗余抓取方案（推荐）
+
+为避免边界时间丢数（上游延迟写入、接口分页抖动、窗口切换等），建议在 cutoff 基础上向前追加 **1 小时** 冗余量：
+
+- 配置：将 `OVERLAP_SECONDS` 设为 3600（默认 120 秒）
+
+```env
+# etl_billiards/.env
+OVERLAP_SECONDS=3600
+```
+
+冗余方案的关键点是“重抓不重落”，依靠各层的去重/幂等机制只落新数据：
+- **ODS 层**：主键/冲突列 UPSERT（重复抓取只会 upsert，不会重复插入）
+- **DWD 层**：事实表增量插入 + 主键冲突不重复落（重复范围会被跳过），维度表按 SCD2 合并
+- **DWS 层**：对指定日期窗口先 delete 再 upsert（窗口内重算幂等）
+
+> 如果你希望“冗余窗口内的数据发生变更也要覆盖更新”，需要把对应层的冲突策略从 `DO NOTHING` 调整为 `DO UPDATE`（当前实现以“只落新数据”为主）。
+
+## DWS（汇总层）入库
+
+本项目已包含 `billiards_dws` 汇总层（当前提供 `dws_order_summary`）：
+
+1) 初始化 DWS 表结构：
+
+```bash
+python -m cli.main --pipeline-flow INGEST_ONLY --tasks INIT_DWS_SCHEMA
+```
+
+2) 生成/刷新汇总表（按窗口重算，建议配合 `--window-start/--window-end`）：
+
+```bash
+python -m cli.main --pipeline-flow INGEST_ONLY --tasks DWS_BUILD_ORDER_SUMMARY \
+  --window-start "2025-10-01 00:00:00" \
+  --window-end "2025-12-26 23:59:59"
+```
+
+3) 推荐串联（ODS -> DWD -> DWS）：
+
+```bash
+# 先跑在线 ODS 抓取（需要有效 API_TOKEN；如果出现 401 请更新 token）
+python -m cli.main --pipeline-flow FULL --tasks ODS_MEMBER,ODS_PAYMENT,ODS_REFUND,ODS_SETTLEMENT_RECORDS
+
+# 再把 ODS 增量同步到 DWD
+python -m cli.main --pipeline-flow INGEST_ONLY --tasks DWD_LOAD_FROM_ODS
+
+# 最后重算 DWS
+python -m cli.main --pipeline-flow INGEST_ONLY --tasks DWS_BUILD_ORDER_SUMMARY
+```
+
+
+## 日志 (UTF-8)
+- 默认日志目录：`etl_billiards/logs/`
+- 每次运行都会生成一个带有时间戳的 `.log` 文件，以便于使用外部工具查看。
+常用选项：
+
+- `--log-file` 自定义日志路径（覆盖默认值）。
+- `--log-dir` 自定义日志目录。
+- `--log-level` 日志级别（`INFO`/`DEBUG`）。
+- `--no-log-console` 禁用控制台日志记录（仅写入文件）。
+
+示例（按桌、按天设置窗口）：
+```bash
+cd etl_billiards
+python scripts/check_ods_gaps.py --start 2025-07-01 --window-days 1 --task-codes ODS_PAYMENT --sleep-per-window-seconds 0.5
+python scripts/reload_ods_windowed.py --tasks ODS_PAYMENT,ODS_TABLE_USE --start 2025-07-01 --window-days 1 --sleep-seconds 1
+python run_update.py --check-ods-gaps --check-ods-window-days 1 --check-ods-sleep-per-window-seconds 0.5
+```

Untitled

@@ -0,0 +1 @@
+README.md

etl_billiards/.env

@@ -12,9 +12,9 @@ STORE_ID=2790685415443269
 TIMEZONE=Asia/Taipei
 
 # API 基础地址，config/env_parser.py -> api.base_url，FETCH 类任务调用
-API_BASE=https://api.example.com
+API_BASE=https://pc.ficoo.vip/apiprod/admin/v1/
 # API 鉴权 Token，config/env_parser.py -> api.token，FETCH 类任务调用
-API_TOKEN=your_token_here
+API_TOKEN=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJjbGllbnQtdHlwZSI6IjQiLCJ1c2VyLXR5cGUiOiIxIiwiaHR0cDovL3NjaGVtYXMubWljcm9zb2Z0LmNvbS93cy8yMDA4LzA2L2lkZW50aXR5L2NsYWltcy9yb2xlIjoiMTIiLCJyb2xlLWlkIjoiMTIiLCJ0ZW5hbnQtaWQiOiIyNzkwNjgzMTYwNzA5OTU3Iiwibmlja25hbWUiOiLnp5_miLfnrqHnkIblkZjvvJrmganmgakxIiwic2l0ZS1pZCI6IjAiLCJtb2JpbGUiOiIxMzgxMDUwMjMwNCIsInNpZCI6IjI5NTA0ODk2NTgzOTU4NDUiLCJzdGFmZi1pZCI6IjMwMDk5MTg2OTE1NTkwNDUiLCJvcmctaWQiOiIwIiwicm9sZS10eXBlIjoiMyIsInJlZnJlc2hUb2tlbiI6IlpWV3grVThBc2FYekFJeTRiaXF6MktwNjMxbTFNRlozV3pLaXNjOHREY289IiwicmVmcmVzaEV4cGlyeVRpbWUiOiIyMDI2LzEvMTcg5LiL5Y2INDoyMjo1OSIsIm5lZWRDaGVja1Rva2VuIjoiZmFsc2UiLCJleHAiOjE3Njg2MzgxNzksImlzcyI6InRlc3QiLCJhdWQiOiJVc2VyIn0.PVeAIx1iHqmHRNaQ4OMFPuOlHBoE47bR5TGJjZP-eOE
 # API 请求超时秒，config/env_parser.py -> api.timeout_sec
 API_TIMEOUT=20
 # API 分页大小，config/env_parser.py -> api.page_size
@@ -40,6 +40,10 @@ PIPELINE_FLOW=FULL
 # 指定任务列表（逗号分隔，覆盖默认），config/env_parser.py -> run.tasks
 # RUN_TASKS=INIT_ODS_SCHEMA,MANUAL_INGEST
 
+# 固定回溯窗口（可选）：同时设置 WINDOW_START + WINDOW_END，将覆盖游标/当前时间窗口
+# WINDOW_START=2025-07-01 00:00:00
+# WINDOW_END=2025-08-01 00:00:00
+
 # 窗口/补偿参数，config/env_parser.py -> run.*
 OVERLAP_SECONDS=120
 WINDOW_BUSY_MIN=30

etl_billiards/api/client.py

@@ -8,6 +8,8 @@ import requests
 from requests.adapters import HTTPAdapter
 from urllib3.util.retry import Retry
 
+from api.endpoint_routing import plan_calls
+
 DEFAULT_BROWSER_HEADERS = {
     "Accept": "application/json, text/plain, */*",
     "Content-Type": "application/json",
@@ -142,7 +144,7 @@ class APIClient:
                 raise ValueError(f"API 返回错误 code={code} msg={msg}")
 
     # ------------------------------------------------------------------ 分页
-    def iter_paginated(
+    def _iter_paginated_single(
         self,
         endpoint: str,
         params: dict | None,
@@ -155,8 +157,7 @@ class APIClient:
         page_end: int | None = None,
     ) -> Iterable[tuple[int, list, dict, dict]]:
         """
-        分页迭代器：逐页拉取数据并产出 (page_no, records, request_params, raw_response)。
-        page_size=None 时不附带分页参数，仅拉取一次。
+        单一 endpoint 的分页迭代器（不包含 recent/former 路由逻辑）。
         """
         base_params = dict(params or {})
         page = page_start
@@ -183,6 +184,42 @@ class APIClient:
 
             page += 1
 
+    def iter_paginated(
+        self,
+        endpoint: str,
+        params: dict | None,
+        page_size: int | None = 200,
+        page_field: str = "page",
+        size_field: str = "limit",
+        data_path: tuple = ("data",),
+        list_key: str | Sequence[str] | None = None,
+        page_start: int = 1,
+        page_end: int | None = None,
+    ) -> Iterable[tuple[int, list, dict, dict]]:
+        """
+        分页迭代器：逐页拉取数据并产出 (page_no, records, request_params, raw_response)。
+        page_size=None 时不附带分页参数，仅拉取一次。
+        """
+        # recent/former 路由：当 params 带时间范围字段时，按“3个月自然月”边界决定走哪个 endpoint，
+        # 跨越边界则拆分为两段请求并顺序产出，确保调用方使用 page_no 命名文件时不会被覆盖。
+        call_plan = plan_calls(endpoint, params)
+        global_page = 1
+
+        for call in call_plan:
+            for _, records, request_params, payload in self._iter_paginated_single(
+                endpoint=call.endpoint,
+                params=call.params,
+                page_size=page_size,
+                page_field=page_field,
+                size_field=size_field,
+                data_path=data_path,
+                list_key=list_key,
+                page_start=page_start,
+                page_end=page_end,
+            ):
+                yield global_page, records, request_params, payload
+                global_page += 1
+
     def get_paginated(
         self,
         endpoint: str,

etl_billiards/api/endpoint_routing.py

@@ -0,0 +1,166 @@
+# -*- coding: utf-8 -*-
+"""
+“近期记录 / 历史记录(Former)”接口路由规则。
+
+需求：
+- 当请求参数包含可定义时间范围的字段时，根据当前时间（北京时间/上海时区）判断：
+  - 3个月（自然月）之前 -> 使用“历史记录”接口
+  - 3个月以内 -> 使用“近期记录”接口
+  - 若时间范围跨越边界 -> 拆分为两段分别请求并合并（由上层分页迭代器顺序产出）
+"""
+
+from __future__ import annotations
+
+from dataclasses import dataclass
+from datetime import datetime
+from typing import Optional
+
+from dateutil import parser as dtparser
+from dateutil.relativedelta import relativedelta
+from zoneinfo import ZoneInfo
+
+
+ROUTING_TZ = ZoneInfo("Asia/Shanghai")
+RECENT_MONTHS = 3
+
+
+# 按 `fetch-test/recent_vs_former_report.md` 更新（“无”表示没有历史接口；相同 path 表示同一个接口可查历史）
+RECENT_TO_FORMER_OVERRIDES: dict[str, str | None] = {
+    "/AssistantPerformance/GetAbolitionAssistant": None,
+    "/Site/GetSiteTableUseDetails": "/Site/GetSiteTableUseDetails",
+    "/GoodsStockManage/QueryGoodsOutboundReceipt": "/GoodsStockManage/QueryFormerGoodsOutboundReceipt",
+    "/Promotion/GetOfflineCouponConsumePageList": "/Promotion/GetOfflineCouponConsumePageList",
+    "/Order/GetRefundPayLogList": None,
+    # 已知特殊
+    "/Site/GetAllOrderSettleList": "/Site/GetFormerOrderSettleList",
+    "/PayLog/GetPayLogListPage": "/PayLog/GetFormerPayLogListPage",
+}
+
+
+TIME_WINDOW_KEYS: tuple[tuple[str, str], ...] = (
+    ("startTime", "endTime"),
+    ("rangeStartTime", "rangeEndTime"),
+    ("StartPayTime", "EndPayTime"),
+)
+
+
+@dataclass(frozen=True)
+class WindowSpec:
+    start_key: str
+    end_key: str
+    start: datetime
+    end: datetime
+
+
+@dataclass(frozen=True)
+class RoutedCall:
+    endpoint: str
+    params: dict
+
+
+def is_former_endpoint(endpoint: str) -> bool:
+    return "Former" in str(endpoint or "")
+
+
+def _parse_dt(value: object, tz: ZoneInfo) -> datetime | None:
+    if value is None:
+        return None
+    s = str(value).strip()
+    if not s:
+        return None
+    dt = dtparser.parse(s)
+    if dt.tzinfo is None:
+        return dt.replace(tzinfo=tz)
+    return dt.astimezone(tz)
+
+
+def _fmt_dt(dt: datetime, tz: ZoneInfo) -> str:
+    return dt.astimezone(tz).strftime("%Y-%m-%d %H:%M:%S")
+
+
+def extract_window_spec(params: dict | None, tz: ZoneInfo = ROUTING_TZ) -> WindowSpec | None:
+    if not isinstance(params, dict) or not params:
+        return None
+    for start_key, end_key in TIME_WINDOW_KEYS:
+        if start_key in params or end_key in params:
+            start = _parse_dt(params.get(start_key), tz)
+            end = _parse_dt(params.get(end_key), tz)
+            if start and end:
+                return WindowSpec(start_key=start_key, end_key=end_key, start=start, end=end)
+    return None
+
+
+def derive_former_endpoint(recent_endpoint: str) -> str | None:
+    endpoint = str(recent_endpoint or "").strip()
+    if not endpoint:
+        return None
+
+    if endpoint in RECENT_TO_FORMER_OVERRIDES:
+        return RECENT_TO_FORMER_OVERRIDES[endpoint]
+
+    if is_former_endpoint(endpoint):
+        return endpoint
+
+    idx = endpoint.find("Get")
+    if idx == -1:
+        return endpoint
+    return f"{endpoint[:idx]}GetFormer{endpoint[idx + 3:]}"
+
+
+def recent_boundary(now: datetime, months: int = RECENT_MONTHS) -> datetime:
+    """
+    3个月（自然月）边界：取 (now - months) 所在月份的 1 号 00:00:00。
+    """
+    if now.tzinfo is None:
+        raise ValueError("now 必须为时区时间")
+    base = now - relativedelta(months=months)
+    return base.replace(day=1, hour=0, minute=0, second=0, microsecond=0)
+
+
+def plan_calls(
+    endpoint: str,
+    params: dict | None,
+    *,
+    now: datetime | None = None,
+    tz: ZoneInfo = ROUTING_TZ,
+    months: int = RECENT_MONTHS,
+) -> list[RoutedCall]:
+    """
+    根据 endpoint + params 的时间窗口，返回要调用的 endpoint/params 列表（可能拆分为两段）。
+    """
+    base_params = dict(params or {})
+    if not base_params:
+        return [RoutedCall(endpoint=endpoint, params=base_params)]
+
+    # 若调用方显式传了 Former 接口，则不二次路由。
+    if is_former_endpoint(endpoint):
+        return [RoutedCall(endpoint=endpoint, params=base_params)]
+
+    window = extract_window_spec(base_params, tz)
+    if not window:
+        return [RoutedCall(endpoint=endpoint, params=base_params)]
+
+    former_endpoint = derive_former_endpoint(endpoint)
+    if former_endpoint is None or former_endpoint == endpoint:
+        return [RoutedCall(endpoint=endpoint, params=base_params)]
+
+    now_dt = (now or datetime.now(tz)).astimezone(tz)
+    boundary = recent_boundary(now_dt, months=months)
+
+    start, end = window.start, window.end
+    if end <= boundary:
+        return [RoutedCall(endpoint=former_endpoint, params=base_params)]
+    if start >= boundary:
+        return [RoutedCall(endpoint=endpoint, params=base_params)]
+
+    # 跨越边界：拆分两段（老数据 -> former，新数据 -> recent）
+    p1 = dict(base_params)
+    p1[window.start_key] = _fmt_dt(start, tz)
+    p1[window.end_key] = _fmt_dt(boundary, tz)
+
+    p2 = dict(base_params)
+    p2[window.start_key] = _fmt_dt(boundary, tz)
+    p2[window.end_key] = _fmt_dt(end, tz)
+
+    return [RoutedCall(endpoint=former_endpoint, params=p1), RoutedCall(endpoint=endpoint, params=p2)]
+

etl_billiards/api/local_json_client.py

@@ -20,6 +20,10 @@ class LocalJsonClient:
         if not self.base_dir.exists():
             raise FileNotFoundError(f"JSON 目录不存在: {self.base_dir}")
 
+    def get_source_hint(self, endpoint: str) -> str:
+        """Return the JSON file path for this endpoint (for source_file lineage)."""
+        return str(self.base_dir / endpoint_to_filename(endpoint))
+
     def iter_paginated(
         self,
         endpoint: str,

etl_billiards/api/recording_client.py

@@ -7,6 +7,7 @@ from pathlib import Path
 from typing import Any, Iterable, Tuple
 
 from api.client import APIClient
+from api.endpoint_routing import plan_calls
 from utils.json_store import dump_json, endpoint_to_filename
 
 
@@ -33,6 +34,10 @@ class RecordingAPIClient:
         self.last_dump: dict[str, Any] | None = None
 
     # ------------------------------------------------------------------ public API
+    def get_source_hint(self, endpoint: str) -> str:
+        """Return the JSON dump path for this endpoint (for source_file lineage)."""
+        return str(self.output_dir / endpoint_to_filename(endpoint))
+
     def iter_paginated(
         self,
         endpoint: str,
@@ -99,11 +104,18 @@ class RecordingAPIClient:
     ):
         filename = endpoint_to_filename(endpoint)
         path = self.output_dir / filename
+        routing_calls = []
+        try:
+            for call in plan_calls(endpoint, params):
+                routing_calls.append({"endpoint": call.endpoint, "params": call.params})
+        except Exception:
+            routing_calls = []
         payload = {
             "task_code": self.task_code,
             "run_id": self.run_id,
             "endpoint": endpoint,
             "params": params or {},
+            "endpoint_routing": {"calls": routing_calls} if routing_calls else None,
             "page_size": page_size,
             "pages": pages,
             "total_records": total_records,

etl_billiards/cli/main.py

@@ -40,6 +40,23 @@ def parse_args():
     parser.add_argument("--api-timeout", type=int, help="API超时(秒)")
     parser.add_argument("--api-page-size", type=int, help="分页大小")
     parser.add_argument("--api-retry-max", type=int, help="API重试最大次数")
+
+    # 回溯/手动窗口
+    parser.add_argument(
+        "--window-start",
+        dest="window_start",
+        help="固定时间窗口开始（优先级高于游标，例如：2025-07-01 00:00:00）",
+    )
+    parser.add_argument(
+        "--window-end",
+        dest="window_end",
+        help="固定时间窗口结束（优先级高于游标，推荐用月末+1，例如：2025-08-01 00:00:00）",
+    )
+    parser.add_argument(
+        "--force-window-override",
+        action="store_true",
+        help="强制使用 window_start/window_end，不走 MAX(fetched_at) 兜底",
+    )
     
     # 目录参数
     parser.add_argument("--export-root", help="导出根目录")
@@ -108,6 +125,16 @@ def build_cli_overrides(args) -> dict:
     if args.write_pretty_json:
         overrides.setdefault("io", {})["write_pretty_json"] = True
 
+    # 回溯/手动窗口
+    if args.window_start or args.window_end:
+        overrides.setdefault("run", {}).setdefault("window_override", {})
+        if args.window_start:
+            overrides["run"]["window_override"]["start"] = args.window_start
+        if args.window_end:
+            overrides["run"]["window_override"]["end"] = args.window_end
+    if args.force_window_override:
+        overrides.setdefault("run", {})["force_window_override"] = True
+
     # 运行窗口
     if args.idle_start:
         overrides.setdefault("run", {}).setdefault("idle_window", {})["start"] = args.idle_start

etl_billiards/config/env_parser.py

@@ -40,6 +40,8 @@ ENV_MAP = {
     "IDLE_WINDOW_END": ("run.idle_window.end",),
     "ALLOW_EMPTY_RESULT_ADVANCE": ("run.allow_empty_result_advance",),
     "ALLOW_EMPTY_ADVANCE": ("run.allow_empty_result_advance",),
+    "WINDOW_START": ("run.window_override.start",),
+    "WINDOW_END": ("run.window_override.end",),
     "PIPELINE_FLOW": ("pipeline.flow",),
     "JSON_FETCH_ROOT": ("pipeline.fetch_root",),
     "JSON_SOURCE_DIR": ("pipeline.ingest_source_dir",),

etl_billiards/database/operations.py

@@ -25,46 +25,54 @@ class DatabaseOperations:
         
         use_returning = "RETURNING" in sql.upper()
         
-        with self.conn.cursor() as c:
-            if not use_returning:
+        # 不带 RETURNING：直接批量执行即可
+        if not use_returning:
+            with self.conn.cursor() as c:
                 psycopg2.extras.execute_batch(c, sql, rows, page_size=page_size)
-                return (0, 0)
-            
-            # 尝试向量化执行
+            return (0, 0)
+
+        # 尝试向量化执行（execute_values + fetch returning）
+        vectorized_failed = False
+        m = re.search(r"VALUES\s*\((.*?)\)", sql, flags=re.IGNORECASE | re.DOTALL)
+        if m:
+            tpl = "(" + m.group(1) + ")"
+            base_sql = sql[:m.start()] + "VALUES %s" + sql[m.end():]
             try:
-                m = re.search(r"VALUES\s*\((.*?)\)", sql, flags=re.IGNORECASE | re.DOTALL)
-                if m:
-                    tpl = "(" + m.group(1) + ")"
-                    base_sql = sql[:m.start()] + "VALUES %s" + sql[m.end():]
-                    
+                with self.conn.cursor() as c:
                     ret = psycopg2.extras.execute_values(
                         c, base_sql, rows, template=tpl, page_size=page_size, fetch=True
                     )
-                    
-                    if not ret:
-                        return (0, 0)
-                    
-                    inserted = sum(1 for rec in ret if self._is_inserted(rec))
-                    return (inserted, len(ret) - inserted)
+                if not ret:
+                    return (0, 0)
+                inserted = sum(1 for rec in ret if self._is_inserted(rec))
+                return (inserted, len(ret) - inserted)
+            except Exception:
+                # 向量化失败后，事务通常处于 aborted 状态，需要先 rollback 才能继续执行。
+                vectorized_failed = True
+
+        if vectorized_failed:
+            try:
+                self.conn.rollback()
             except Exception:
                 pass
-            
-            # 回退：逐行执行
-            inserted = 0
-            updated = 0
+
+        # 回退：逐行执行
+        inserted = 0
+        updated = 0
+        with self.conn.cursor() as c:
             for r in rows:
                 c.execute(sql, r)
                 try:
                     rec = c.fetchone()
                 except Exception:
                     rec = None
-                
+
                 if self._is_inserted(rec):
                     inserted += 1
                 else:
                     updated += 1
-            
-            return (inserted, updated)
+
+        return (inserted, updated)
     
     @staticmethod
     def _is_inserted(rec) -> bool:

etl_billiards/database/schema_dws.sql

@@ -0,0 +1,50 @@
+-- DWS schema for aggregated / serving tables.
+CREATE SCHEMA IF NOT EXISTS billiards_dws;
+
+CREATE TABLE IF NOT EXISTS billiards_dws.dws_order_summary (
+    site_id BIGINT NOT NULL,
+    order_settle_id BIGINT NOT NULL,
+    order_trade_no TEXT,
+    order_date DATE,
+    tenant_id BIGINT,
+    member_id BIGINT,
+    member_flag BOOLEAN,
+    recharge_order_flag BOOLEAN,
+    item_count INT,
+    total_item_quantity NUMERIC,
+    table_fee_amount NUMERIC,
+    assistant_service_amount NUMERIC,
+    goods_amount NUMERIC,
+    group_amount NUMERIC,
+    total_coupon_deduction NUMERIC,
+    member_discount_amount NUMERIC,
+    manual_discount_amount NUMERIC,
+    order_original_amount NUMERIC,
+    order_final_amount NUMERIC,
+    stored_card_deduct NUMERIC,
+    external_paid_amount NUMERIC,
+    total_paid_amount NUMERIC,
+    book_table_flow NUMERIC,
+    book_assistant_flow NUMERIC,
+    book_goods_flow NUMERIC,
+    book_group_flow NUMERIC,
+    book_order_flow NUMERIC,
+    order_effective_consume_cash NUMERIC,
+    order_effective_recharge_cash NUMERIC,
+    order_effective_flow NUMERIC,
+    refund_amount NUMERIC,
+    net_income NUMERIC,
+    created_at TIMESTAMPTZ DEFAULT now(),
+    updated_at TIMESTAMPTZ DEFAULT now(),
+    PRIMARY KEY (site_id, order_settle_id)
+);
+
+CREATE INDEX IF NOT EXISTS idx_dws_order_summary_order_date
+    ON billiards_dws.dws_order_summary (order_date);
+
+CREATE INDEX IF NOT EXISTS idx_dws_order_summary_tenant_date
+    ON billiards_dws.dws_order_summary (tenant_id, order_date);
+
+CREATE INDEX IF NOT EXISTS idx_dws_order_summary_member_date
+    ON billiards_dws.dws_order_summary (member_id, order_date);
+

etl_billiards/database/seed_ods_tasks.sql

@@ -8,28 +8,30 @@ WITH target_store AS (
 ),
 task_codes AS (
     SELECT unnest(ARRAY[
-        'assistant_accounts_masterS',
-        'assistant_service_records',
-        'assistant_cancellation_records',
-        'goods_stock_movements',
-        'ODS_INVENTORY_STOCK',
-        'ODS_PACKAGE',
-        'ODS_GROUP_BUY_REDEMPTION',
-        'ODS_MEMBER',
-        'ODS_MEMBER_BALANCE',
-        'member_stored_value_cards',
+        -- Must match tasks/ods_tasks.py (ENABLED_ODS_CODES)
+        'ODS_ASSISTANT_ACCOUNT',
+        'ODS_ASSISTANT_LEDGER',
+        'ODS_ASSISTANT_ABOLISH',
+        'ODS_SETTLEMENT_RECORDS',
+        'ODS_TABLE_USE',
         'ODS_PAYMENT',
         'ODS_REFUND',
-        'platform_coupon_redemption_records',
-        'recharge_settlements',
+        'ODS_PLATFORM_COUPON',
+        'ODS_MEMBER',
+        'ODS_MEMBER_CARD',
+        'ODS_MEMBER_BALANCE',
+        'ODS_RECHARGE_SETTLE',
+        'ODS_GROUP_PACKAGE',
+        'ODS_GROUP_BUY_REDEMPTION',
+        'ODS_INVENTORY_STOCK',
+        'ODS_INVENTORY_CHANGE',
         'ODS_TABLES',
         'ODS_GOODS_CATEGORY',
         'ODS_STORE_GOODS',
-        'table_fee_discount_records',
+        'ODS_STORE_GOODS_SALES',
+        'ODS_TABLE_FEE_DISCOUNT',
         'ODS_TENANT_GOODS',
-        'ODS_SETTLEMENT_TICKET',
-        'settlement_records',
-        'INIT_ODS_SCHEMA'
+        'ODS_SETTLEMENT_TICKET'
     ]) AS task_code
 )
 INSERT INTO etl_admin.etl_task (task_code, store_id, enabled)

etl_billiards/database/seed_scheduler_tasks.sql

@@ -0,0 +1,49 @@
+-- Seed scheduler-compatible tasks into etl_admin.etl_task.
+--
+-- Notes:
+-- - These task_code values must match orchestration/task_registry.py.
+-- - ODS_* tasks are intentionally excluded here because they don't follow the
+--   BaseTask(cursor_data) scheduler interface in this repo version.
+--
+-- Usage (example):
+--   psql "%PG_DSN%" -f etl_billiards/database/seed_scheduler_tasks.sql
+--
+WITH target_store AS (
+    SELECT 2790685415443269::bigint AS store_id  -- TODO: replace with your store_id
+),
+task_codes AS (
+    SELECT unnest(ARRAY[
+        'ASSISTANT_ABOLISH',
+        'ASSISTANTS',
+        'COUPON_USAGE',
+        'CHECK_CUTOFF',
+        'DWD_LOAD_FROM_ODS',
+        'DWD_QUALITY_CHECK',
+        'INIT_DWD_SCHEMA',
+        'INIT_DWS_SCHEMA',
+        'INIT_ODS_SCHEMA',
+        'INVENTORY_CHANGE',
+        'LEDGER',
+        'MANUAL_INGEST',
+        'MEMBERS',
+        'MEMBERS_DWD',
+        'ODS_JSON_ARCHIVE',
+        'ORDERS',
+        'PACKAGES_DEF',
+        'PAYMENTS',
+        'PAYMENTS_DWD',
+        'PRODUCTS',
+        'REFUNDS',
+        'TABLE_DISCOUNT',
+        'TABLES',
+        'TICKET_DWD',
+        'TOPUPS',
+        'DWS_BUILD_ORDER_SUMMARY'
+    ]) AS task_code
+)
+INSERT INTO etl_admin.etl_task (task_code, store_id, enabled)
+SELECT t.task_code, s.store_id, TRUE
+FROM task_codes t CROSS JOIN target_store s
+ON CONFLICT (task_code, store_id) DO UPDATE
+SET enabled = EXCLUDED.enabled,
+    updated_at = now();

etl_billiards/docs/table_2025-12-19/_generate_assistant_tables.py

@@ -0,0 +1,585 @@
+# -*- coding: utf-8 -*-
+"""生成 2025年10-12月 助教排行榜 + 助教详情表（CSV + MD）。
+
+输出目录：etl_billiards/docs/table_2025-12-19
+
+注意：客户流水/充值归因涉及“多助教/多订单命中”时按全额复制计入，会导致助教汇总>门店汇总，表格说明会写明。
+"""
+
+from __future__ import annotations
+
+import csv
+import re
+from dataclasses import dataclass
+from decimal import Decimal
+from pathlib import Path
+from statistics import median
+from typing import Any
+
+import psycopg2
+import psycopg2.extras
+
+
+SITE_ID = 2790685415443269
+TZ = "Asia/Shanghai"
+
+WIN_OCT = ("2025-10-01 00:00:00+08", "2025-11-01 00:00:00+08")
+WIN_NOV = ("2025-11-01 00:00:00+08", "2025-12-01 00:00:00+08")
+WIN_DEC = ("2025-12-01 00:00:00+08", "2026-01-01 00:00:00+08")
+WIN_ALL = (WIN_OCT[0], WIN_DEC[1])
+
+MONTHS = [
+    ("2025-10", "10月", WIN_OCT),
+    ("2025-11", "11月", WIN_NOV),
+    ("2025-12", "12月", WIN_DEC),
+]
+
+REPO_ROOT = Path(__file__).resolve().parents[3]
+ENV_PATH = REPO_ROOT / "etl_billiards" / ".env"
+OUT_DIR = Path(__file__).resolve().parent
+
+
+@dataclass(frozen=True)
+class SqlBlock:
+    title: str
+    sql: str
+
+
+def read_pg_dsn() -> str:
+    text = ENV_PATH.read_text(encoding="utf-8")
+    m = re.search(r"^PG_DSN=(.+)$", text, re.M)
+    if not m:
+        raise RuntimeError(f"未在 {ENV_PATH} 中找到 PG_DSN")
+    return m.group(1).strip()
+
+
+def conn():
+    return psycopg2.connect(read_pg_dsn(), connect_timeout=10)
+
+
+def sanitize_filename(name: str) -> str:
+    name = name.strip()
+    name = re.sub(r"[<>:\"/\\|?*]+", "_", name)
+    name = re.sub(r"\s+", " ", name)
+    return name
+
+
+def d(v: Any) -> Decimal:
+    if v is None:
+        return Decimal("0")
+    if isinstance(v, Decimal):
+        return v
+    return Decimal(str(v))
+
+
+def fmt_money(v: Any) -> str:
+    return f"{d(v):.2f}"
+
+
+def fmt_hours(v: Any, digits: int = 2) -> str:
+    q = Decimal("1").scaleb(-digits)
+    return f"{d(v).quantize(q):f}h"
+
+
+def write_csv(path: Path, title: str, description: str, header_rows: list[list[str]], rows: list[list[Any]]) -> None:
+    path.parent.mkdir(parents=True, exist_ok=True)
+    with path.open("w", newline="", encoding="utf-8") as f:
+        w = csv.writer(f)
+        w.writerow([title])
+        w.writerow([description])
+        w.writerow([])
+        for hr in header_rows:
+            w.writerow(hr)
+        for r in rows:
+            w.writerow(["" if v is None else v for v in r])
+
+
+def write_csv_sections(path: Path, title: str, description: str, section_rows: list[list[Any]]) -> None:
+    path.parent.mkdir(parents=True, exist_ok=True)
+    with path.open("w", newline="", encoding="utf-8") as f:
+        w = csv.writer(f)
+        w.writerow([title])
+        w.writerow([description])
+        w.writerow([])
+        for r in section_rows:
+            w.writerow(["" if v is None else v for v in r])
+
+
+def write_md(path: Path, title: str, thinking: str, description: str, sql_blocks: list[SqlBlock]) -> None:
+    parts: list[str] = []
+    parts.append(f"# {title}\n")
+    parts.append("## 思考过程\n")
+    parts.append(thinking.strip() + "\n")
+    parts.append("\n## 查询说明\n")
+    parts.append(description.strip() + "\n")
+    parts.append("\n## SQL\n")
+    for b in sql_blocks:
+        parts.append(f"\n### {b.title}\n")
+        parts.append("```sql\n")
+        parts.append(b.sql.strip() + "\n")
+        parts.append("```\n")
+    path.parent.mkdir(parents=True, exist_ok=True)
+    path.write_text("".join(parts), encoding="utf-8")
+
+
+def fetch_all(cur, sql: str, params: dict[str, Any]) -> list[dict[str, Any]]:
+    cur.execute(sql, params)
+    return list(cur.fetchall())
+
+
+def month_case(ts_expr: str) -> str:
+    parts = []
+    for month_key, _, (ws, we) in MONTHS:
+        parts.append(
+            f"when {ts_expr} >= '{ws}'::timestamptz and {ts_expr} < '{we}'::timestamptz then '{month_key}'"
+        )
+    return "case " + " ".join(parts) + " else null end"
+
+
+def sql_order_base(window_start: str, window_end: str) -> str:
+    return f"""
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '{window_start}'::timestamptz
+    and tfl.start_use_time <  '{window_end}'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+"""
+
+
+def dense_rank_desc(values: dict[str, Decimal]) -> dict[str, int]:
+    uniq = sorted({v for v in values.values() if v > 0}, reverse=True)
+    rank_map = {v: i + 1 for i, v in enumerate(uniq)}
+    return {k: rank_map.get(v, 0) for k, v in values.items()}
+
+
+def calc_diff(all_values: dict[str, Decimal], current: Decimal) -> tuple[Decimal, Decimal]:
+    xs = [v for v in all_values.values() if v > 0]
+    if not xs or current <= 0:
+        return Decimal("0"), Decimal("0")
+    avg = sum(xs) / Decimal(len(xs))
+    med = Decimal(str(median([float(v) for v in xs])))
+    return current - avg, current - med
+
+
+def main() -> None:
+    OUT_DIR.mkdir(parents=True, exist_ok=True)
+
+    with conn() as c, c.cursor(cursor_factory=psycopg2.extras.RealDictCursor) as cur:
+        assistants_rows = fetch_all(
+            cur,
+            """
+select distinct nickname as assistant
+from billiards_dwd.dwd_assistant_service_log
+where site_id=%(site_id)s and coalesce(is_delete,0)=0
+  and start_use_time >= %(window_start)s::timestamptz
+  and start_use_time <  %(window_end)s::timestamptz
+order by assistant;
+""",
+            {"site_id": SITE_ID, "window_start": WIN_ALL[0], "window_end": WIN_ALL[1]},
+        )
+        assistants = [r["assistant"] for r in assistants_rows if r.get("assistant")]
+
+        # 助教-客户-月份：服务时长
+        sql_svc = f"""
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    {month_case('asl.start_use_time')} as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+"""
+        svc_rows = fetch_all(cur, sql_svc, {"site_id": SITE_ID, "window_start": WIN_ALL[0], "window_end": WIN_ALL[1]})
+
+        # 助教-客户-月份：客户流水
+        sql_rev = sql_order_base(WIN_ALL[0], WIN_ALL[1]) + f"""
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    {month_case('o.order_start_time')} as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+"""
+        rev_rows = fetch_all(cur, sql_rev, {"site_id": SITE_ID, "window_start": WIN_ALL[0], "window_end": WIN_ALL[1]})
+
+        # 助教-客户-月份：充值归因
+        sql_rech = f"""
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    {month_case('m.pay_time')} as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+"""
+        rech_rows = fetch_all(cur, sql_rech, {"site_id": SITE_ID, "window_start": WIN_ALL[0], "window_end": WIN_ALL[1]})
+
+        # 汇总：月度助教指标
+        svc_map = {mk: {a: {"base": Decimal('0'), "extra": Decimal('0')} for a in assistants} for mk,_,_ in MONTHS}
+        for r in svc_rows:
+            mk = r["month_key"]; a = r["assistant"]
+            if mk in svc_map and a in svc_map[mk]:
+                svc_map[mk][a]["base"] += d(r["base_hours"])
+                svc_map[mk][a]["extra"] += d(r["extra_hours"])
+
+        revenue_map = {mk: {a: Decimal('0') for a in assistants} for mk,_,_ in MONTHS}
+        for r in rev_rows:
+            mk = r["month_key"]; a = r["assistant"]
+            if mk in revenue_map and a in revenue_map[mk]:
+                revenue_map[mk][a] += d(r["revenue_amount"])
+
+        recharge_map = {mk: {a: Decimal('0') for a in assistants} for mk,_,_ in MONTHS}
+        for r in rech_rows:
+            mk = r["month_key"]; a = r["assistant"]
+            if mk in recharge_map and a in recharge_map[mk]:
+                recharge_map[mk][a] += d(r["recharge_amount"])
+
+        # ====== 输出4张排行榜 ======
+        def write_rank(file_stem: str, title: str, desc: str, rows: list[list[Any]]):
+            write_csv(OUT_DIR / f"{file_stem}.csv", title, desc, [["月份", "排名", "助教昵称", "指标"]], rows)
+            write_md(OUT_DIR / f"{file_stem}.md", title, "按月聚合并做dense_rank排名。", desc, [])
+
+        rows = []
+        for mk,_,_ in MONTHS:
+            values = {a: svc_map[mk][a]["base"] for a in assistants}
+            ranks = dense_rank_desc(values)
+            for a in sorted(assistants, key=lambda x: (ranks[x] if ranks[x] else 999999, x)):
+                v = values[a]
+                if v > 0:
+                    rows.append([mk, ranks[a], a, fmt_hours(v, 2)])
+        write_rank(
+            "助教_基础课时长排行_2025年10-12月",
+            "2025年10-12月 助教基础课时长排行榜",
+            "口径：order_assistant_type=1，时长=income_seconds/3600（小时），按月排名。",
+            rows,
+        )
+
+        rows = []
+        for mk,_,_ in MONTHS:
+            values = {a: svc_map[mk][a]["extra"] for a in assistants}
+            ranks = dense_rank_desc(values)
+            for a in sorted(assistants, key=lambda x: (ranks[x] if ranks[x] else 999999, x)):
+                v = values[a]
+                if v > 0:
+                    rows.append([mk, ranks[a], a, fmt_hours(v, 2)])
+        write_rank(
+            "助教_附加课时长排行_2025年10-12月",
+            "2025年10-12月 助教附加课(超休)时长排行榜",
+            "口径：order_assistant_type=2，超休时长=income_seconds/3600（小时），按月排名。",
+            rows,
+        )
+
+        rows = []
+        for mk,_,_ in MONTHS:
+            values = revenue_map[mk]
+            ranks = dense_rank_desc(values)
+            for a in sorted(assistants, key=lambda x: (ranks[x] if ranks[x] else 999999, x)):
+                v = values[a]
+                if v > 0:
+                    rows.append([mk, ranks[a], a, fmt_money(v)])
+        write_rank(
+            "助教_客户流水排行_2025年10-12月",
+            "2025年10-12月 助教客户流水排行榜（全额复制口径）",
+            "口径：客户流水=台费+助教+商品应付金额按订单归集后，全额计入订单内每位助教；多助教会导致汇总>门店总额。",
+            rows,
+        )
+
+        rows = []
+        for mk,_,_ in MONTHS:
+            values = recharge_map[mk]
+            ranks = dense_rank_desc(values)
+            for a in sorted(assistants, key=lambda x: (ranks[x] if ranks[x] else 999999, x)):
+                v = values[a]
+                if v > 0:
+                    rows.append([mk, ranks[a], a, fmt_money(v)])
+        write_rank(
+            "助教_客户充值归因排行_2025年10-12月",
+            "2025年10-12月 助教客户充值归因排行榜（全额复制口径）",
+            "口径：充值支付(dwd_payment.relate_type=5)在消费窗口±30分钟内命中且订单有助教，则全额计入助教；多助教/多订单命中会重复计入。",
+            rows,
+        )
+
+        # ====== 输出助教详情（每人一份） ======
+        # 会员昵称
+        cur.execute("select member_id, nickname from billiards_dwd.dim_member where scd2_is_current=1")
+        member_name = {r["member_id"]: (r.get("nickname") or "") for r in cur.fetchall()}
+
+        # 索引：assistant->member->month
+        svc_idx = {a: {} for a in assistants}
+        for r in svc_rows:
+            a = r["assistant"]; mid = int(r["member_id"]); mk = r["month_key"]
+            svc_idx.setdefault(a, {}).setdefault(mid, {})[mk] = {"base": d(r["base_hours"]), "extra": d(r["extra_hours"])}
+
+        rev_idx = {a: {} for a in assistants}
+        for r in rev_rows:
+            a = r["assistant"]; mid = int(r["member_id"]); mk = r["month_key"]
+            rev_idx.setdefault(a, {}).setdefault(mid, {})[mk] = d(r["revenue_amount"])
+
+        rech_idx = {a: {} for a in assistants}
+        for r in rech_rows:
+            a = r["assistant"]; mid = int(r["member_id"]); mk = r["month_key"]
+            rech_idx.setdefault(a, {}).setdefault(mid, {})[mk] = d(r["recharge_amount"])
+
+        for a in assistants:
+            safe = sanitize_filename(a)
+            csv_path = OUT_DIR / f"助教详情_{safe}.csv"
+            md_path = OUT_DIR / f"助教详情_{safe}.md"
+
+            # 评价（简短）
+            base_total = sum((svc_map[mk][a]["base"] for mk,_,_ in MONTHS), Decimal('0'))
+            extra_total = sum((svc_map[mk][a]["extra"] for mk,_,_ in MONTHS), Decimal('0'))
+            rev_total = sum((revenue_map[mk][a] for mk,_,_ in MONTHS), Decimal('0'))
+            rech_total = sum((recharge_map[mk][a] for mk,_,_ in MONTHS), Decimal('0'))
+
+            # 头部客户 Top100（按12月消费业绩）
+            members = set(rev_idx.get(a, {}).keys()) | set(svc_idx.get(a, {}).keys()) | set(rech_idx.get(a, {}).keys())
+            def rev_dec(mid: int) -> Decimal:
+                return rev_idx.get(a, {}).get(mid, {}).get('2025-12', Decimal('0'))
+            top_members = sorted(members, key=lambda mid: rev_dec(mid), reverse=True)[:100]
+
+            top3 = '、'.join([(member_name.get(mid) or str(mid)) for mid in top_members[:3]])
+            assistant_review = (
+                f"评价：基础{fmt_hours(base_total,1)}，附加{fmt_hours(extra_total,1)}；"
+                f"客户流水¥{rev_total:.2f}，充值归因¥{rech_total:.2f}；"
+                f"头部客户(12月)Top3：{top3 or '无'}。"
+            )
+
+            # Part1-4
+            part1=[]; part2=[]; part3=[]; part4=[]
+            for mk, mcn, _ in MONTHS:
+                base_v = svc_map[mk][a]["base"]
+                extra_v = svc_map[mk][a]["extra"]
+                rev_v = revenue_map[mk][a]
+                rech_v = recharge_map[mk][a]
+
+                base_all = {x: svc_map[mk][x]["base"] for x in assistants}
+                extra_all = {x: svc_map[mk][x]["extra"] for x in assistants}
+                rev_all = {x: revenue_map[mk][x] for x in assistants}
+                rech_all = {x: recharge_map[mk][x] for x in assistants}
+
+                base_rank = dense_rank_desc(base_all).get(a, 0)
+                extra_rank = dense_rank_desc(extra_all).get(a, 0)
+                rev_rank = dense_rank_desc(rev_all).get(a, 0)
+                rech_rank = dense_rank_desc(rech_all).get(a, 0)
+
+                base_da, base_dm = calc_diff(base_all, base_v)
+                extra_da, extra_dm = calc_diff(extra_all, extra_v)
+                rev_da, rev_dm = calc_diff(rev_all, rev_v)
+                rech_da, rech_dm = calc_diff(rech_all, rech_v)
+
+                part1.append([mcn, fmt_hours(base_v,2), base_rank or "", fmt_hours(base_da,2), fmt_hours(base_dm,2)])
+                part2.append([mcn, fmt_hours(extra_v,2), extra_rank or "", fmt_hours(extra_da,2), fmt_hours(extra_dm,2)])
+                part3.append([mcn, fmt_money(rev_v), rev_rank or "", fmt_money(rev_da), fmt_money(rev_dm)])
+                part4.append([mcn, fmt_money(rech_v), rech_rank or "", fmt_money(rech_da), fmt_money(rech_dm)])
+
+            # Part5 rows
+            part5=[]
+            for i, mid in enumerate(top_members, start=1):
+                def h_pair(month_key: str) -> str:
+                    v = svc_idx.get(a, {}).get(mid, {}).get(month_key, {})
+                    return f"{fmt_hours(v.get('base',Decimal('0')),1)} / {fmt_hours(v.get('extra',Decimal('0')),1)}"
+                def rev_m(month_key: str) -> Decimal:
+                    return rev_idx.get(a, {}).get(mid, {}).get(month_key, Decimal('0'))
+                def rech_m(month_key: str) -> Decimal:
+                    return rech_idx.get(a, {}).get(mid, {}).get(month_key, Decimal('0'))
+                name = member_name.get(mid) or str(mid)
+                part5.append([
+                    i,
+                    name,
+                    h_pair('2025-12'), fmt_money(rev_m('2025-12')), fmt_money(rech_m('2025-12')),
+                    h_pair('2025-11'), fmt_money(rev_m('2025-11')), fmt_money(rech_m('2025-11')),
+                    h_pair('2025-10'), fmt_money(rev_m('2025-10')), fmt_money(rech_m('2025-10')),
+                ])
+
+            title = f"助教详情：{a}（2025年10-12月）"
+            desc = (
+                "本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。"
+                "均值/中位数差值对比集合为当月该指标>0的助教。"
+                "充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。"
+            )
+
+            rows=[]
+            rows += [["一、基础课业绩"], ["说明：" + assistant_review], []]
+            rows += [["月份", "基础课业绩", "基础课业绩", "基础课业绩", "基础课业绩"], ["月份", "小时数", "排名", "平均值差值小时数", "中位数值差值小时数"]]
+            rows += part1
+            rows += [[], ["二、附加课业绩"], ["说明：附加课=order_assistant_type=2。"], []]
+            rows += [["月份", "附加课业绩", "附加课业绩", "附加课业绩", "附加课业绩"], ["月份", "小时数", "排名", "平均值差值小时数", "中位数值差值小时数"]]
+            rows += part2
+            rows += [[], ["三、客户消费业绩"], ["说明：订单台费+助教+商品应付金额全额计入订单内助教。"], []]
+            rows += [["月份", "客户消费业绩", "客户消费业绩", "客户消费业绩", "客户消费业绩"], ["月份", "合计元", "排名", "平均值差值元", "中位数值差值元"]]
+            rows += part3
+            rows += [[], ["四、客户充值业绩"], ["说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。"], []]
+            rows += [["月份", "客户充值业绩", "客户充值业绩", "客户充值业绩", "客户充值业绩"], ["月份", "合计元", "排名", "平均值差值元", "中位数值差值元"]]
+            rows += part4
+            rows += [[], ["五、头部客户（按12月消费业绩排序，Top100）"], ["说明：基础/附加课时=基础h/附加h。"], []]
+            rows += [["排名", "客户名称", "12月", "12月", "12月", "11月", "11月", "11月", "10月", "10月", "10月"],
+                     ["排名", "客户名称", "基础/附加课时", "消费业绩(元)", "客户充值(元)", "基础/附加课时", "消费业绩(元)", "客户充值(元)", "基础/附加课时", "消费业绩(元)", "客户充值(元)"]]
+            rows += part5
+
+            write_csv_sections(csv_path, title, desc, rows)
+            write_md(
+                md_path,
+                title,
+                "按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。",
+                desc + "\n" + assistant_review,
+                [
+                    SqlBlock("服务时长（助教-客户-月份）", sql_svc),
+                    SqlBlock("客户流水（助教-客户-月份）", sql_rev),
+                    SqlBlock("充值归因（助教-客户-月份）", sql_rech),
+                ],
+            )
+
+    print(f"完成：{OUT_DIR}")
+
+
+if __name__ == "__main__":
+    main()

etl_billiards/docs/table_2025-12-19/助教_基础课时长排行_2025年10-12月.csv

@@ -0,0 +1,73 @@
+2025年10-12月 助教基础课时长排行榜
+口径：order_assistant_type=1，时长=income_seconds/3600（小时），按月排名。
+
+月份,排名,助教昵称,指标
+2025-10,1,佳怡,139.55h
+2025-10,2,璇子,120.20h
+2025-10,3,婉婉,90.68h
+2025-10,4,七七,64.70h
+2025-10,5,小柔,63.87h
+2025-10,6,球球,57.45h
+2025-10,7,小敌,55.82h
+2025-10,8,涛涛,50.13h
+2025-10,9,周周,41.33h
+2025-10,10,素素,40.45h
+2025-10,11,乔西,32.60h
+2025-10,12,苏苏,25.15h
+2025-10,13,奈千,23.62h
+2025-10,14,年糕,21.23h
+2025-10,15,欣怡,19.50h
+2025-10,16,饭团,16.00h
+2025-10,17,Amy,11.97h
+2025-10,18,姜姜,6.60h
+2025-10,19,希希,5.02h
+2025-10,20,悦悦,2.30h
+2025-11,1,佳怡,176.25h
+2025-11,2,璇子,147.92h
+2025-11,3,小燕,109.28h
+2025-11,4,Amy,93.53h
+2025-11,5,七七,91.90h
+2025-11,6,小柔,88.65h
+2025-11,7,涛涛,74.40h
+2025-11,8,阿清,73.48h
+2025-11,9,小敌,72.90h
+2025-11,10,周周,71.27h
+2025-11,11,球球,66.50h
+2025-11,12,婉婉,46.03h
+2025-11,13,小侯,42.58h
+2025-11,14,千千,38.88h
+2025-11,15,年糕,35.80h
+2025-11,16,柚子,35.40h
+2025-11,17,素素,35.03h
+2025-11,18,瑶瑶,34.25h
+2025-11,19,奈千,32.83h
+2025-11,20,乔西,30.57h
+2025-11,21,泡芙,21.38h
+2025-11,22,梦梦,19.60h
+2025-11,23,苏苏,13.52h
+2025-11,24,欣怡,10.33h
+2025-11,25,QQ,5.17h
+2025-11,26,西子,1.82h
+2025-11,27,希希,1.58h
+2025-12,1,小燕,159.02h
+2025-12,2,佳怡,109.40h
+2025-12,3,璇子,90.75h
+2025-12,4,七七,77.72h
+2025-12,5,阿清,66.45h
+2025-12,6,周周,60.02h
+2025-12,7,小柔,54.93h
+2025-12,8,小侯,49.57h
+2025-12,9,球球,48.58h
+2025-12,10,涛涛,44.08h
+2025-12,11,苏苏,43.90h
+2025-12,12,千千,38.28h
+2025-12,13,乔西,25.82h
+2025-12,14,年糕,25.62h
+2025-12,15,瑶瑶,19.48h
+2025-12,16,Amy,18.08h
+2025-12,17,婉婉,17.83h
+2025-12,18,梦梦,16.08h
+2025-12,19,素素,9.98h
+2025-12,20,小敌,6.40h
+2025-12,21,奈千,2.58h
+2025-12,22,QQ,1.22h

etl_billiards/docs/table_2025-12-19/助教_基础课时长排行_2025年10-12月.md

@@ -0,0 +1,31 @@
+# 2025年10-12月 助教基础课时长排行榜
+## 思考过程
+按月汇总助教基础课时长，并用 dense_rank 做排名。
+
+## 查询说明
+口径：order_assistant_type=1；时长=income_seconds/3600（小时）。
+
+## SQL
+
+### 基础课时长（助教+月份汇总）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0)=0
+    and asl.order_assistant_type=1
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+)
+select
+  assistant,
+  month_key,
+  sum(income_seconds)/3600.0 as hours
+from raw
+where month_key is not null
+group by assistant, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教_客户充值归因排行_2025年10-12月.csv

@@ -0,0 +1,35 @@
+2025年10-12月 助教客户充值归因排行榜（全额复制口径）
+口径：充值支付(dwd_payment.relate_type=5)在消费窗口±30分钟内命中且订单有助教，则全额计入助教；多助教/多订单命中会重复计入。
+
+月份,排名,助教昵称,指标
+2025-10,1,璇子,34700.00
+2025-10,2,小柔,31700.00
+2025-10,3,佳怡,27000.00
+2025-10,4,婉婉,24000.00
+2025-10,5,小敌,21000.00
+2025-10,5,涛涛,21000.00
+2025-10,6,奈千,18000.00
+2025-10,7,乔西,17000.00
+2025-10,8,球球,15000.00
+2025-10,9,周周,11000.00
+2025-10,10,年糕,9000.00
+2025-10,11,七七,6000.00
+2025-10,11,素素,6000.00
+2025-10,11,苏苏,6000.00
+2025-10,12,姜姜,4000.00
+2025-10,13,Amy,3000.00
+2025-10,13,悦悦,3000.00
+2025-10,13,欣怡,3000.00
+2025-11,1,佳怡,20000.00
+2025-11,2,小柔,11000.00
+2025-11,3,璇子,10000.00
+2025-11,4,Amy,9000.00
+2025-11,4,周周,9000.00
+2025-11,4,婉婉,9000.00
+2025-11,4,球球,9000.00
+2025-11,5,小敌,8000.00
+2025-11,6,涛涛,5000.00
+2025-11,7,欣怡,4000.00
+2025-11,8,乔西,3000.00
+2025-11,8,柚子,3000.00
+2025-11,9,素素,1000.00

etl_billiards/docs/table_2025-12-19/助教_客户充值归因排行_2025年10-12月.md

@@ -0,0 +1,85 @@
+# 2025年10-12月 助教客户充值归因排行榜（全额复制口径）
+## 思考过程
+按“消费窗口±30分钟”把充值支付命中到订单，再全额计入订单内助教，并按月排名。
+
+## 查询说明
+注意：多助教/多订单命中按全额复制，充值会重复计入，故助教汇总可能大于门店总额。
+
+## SQL
+
+### 充值归因（助教+月份汇总，全额复制）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select rp.pay_time, ow.order_settle_id, rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select assistant, month_key, sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教_客户流水排行_2025年10-12月.csv

@@ -0,0 +1,51 @@
+2025年10-12月 助教客户流水排行榜（全额复制口径）
+口径：客户流水=台费+助教+商品应付金额按订单归集后，全额计入订单内每位助教；多助教会导致汇总>门店总额。
+
+月份,排名,助教昵称,指标
+2025-10,1,璇子,80804.14
+2025-10,2,婉婉,62187.64
+2025-10,3,小柔,52623.85
+2025-10,4,小敌,44510.29
+2025-10,5,佳怡,44466.03
+2025-10,6,七七,38810.44
+2025-10,7,奈千,38653.58
+2025-10,8,涛涛,35940.84
+2025-10,9,素素,34135.89
+2025-10,10,球球,33923.75
+2025-10,11,周周,27375.91
+2025-10,12,年糕,26289.89
+2025-10,13,乔西,17649.50
+2025-10,14,Amy,15810.80
+2025-10,15,苏苏,11236.84
+2025-10,16,饭团,7955.28
+2025-10,17,欣怡,4824.69
+2025-10,18,希希,3086.34
+2025-10,19,悦悦,2970.96
+2025-10,20,姜姜,2333.94
+2025-11,1,璇子,154486.83
+2025-11,2,Amy,121568.32
+2025-11,3,小柔,110137.94
+2025-11,4,涛涛,88677.55
+2025-11,5,七七,84500.79
+2025-11,6,佳怡,79249.31
+2025-11,7,奈千,68543.08
+2025-11,8,瑶瑶,65924.36
+2025-11,9,小敌,47986.57
+2025-11,10,球球,41907.39
+2025-11,11,梦梦,39768.09
+2025-11,12,小燕,39426.42
+2025-11,13,阿清,37302.04
+2025-11,14,婉婉,33326.32
+2025-11,15,周周,31436.74
+2025-11,16,小侯,27313.21
+2025-11,17,千千,24684.71
+2025-11,18,柚子,23234.98
+2025-11,19,素素,18707.30
+2025-11,20,年糕,15696.08
+2025-11,21,乔西,15536.78
+2025-11,22,苏苏,10254.59
+2025-11,23,泡芙,8323.03
+2025-11,24,欣怡,5157.29
+2025-11,25,QQ,1134.18
+2025-11,26,西子,303.51
+2025-11,27,希希,281.22

etl_billiards/docs/table_2025-12-19/助教_客户流水排行_2025年10-12月.md

@@ -0,0 +1,73 @@
+# 2025年10-12月 助教客户流水排行榜（全额复制口径）
+## 思考过程
+先把订单应付金额汇总为 order_amount，再把该订单全额计入订单内每位助教，并按月排名。
+
+## 查询说明
+注意：多助教按全额复制计入，导致助教汇总>门店总额，这是刻意口径。
+
+## SQL
+
+### 客户流水（助教+月份汇总，全额复制）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id=g.order_settle_id
+  where g.site_id=%(site_id)s and coalesce(g.is_delete,0)=0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    coalesce(bo.table_amount,0)+coalesce(a.assistant_amount,0)+coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id=bo.order_settle_id
+  left join goods_amount g on g.order_settle_id=bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+)
+select assistant, month_key, sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教_附加课时长排行_2025年10-12月.csv

@@ -0,0 +1,37 @@
+2025年10-12月 助教附加课(超休)时长排行榜
+口径：order_assistant_type=2，超休时长=income_seconds/3600（小时），按月排名。
+
+月份,排名,助教昵称,指标
+2025-10,1,球球,11.00h
+2025-10,2,佳怡,9.00h
+2025-10,3,璇子,8.00h
+2025-10,3,苏苏,8.00h
+2025-10,4,婉婉,4.00h
+2025-10,5,姜姜,3.00h
+2025-10,5,小敌,3.00h
+2025-11,1,周周,25.00h
+2025-11,1,球球,25.00h
+2025-11,2,婉婉,15.00h
+2025-11,3,佳怡,10.00h
+2025-11,3,柚子,10.00h
+2025-11,3,璇子,10.00h
+2025-11,3,素素,10.00h
+2025-11,4,小柔,7.00h
+2025-11,4,年糕,7.00h
+2025-11,5,泡芙,3.00h
+2025-11,5,涛涛,3.00h
+2025-11,5,瑶瑶,3.00h
+2025-11,6,小燕,2.00h
+2025-11,7,乔西,1.00h
+2025-11,7,梦梦,1.00h
+2025-12,1,七七,22.00h
+2025-12,2,小燕,21.00h
+2025-12,3,婉婉,15.00h
+2025-12,3,小柔,15.00h
+2025-12,4,璇子,14.00h
+2025-12,5,周周,8.00h
+2025-12,6,千千,5.00h
+2025-12,6,球球,5.00h
+2025-12,7,佳怡,4.00h
+2025-12,8,QQ,3.00h
+2025-12,9,苏苏,1.00h

etl_billiards/docs/table_2025-12-19/助教_附加课时长排行_2025年10-12月.md

@@ -0,0 +1,31 @@
+# 2025年10-12月 助教附加课(超休)时长排行榜
+## 思考过程
+按月汇总助教附加课时长，并用 dense_rank 做排名。
+
+## 查询说明
+口径：order_assistant_type=2；时长=income_seconds/3600（小时）。
+
+## SQL
+
+### 附加课时长（助教+月份汇总）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0)=0
+    and asl.order_assistant_type=2
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+)
+select
+  assistant,
+  month_key,
+  sum(income_seconds)/3600.0 as hours
+from raw
+where month_key is not null
+group by assistant, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_Amy.csv

@@ -0,0 +1,56 @@
+助教详情：Amy（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础123.6h，附加0.0h；客户流水¥150692.64，充值归因¥12000.00；头部客户(12月)Top3：轩哥、明哥、江先生。
+
+一、基础课业绩
+说明：评价：基础123.6h，附加0.0h；客户流水¥150692.64，充值归因¥12000.00；头部客户(12月)Top3：轩哥、明哥、江先生。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,11.97h,17,-32.44h,-24.56h
+11月,93.53h,4,39.06h,54.65h
+12月,18.08h,16,-26.73h,-23.01h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,0.00h,,0.00h,0.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,15810.80,14,-13468.73,-14839.03
+11月,121568.32,2,77313.93,88242.00
+12月,13313.52,12,-7417.01,-327.12
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,3000.00,13,-11466.67,-10000.00
+11月,9000.00,4,1230.77,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,轩哥,8.8h / 0.0h,4500.95,0.00,36.0h / 0.0h,38175.17,0.00,4.6h / 0.0h,8281.61,0.00
+2,明哥,0.1h / 0.0h,4190.45,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+3,江先生,2.2h / 0.0h,3202.78,0.00,4.5h / 0.0h,1719.24,0.00,0.0h / 0.0h,0.00,0.00
+4,amy,4.8h / 0.0h,1105.90,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+5,小燕,1.0h / 0.0h,313.44,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+6,叶先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,7.4h / 0.0h,7529.19,3000.00
+7,羊,0.0h / 0.0h,0.00,0.00,4.1h / 0.0h,1017.50,1000.00,0.0h / 0.0h,0.00,0.00
+8,葛先生,0.0h / 0.0h,0.00,0.00,1.6h / 0.0h,1585.38,0.00,0.0h / 0.0h,0.00,0.00
+9,陈先生,0.0h / 0.0h,0.00,0.00,4.2h / 0.0h,2566.75,3000.00,0.0h / 0.0h,0.00,0.00
+10,李先生,0.0h / 0.0h,0.00,0.00,1.0h / 0.0h,816.28,0.00,0.0h / 0.0h,0.00,0.00
+11,蔡总,1.2h / 0.0h,0.00,0.00,40.2h / 0.0h,75063.98,5000.00,0.0h / 0.0h,0.00,0.00
+12,昌哥,0.0h / 0.0h,0.00,0.00,2.0h / 0.0h,624.02,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_Amy.md

@@ -0,0 +1,196 @@
+# 助教详情：Amy（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础123.6h，附加0.0h；客户流水¥150692.64，充值归因¥12000.00；头部客户(12月)Top3：轩哥、明哥、江先生。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_QQ.csv

@@ -0,0 +1,47 @@
+助教详情：QQ（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础6.4h，附加3.0h；客户流水¥1517.46，充值归因¥0.00；头部客户(12月)Top3：游、张先生、黄先生。
+
+一、基础课业绩
+说明：评价：基础6.4h，附加3.0h；客户流水¥1517.46，充值归因¥0.00；头部客户(12月)Top3：游、张先生、黄先生。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,5.17h,25,-49.31h,-33.72h
+12月,1.22h,22,-43.59h,-39.88h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,0.00h,,0.00h,0.00h
+12月,3.00h,8,-7.27h,-5.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,1134.18,25,-43120.21,-32192.14
+12月,383.28,21,-20347.25,-13257.36
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,游,1.2h / 0.0h,383.28,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+2,张先生,0.0h / 0.0h,0.00,0.00,2.8h / 0.0h,502.16,0.00,0.0h / 0.0h,0.00,0.00
+3,黄先生,0.0h / 3.0h,0.00,0.00,2.4h / 0.0h,632.02,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_QQ.md

@@ -0,0 +1,196 @@
+# 助教详情：QQ（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础6.4h，附加3.0h；客户流水¥1517.46，充值归因¥0.00；头部客户(12月)Top3：游、张先生、黄先生。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_七七.csv

@@ -0,0 +1,56 @@
+助教详情：七七（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础234.3h，附加22.0h；客户流水¥187346.62，充值归因¥6000.00；头部客户(12月)Top3：蔡总、轩哥、林先生。
+
+一、基础课业绩
+说明：评价：基础234.3h，附加22.0h；客户流水¥187346.62，充值归因¥6000.00；头部客户(12月)Top3：蔡总、轩哥、林先生。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,64.70h,4,20.29h,28.17h
+11月,91.90h,5,37.42h,53.02h
+12月,77.72h,4,32.91h,36.62h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,0.00h,,0.00h,0.00h
+12月,22.00h,1,11.73h,14.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,38810.44,6,9530.91,8160.61
+11月,84500.79,5,40246.40,51174.47
+12月,64035.39,2,43304.86,50394.74
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,6000.00,11,-8466.67,-7000.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,蔡总,26.6h / 0.0h,29660.33,0.00,45.7h / 0.0h,52373.96,0.00,0.0h / 0.0h,0.00,0.00
+2,轩哥,32.8h / 22.0h,27236.57,0.00,24.2h / 0.0h,21849.91,0.00,39.5h / 0.0h,29630.52,3000.00
+3,林先生,14.0h / 0.0h,3808.56,0.00,5.4h / 0.0h,1623.92,0.00,0.0h / 0.0h,0.00,0.00
+4,江先生,3.2h / 0.0h,3042.99,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+5,张先生,1.0h / 0.0h,286.94,0.00,10.9h / 0.0h,3995.10,0.00,13.2h / 0.0h,5315.81,0.00
+6,游,0.0h / 0.0h,0.00,0.00,0.2h / 0.0h,3544.42,0.00,0.0h / 0.0h,0.00,0.00
+7,罗先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,1.1h / 0.0h,545.55,0.00
+8,小熊,0.0h / 0.0h,0.00,0.00,1.3h / 0.0h,314.44,0.00,0.0h / 0.0h,0.00,0.00
+9,叶总,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,3.0h / 0.0h,862.68,0.00
+10,陶,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,4.2h / 0.0h,1278.72,0.00
+11,T,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,3.8h / 0.0h,1177.16,3000.00
+12,胡先生,0.0h / 0.0h,0.00,0.00,4.2h / 0.0h,799.04,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_七七.md

@@ -0,0 +1,196 @@
+# 助教详情：七七（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础234.3h，附加22.0h；客户流水¥187346.62，充值归因¥6000.00；头部客户(12月)Top3：蔡总、轩哥、林先生。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_乔西.csv

@@ -0,0 +1,59 @@
+助教详情：乔西（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础89.0h，附加1.0h；客户流水¥46219.50，充值归因¥20000.00；头部客户(12月)Top3：轩哥、T、林先生。
+
+一、基础课业绩
+说明：评价：基础89.0h，附加1.0h；客户流水¥46219.50，充值归因¥20000.00；头部客户(12月)Top3：轩哥、T、林先生。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,32.60h,11,-11.81h,-3.93h
+11月,30.57h,20,-23.91h,-8.32h
+12月,25.82h,13,-18.99h,-15.28h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,1.00h,7,-7.80h,-6.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,17649.50,13,-11630.03,-13000.33
+11月,15536.78,21,-28717.61,-17789.54
+12月,13033.22,13,-7697.31,-607.42
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,17000.00,7,2533.33,4000.00
+11月,3000.00,8,-4769.23,-6000.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,轩哥,8.2h / 0.0h,8244.84,0.00,8.3h / 0.0h,7623.25,0.00,9.3h / 0.0h,11503.53,8000.00
+2,T,6.8h / 0.0h,1789.02,0.00,4.4h / 0.0h,1100.80,0.00,0.0h / 0.0h,0.00,0.00
+3,林先生,4.6h / 0.0h,1369.51,0.00,5.0h / 0.0h,1645.89,0.00,0.0h / 0.0h,0.00,0.00
+4,张先生,3.3h / 0.0h,1066.81,0.00,0.0h / 0.0h,0.00,0.00,2.5h / 0.0h,489.66,0.00
+5,候,2.9h / 0.0h,563.04,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+6,罗先生,0.0h / 0.0h,0.00,0.00,1.4h / 0.0h,454.46,0.00,0.0h / 0.0h,0.00,0.00
+7,陈腾鑫,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,0.9h / 0.0h,162.29,3000.00
+8,葛先生,0.0h / 0.0h,0.00,0.00,2.2h / 0.0h,707.97,0.00,0.0h / 0.0h,0.00,0.00
+9,陈先生,0.0h / 0.0h,0.00,0.00,4.2h / 0.0h,2566.75,3000.00,0.0h / 0.0h,0.00,0.00
+10,陈淑涛,0.0h / 0.0h,0.00,0.00,3.4h / 0.0h,1021.49,0.00,0.0h / 0.0h,0.00,0.00
+11,周先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,7.5h / 0.0h,2726.01,0.00
+12,陈德韩,0.0h / 0.0h,0.00,0.00,0.0h / 1.0h,0.00,0.00,7.6h / 0.0h,1568.91,5000.00
+13,黄先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,1.6h / 0.0h,300.37,0.00
+14,陈先生,0.0h / 0.0h,0.00,0.00,1.8h / 0.0h,416.17,0.00,0.0h / 0.0h,0.00,0.00
+15,方先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,3.2h / 0.0h,898.73,1000.00

etl_billiards/docs/table_2025-12-19/助教详情_乔西.md

@@ -0,0 +1,196 @@
+# 助教详情：乔西（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础89.0h，附加1.0h；客户流水¥46219.50，充值归因¥20000.00；头部客户(12月)Top3：轩哥、T、林先生。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_佳怡.csv

@@ -0,0 +1,63 @@
+助教详情：佳怡（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础425.2h，附加23.0h；客户流水¥156229.68，充值归因¥47000.00；头部客户(12月)Top3：罗先生、周周、轩哥。
+
+一、基础课业绩
+说明：评价：基础425.2h，附加23.0h；客户流水¥156229.68，充值归因¥47000.00；头部客户(12月)Top3：罗先生、周周、轩哥。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,139.55h,1,95.14h,103.02h
+11月,176.25h,1,121.77h,137.37h
+12月,109.40h,2,64.59h,68.31h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,9.00h,2,2.43h,1.00h
+11月,10.00h,3,1.20h,3.00h
+12月,4.00h,7,-6.27h,-4.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,44466.03,5,15186.50,13816.20
+11月,79249.31,6,34994.92,45922.99
+12月,32514.34,6,11783.81,18873.70
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,27000.00,3,12533.33,14000.00
+11月,20000.00,1,12230.77,11000.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,罗先生,67.8h / 4.0h,16680.01,0.00,57.5h / 10.0h,12309.26,0.00,46.9h / 8.0h,12047.56,7000.00
+2,周周,13.2h / 0.0h,3866.19,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+3,轩哥,6.4h / 0.0h,3741.42,0.00,18.0h / 0.0h,24036.03,0.00,24.6h / 1.0h,17999.86,3000.00
+4,大G,9.5h / 0.0h,2623.97,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+5,T,1.9h / 0.0h,2053.02,0.00,5.0h / 0.0h,2490.70,0.00,8.1h / 0.0h,1323.17,3000.00
+6,林先生,4.5h / 0.0h,1720.12,0.00,10.4h / 0.0h,3269.81,0.00,0.0h / 0.0h,0.00,0.00
+7,游,3.2h / 0.0h,1307.16,0.00,10.3h / 0.0h,4754.69,0.00,0.0h / 0.0h,0.00,0.00
+8,胡先生,3.0h / 0.0h,522.45,0.00,26.4h / 0.0h,9712.14,13000.00,0.0h / 0.0h,0.00,0.00
+9,江先生,0.0h / 0.0h,0.00,0.00,3.8h / 0.0h,1374.85,0.00,0.0h / 0.0h,0.00,0.00
+10,陈先生,0.0h / 0.0h,0.00,0.00,7.6h / 0.0h,2566.75,3000.00,0.0h / 0.0h,0.00,0.00
+11,陈腾鑫,0.0h / 0.0h,0.00,0.00,19.2h / 0.0h,4276.97,1000.00,38.5h / 0.0h,7626.75,12000.00
+12,张先生,0.0h / 0.0h,0.00,0.00,4.9h / 0.0h,2233.65,0.00,0.0h / 0.0h,0.00,0.00
+13,歌神,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,2.0h / 0.0h,361.38,0.00
+14,夏,0.0h / 0.0h,0.00,0.00,2.0h / 0.0h,455.03,0.00,5.8h / 0.0h,2595.54,1000.00
+15,小熊,0.0h / 0.0h,0.00,0.00,6.4h / 0.0h,2072.54,3000.00,6.8h / 0.0h,1213.06,1000.00
+16,蔡总,0.0h / 0.0h,0.00,0.00,4.5h / 0.0h,9696.89,0.00,0.0h / 0.0h,0.00,0.00
+17,吕先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,1.7h / 0.0h,337.17,0.00
+18,贺斌,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,1.0h / 0.0h,247.34,0.00
+19,陶,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,4.1h / 0.0h,714.20,0.00

etl_billiards/docs/table_2025-12-19/助教详情_佳怡.md

@@ -0,0 +1,196 @@
+# 助教详情：佳怡（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础425.2h，附加23.0h；客户流水¥156229.68，充值归因¥47000.00；头部客户(12月)Top3：罗先生、周周、轩哥。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_千千.csv

@@ -0,0 +1,55 @@
+助教详情：千千（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础77.2h，附加5.0h；客户流水¥34297.90，充值归因¥0.00；头部客户(12月)Top3：张先生、周先生、陈腾鑫。
+
+一、基础课业绩
+说明：评价：基础77.2h，附加5.0h；客户流水¥34297.90，充值归因¥0.00；头部客户(12月)Top3：张先生、周先生、陈腾鑫。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,38.88h,14,-15.59h,0.00h
+12月,38.28h,12,-6.53h,-2.81h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,0.00h,,0.00h,0.00h
+12月,5.00h,6,-5.27h,-3.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,24684.71,17,-19569.68,-8641.61
+12月,9613.19,16,-11117.34,-4027.46
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,张先生,10.0h / 0.0h,2622.05,0.00,4.4h / 0.0h,1623.68,0.00,0.0h / 0.0h,0.00,0.00
+2,周先生,8.6h / 0.0h,1577.94,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+3,陈腾鑫,4.8h / 0.0h,1375.20,0.00,2.9h / 0.0h,2418.94,0.00,0.0h / 0.0h,0.00,0.00
+4,梅,3.3h / 5.0h,1356.34,0.00,5.8h / 0.0h,2007.81,0.00,0.0h / 0.0h,0.00,0.00
+5,清,3.0h / 0.0h,1128.06,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+6,黄先生,6.2h / 0.0h,1048.14,0.00,6.8h / 0.0h,1251.92,0.00,0.0h / 0.0h,0.00,0.00
+7,游,2.2h / 0.0h,505.46,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+8,葛先生,0.0h / 0.0h,0.00,0.00,14.4h / 0.0h,7500.34,0.00,0.0h / 0.0h,0.00,0.00
+9,林先生,0.0h / 0.0h,0.00,0.00,2.7h / 0.0h,499.11,0.00,0.0h / 0.0h,0.00,0.00
+10,轩哥,0.0h / 0.0h,0.00,0.00,1.2h / 0.0h,415.60,0.00,0.0h / 0.0h,0.00,0.00
+11,蔡总,0.0h / 0.0h,0.00,0.00,0.7h / 0.0h,8967.31,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_千千.md

@@ -0,0 +1,196 @@
+# 助教详情：千千（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础77.2h，附加5.0h；客户流水¥34297.90，充值归因¥0.00；头部客户(12月)Top3：张先生、周先生、陈腾鑫。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_周周.csv

@@ -0,0 +1,66 @@
+助教详情：周周（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础172.6h，附加33.0h；客户流水¥80429.10，充值归因¥20000.00；头部客户(12月)Top3：周周、明哥、T。
+
+一、基础课业绩
+说明：评价：基础172.6h，附加33.0h；客户流水¥80429.10，充值归因¥20000.00；头部客户(12月)Top3：周周、明哥、T。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,41.33h,9,-3.08h,4.81h
+11月,71.27h,10,16.79h,32.38h
+12月,60.02h,6,15.21h,18.92h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,25.00h,1,16.20h,18.00h
+12月,8.00h,5,-2.27h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,27375.91,11,-1903.62,-3273.92
+11月,31436.74,15,-12817.65,-1889.58
+12月,21616.45,7,885.92,7975.80
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,11000.00,9,-3466.67,-2000.00
+11月,9000.00,4,1230.77,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,周周,28.8h / 8.0h,8105.19,0.00,0.0h / 20.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+2,明哥,0.4h / 0.0h,4190.45,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+3,T,10.1h / 0.0h,2968.13,0.00,0.0h / 0.0h,0.00,0.00,4.5h / 0.0h,1300.23,0.00
+4,大G,14.8h / 0.0h,2724.15,0.00,2.8h / 0.0h,1783.61,0.00,0.0h / 0.0h,0.00,0.00
+5,罗先生,2.6h / 0.0h,1584.22,0.00,9.0h / 0.0h,2415.09,0.00,0.0h / 0.0h,0.00,0.00
+6,游,2.4h / 0.0h,1307.16,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+7,张先生,0.0h / 0.0h,449.03,0.00,6.3h / 0.0h,2092.46,0.00,4.1h / 0.0h,1650.08,0.00
+8,轩哥,0.7h / 0.0h,288.12,0.00,5.0h / 3.0h,2175.94,5000.00,20.2h / 0.0h,16154.38,10000.00
+9,江先生,0.0h / 0.0h,0.00,0.00,4.3h / 0.0h,3588.88,0.00,0.0h / 0.0h,0.00,0.00
+10,罗超杰,0.0h / 0.0h,0.00,0.00,1.2h / 0.0h,255.63,0.00,0.0h / 0.0h,0.00,0.00
+11,陈腾鑫,0.0h / 0.0h,0.00,0.00,4.6h / 0.0h,1210.78,0.00,0.0h / 0.0h,197.60,1000.00
+12,林总,0.0h / 0.0h,0.00,0.00,2.2h / 0.0h,439.96,0.00,0.0h / 0.0h,0.00,0.00
+13,林先生,0.0h / 0.0h,0.00,0.00,0.7h / 0.0h,747.44,0.00,0.0h / 0.0h,0.00,0.00
+14,葛先生,0.0h / 0.0h,0.00,0.00,11.0h / 0.0h,3073.27,0.00,0.0h / 0.0h,0.00,0.00
+15,羊,0.0h / 0.0h,0.00,0.00,1.4h / 0.0h,602.50,0.00,0.0h / 0.0h,0.00,0.00
+16,小熊,0.0h / 0.0h,0.00,0.00,10.7h / 0.0h,2612.37,4000.00,0.0h / 0.0h,0.00,0.00
+17,陈德韩,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,4.3h / 0.0h,5714.01,0.00
+18,蔡总,0.0h / 0.0h,0.00,0.00,8.2h / 0.0h,9385.22,0.00,0.0h / 0.0h,0.00,0.00
+19,万先生,0.0h / 0.0h,0.00,0.00,0.0h / 2.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+20,吕先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,3.7h / 0.0h,1080.89,0.00
+21,林先生,0.0h / 0.0h,0.00,0.00,4.0h / 0.0h,1053.59,0.00,0.0h / 0.0h,0.00,0.00
+22,陶,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,4.4h / 0.0h,1278.72,0.00

etl_billiards/docs/table_2025-12-19/助教详情_周周.md

@@ -0,0 +1,196 @@
+# 助教详情：周周（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础172.6h，附加33.0h；客户流水¥80429.10，充值归因¥20000.00；头部客户(12月)Top3：周周、明哥、T。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_奈千.csv

@@ -0,0 +1,58 @@
+助教详情：奈千（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础59.0h，附加0.0h；客户流水¥107484.78，充值归因¥18000.00；头部客户(12月)Top3：轩哥、黎先生、陈先生。
+
+一、基础课业绩
+说明：评价：基础59.0h，附加0.0h；客户流水¥107484.78，充值归因¥18000.00；头部客户(12月)Top3：轩哥、黎先生、陈先生。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,23.62h,13,-20.79h,-12.91h
+11月,32.83h,19,-21.64h,-6.05h
+12月,2.58h,21,-42.23h,-38.51h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,0.00h,,0.00h,0.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,38653.58,7,9374.05,8003.75
+11月,68543.08,7,24288.69,35216.76
+12月,288.12,22,-20442.41,-13352.52
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,18000.00,6,3533.33,5000.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,轩哥,0.7h / 0.0h,288.12,0.00,4.3h / 0.0h,14349.45,0.00,6.2h / 0.0h,22341.76,13000.00
+2,黎先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,1.7h / 0.0h,470.19,0.00
+3,陈先生,0.0h / 0.0h,0.00,0.00,5.1h / 0.0h,1150.83,0.00,0.0h / 0.0h,0.00,0.00
+4,羊,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,1.4h / 0.0h,3620.75,0.00
+5,陈腾鑫,0.0h / 0.0h,0.00,0.00,2.6h / 0.0h,441.39,0.00,2.1h / 0.0h,1084.69,1000.00
+6,张先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,2.2h / 0.0h,474.93,0.00
+7,夏,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,2.0h / 0.0h,2595.54,1000.00
+8,罗超,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,2.5h / 0.0h,1013.51,0.00
+9,罗先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,1143.87,3000.00
+10,陈德韩,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,4.4h / 0.0h,5714.01,0.00
+11,阿亮,0.0h / 0.0h,0.00,0.00,2.9h / 0.0h,493.02,0.00,0.0h / 0.0h,0.00,0.00
+12,曾先生,0.0h / 0.0h,0.00,0.00,1.0h / 0.0h,206.16,0.00,0.0h / 0.0h,0.00,0.00
+13,蔡总,1.8h / 0.0h,0.00,0.00,17.0h / 0.0h,51902.23,0.00,0.0h / 0.0h,0.00,0.00
+14,T,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,1.2h / 0.0h,194.33,0.00

etl_billiards/docs/table_2025-12-19/助教详情_奈千.md

@@ -0,0 +1,196 @@
+# 助教详情：奈千（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础59.0h，附加0.0h；客户流水¥107484.78，充值归因¥18000.00；头部客户(12月)Top3：轩哥、黎先生、陈先生。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_姜姜.csv

@@ -0,0 +1,46 @@
+助教详情：姜姜（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础6.6h，附加3.0h；客户流水¥2333.94，充值归因¥4000.00；头部客户(12月)Top3：罗先生、汪先生。
+
+一、基础课业绩
+说明：评价：基础6.6h，附加3.0h；客户流水¥2333.94，充值归因¥4000.00；头部客户(12月)Top3：罗先生、汪先生。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,6.60h,18,-37.81h,-29.93h
+11月,0.00h,,0.00h,0.00h
+12月,0.00h,,0.00h,0.00h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,3.00h,5,-3.57h,-5.00h
+11月,0.00h,,0.00h,0.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,2333.94,20,-26945.59,-28315.89
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,4000.00,12,-10466.67,-9000.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,罗先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,1.5h / 0.0h,1143.87,3000.00
+2,汪先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,5.1h / 3.0h,1190.07,1000.00

etl_billiards/docs/table_2025-12-19/助教详情_姜姜.md

@@ -0,0 +1,196 @@
+# 助教详情：姜姜（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础6.6h，附加3.0h；客户流水¥2333.94，充值归因¥4000.00；头部客户(12月)Top3：罗先生、汪先生。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_婉婉.csv

@@ -0,0 +1,65 @@
+助教详情：婉婉（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础154.6h，附加34.0h；客户流水¥106476.37，充值归因¥33000.00；头部客户(12月)Top3：江先生、明哥、候。
+
+一、基础课业绩
+说明：评价：基础154.6h，附加34.0h；客户流水¥106476.37，充值归因¥33000.00；头部客户(12月)Top3：江先生、明哥、候。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,90.68h,3,46.27h,54.16h
+11月,46.03h,12,-8.44h,7.15h
+12月,17.83h,17,-26.98h,-23.26h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,4.00h,4,-2.57h,-4.00h
+11月,15.00h,2,6.20h,8.00h
+12月,15.00h,3,4.73h,7.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,62187.64,2,32908.11,31537.81
+11月,33326.32,14,-10928.07,0.00
+12月,10962.41,15,-9768.12,-2678.24
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,24000.00,4,9533.33,11000.00
+11月,9000.00,4,1230.77,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,江先生,4.7h / 0.0h,5889.30,0.00,9.4h / 0.0h,6947.07,5000.00,2.7h / 0.0h,1538.09,0.00
+2,明哥,11.7h / 10.0h,4822.90,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+3,候,1.1h / 0.0h,195.75,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+4,刘哥,0.3h / 0.0h,54.46,0.00,5.0h / 0.0h,2982.34,0.00,0.0h / 0.0h,0.00,0.00
+5,林总,0.0h / 0.0h,0.00,0.00,0.9h / 0.0h,244.48,0.00,0.0h / 0.0h,0.00,0.00
+6,羊,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,4.3h / 0.0h,3620.75,0.00
+7,叶先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,3.7h / 0.0h,1278.01,0.00
+8,邓飛,0.0h / 0.0h,0.00,0.00,0.8h / 0.0h,925.47,1000.00,0.0h / 4.0h,0.00,0.00
+9,歌神,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,6.7h / 0.0h,4718.40,0.00
+10,夏,0.0h / 0.0h,0.00,0.00,5.4h / 4.0h,2991.13,0.00,18.2h / 0.0h,11826.32,1000.00
+11,轩哥,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,45.2h / 0.0h,33043.15,16000.00
+12,罗先生,0.0h / 5.0h,0.00,0.00,7.9h / 0.0h,2086.94,0.00,3.4h / 0.0h,1143.87,3000.00
+13,蔡总,0.0h / 0.0h,0.00,0.00,0.0h / 3.0h,6196.43,0.00,0.0h / 0.0h,0.00,0.00
+14,吕先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,2.8h / 0.0h,3924.18,3000.00
+15,万先生,0.0h / 0.0h,0.00,0.00,0.0h / 4.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+16,老宋,0.0h / 0.0h,0.00,0.00,1.0h / 0.0h,465.98,0.00,0.0h / 0.0h,0.00,0.00
+17,黎先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,0.8h / 0.0h,470.19,0.00
+18,君姐,0.0h / 0.0h,0.00,0.00,4.2h / 0.0h,1864.86,0.00,0.0h / 0.0h,0.00,0.00
+19,林先生,0.0h / 0.0h,0.00,0.00,6.0h / 4.0h,2690.52,0.00,0.0h / 0.0h,0.00,0.00
+20,婉婉,0.0h / 0.0h,0.00,0.00,1.6h / 0.0h,242.81,0.00,2.8h / 0.0h,624.68,1000.00
+21,胡先生,0.0h / 0.0h,0.00,0.00,3.8h / 0.0h,5688.29,3000.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_婉婉.md

@@ -0,0 +1,196 @@
+# 助教详情：婉婉（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础154.6h，附加34.0h；客户流水¥106476.37，充值归因¥33000.00；头部客户(12月)Top3：江先生、明哥、候。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_小侯.csv

@@ -0,0 +1,57 @@
+助教详情：小侯（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础92.2h，附加0.0h；客户流水¥41280.98，充值归因¥0.00；头部客户(12月)Top3：张先生、陈腾鑫、李先生。
+
+一、基础课业绩
+说明：评价：基础92.2h，附加0.0h；客户流水¥41280.98，充值归因¥0.00；头部客户(12月)Top3：张先生、陈腾鑫、李先生。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,42.58h,13,-11.89h,3.70h
+12月,49.57h,8,4.76h,8.47h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,0.00h,,0.00h,0.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,27313.21,16,-16941.18,-6013.11
+12月,13967.77,11,-6762.76,327.12
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,张先生,14.4h / 0.0h,4161.57,0.00,5.8h / 0.0h,1583.86,0.00,0.0h / 0.0h,0.00,0.00
+2,陈腾鑫,12.1h / 0.0h,3984.45,0.00,7.0h / 0.0h,2965.62,0.00,0.0h / 0.0h,0.00,0.00
+3,李先生,9.3h / 0.0h,1729.57,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+4,梅,3.3h / 0.0h,1356.34,0.00,1.4h / 0.0h,1573.10,0.00,0.0h / 0.0h,0.00,0.00
+5,清,3.0h / 0.0h,1128.06,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+6,T,3.9h / 0.0h,938.16,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+7,候,3.4h / 0.0h,669.62,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+8,林志铭,0.0h / 0.0h,0.00,0.00,1.2h / 0.0h,220.15,0.00,0.0h / 0.0h,0.00,0.00
+9,林先生,0.0h / 0.0h,0.00,0.00,9.7h / 0.0h,3619.37,0.00,0.0h / 0.0h,0.00,0.00
+10,艾宇民,0.0h / 0.0h,0.00,0.00,7.6h / 0.0h,3872.24,0.00,0.0h / 0.0h,0.00,0.00
+11,蔡总,0.0h / 0.0h,0.00,0.00,2.6h / 0.0h,10419.30,0.00,0.0h / 0.0h,0.00,0.00
+12,钟智豪,0.0h / 0.0h,0.00,0.00,1.8h / 0.0h,274.34,0.00,0.0h / 0.0h,0.00,0.00
+13,李先生,0.0h / 0.0h,0.00,0.00,5.6h / 0.0h,2785.23,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_小侯.md

@@ -0,0 +1,196 @@
+# 助教详情：小侯（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础92.2h，附加0.0h；客户流水¥41280.98，充值归因¥0.00；头部客户(12月)Top3：张先生、陈腾鑫、李先生。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_小敌.csv

@@ -0,0 +1,55 @@
+助教详情：小敌（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础135.1h，附加3.0h；客户流水¥93760.82，充值归因¥29000.00；头部客户(12月)Top3：郑先生、张先生、轩哥。
+
+一、基础课业绩
+说明：评价：基础135.1h，附加3.0h；客户流水¥93760.82，充值归因¥29000.00；头部客户(12月)Top3：郑先生、张先生、轩哥。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,55.82h,7,11.41h,19.29h
+11月,72.90h,9,18.42h,34.02h
+12月,6.40h,20,-38.41h,-34.69h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,3.00h,5,-3.57h,-5.00h
+11月,0.00h,,0.00h,0.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,44510.29,4,15230.76,13860.46
+11月,47986.57,9,3732.18,14660.25
+12月,1263.96,20,-19466.57,-12376.68
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,21000.00,5,6533.33,8000.00
+11月,8000.00,5,230.77,-1000.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,郑先生,4.9h / 0.0h,814.93,0.00,27.3h / 0.0h,4745.90,0.00,0.0h / 0.0h,0.00,0.00
+2,张先生,1.5h / 0.0h,449.03,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+3,轩哥,0.0h / 0.0h,0.00,0.00,11.2h / 0.0h,15962.93,0.00,43.1h / 0.0h,32148.10,21000.00
+4,游,0.0h / 0.0h,0.00,0.00,3.6h / 0.0h,3791.20,0.00,0.0h / 0.0h,0.00,0.00
+5,叶先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,2.1h / 0.0h,3922.17,0.00
+6,李先生,0.0h / 0.0h,0.00,0.00,11.8h / 0.0h,2997.53,3000.00,0.0h / 0.0h,0.00,0.00
+7,周先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,6.6h / 0.0h,2726.01,0.00
+8,陈德韩,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,4.0h / 0.0h,5714.01,0.00
+9,蔡总,0.0h / 0.0h,0.00,0.00,17.2h / 0.0h,19881.95,5000.00,0.0h / 0.0h,0.00,0.00
+10,林先生,0.0h / 0.0h,0.00,0.00,1.9h / 0.0h,607.06,0.00,0.0h / 0.0h,0.00,0.00
+11,邓飛,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 3.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_小敌.md

@@ -0,0 +1,196 @@
+# 助教详情：小敌（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础135.1h，附加3.0h；客户流水¥93760.82，充值归因¥29000.00；头部客户(12月)Top3：郑先生、张先生、轩哥。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_小柔.csv

@@ -0,0 +1,66 @@
+助教详情：小柔（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础207.4h，附加22.0h；客户流水¥209918.77，充值归因¥42700.00；头部客户(12月)Top3：蔡总、轩哥、明哥。
+
+一、基础课业绩
+说明：评价：基础207.4h，附加22.0h；客户流水¥209918.77，充值归因¥42700.00；头部客户(12月)Top3：蔡总、轩哥、明哥。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,63.87h,5,19.46h,27.34h
+11月,88.65h,6,34.17h,49.77h
+12月,54.93h,7,10.12h,13.84h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,7.00h,4,-1.80h,0.00h
+12月,15.00h,3,4.73h,7.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,52623.85,3,23344.32,21974.02
+11月,110137.94,3,65883.55,76811.62
+12月,47156.98,3,26426.45,33516.34
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,31700.00,2,17233.33,18700.00
+11月,11000.00,2,3230.77,2000.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,蔡总,30.2h / 0.0h,25658.77,0.00,62.2h / 3.0h,88451.79,5000.00,0.0h / 0.0h,0.00,0.00
+2,轩哥,8.4h / 0.0h,13202.16,0.00,2.0h / 0.0h,4130.37,0.00,35.2h / 0.0h,33211.31,23000.00
+3,明哥,7.7h / 12.0h,4190.45,0.00,5.8h / 0.0h,2258.14,0.00,0.0h / 0.0h,0.00,0.00
+4,江先生,3.2h / 0.0h,3042.99,0.00,6.1h / 0.0h,7578.00,5000.00,2.7h / 0.0h,1538.09,0.00
+5,T,2.0h / 0.0h,434.21,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+6,昌哥,1.8h / 0.0h,318.40,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+7,陈腾鑫,1.6h / 0.0h,310.00,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+8,罗先生,0.0h / 0.0h,0.00,0.00,4.1h / 0.0h,722.25,0.00,0.0h / 0.0h,0.00,0.00
+9,邓飛,0.0h / 0.0h,0.00,0.00,1.2h / 0.0h,925.47,1000.00,0.0h / 0.0h,0.00,0.00
+10,吴生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,2.8h / 0.0h,685.89,0.00
+11,歌神,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,3.8h / 0.0h,2789.21,0.00
+12,羊,0.0h / 0.0h,0.00,0.00,1.6h / 0.0h,383.40,0.00,0.0h / 0.0h,0.00,0.00
+13,张先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,3.3h / 0.0h,568.87,0.00
+14,游,0.0h / 0.0h,0.00,0.00,2.2h / 0.0h,935.84,0.00,0.0h / 0.0h,0.00,0.00
+15,陈德韩,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,3.0h / 0.0h,5714.01,0.00
+16,吕先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,7.5h / 0.0h,2750.70,3000.00
+17,罗先生,0.0h / 3.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+18,万先生,0.0h / 0.0h,0.00,0.00,0.0h / 4.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+19,林先生,0.0h / 0.0h,0.00,0.00,3.2h / 0.0h,2398.29,0.00,0.0h / 0.0h,0.00,0.00
+20,陶,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,1.7h / 0.0h,362.17,0.00
+21,叶总,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,3.8h / 0.0h,5003.60,5700.00
+22,胡先生,0.0h / 0.0h,0.00,0.00,0.1h / 0.0h,2354.39,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_小柔.md

@@ -0,0 +1,196 @@
+# 助教详情：小柔（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础207.4h，附加22.0h；客户流水¥209918.77，充值归因¥42700.00；头部客户(12月)Top3：蔡总、轩哥、明哥。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_小柳.csv

@@ -0,0 +1,44 @@
+助教详情：小柳（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础0.0h，附加0.0h；客户流水¥0.00，充值归因¥0.00；头部客户(12月)Top3：无。
+
+一、基础课业绩
+说明：评价：基础0.0h，附加0.0h；客户流水¥0.00，充值归因¥0.00；头部客户(12月)Top3：无。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,0.00h,,0.00h,0.00h
+12月,0.00h,,0.00h,0.00h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,0.00h,,0.00h,0.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)

etl_billiards/docs/table_2025-12-19/助教详情_小柳.md

@@ -0,0 +1,196 @@
+# 助教详情：小柳（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础0.0h，附加0.0h；客户流水¥0.00，充值归因¥0.00；头部客户(12月)Top3：无。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_小燕.csv

@@ -0,0 +1,48 @@
+助教详情：小燕（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础268.3h，附加23.0h；客户流水¥77172.28，充值归因¥0.00；头部客户(12月)Top3：葛先生、小燕、梅。
+
+一、基础课业绩
+说明：评价：基础268.3h，附加23.0h；客户流水¥77172.28，充值归因¥0.00；头部客户(12月)Top3：葛先生、小燕、梅。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,109.28h,3,54.81h,70.40h
+12月,159.02h,1,114.21h,117.92h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,2.00h,6,-6.80h,-5.00h
+12月,21.00h,2,10.73h,13.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,39426.42,12,-4827.97,6100.10
+12月,37745.86,5,17015.33,24105.22
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,葛先生,80.8h / 16.0h,22323.38,0.00,53.5h / 0.0h,17941.19,0.00,0.0h / 0.0h,0.00,0.00
+2,小燕,78.2h / 5.0h,15422.48,0.00,51.4h / 2.0h,10944.82,0.00,0.0h / 0.0h,0.00,0.00
+3,梅,0.0h / 0.0h,0.00,0.00,3.8h / 0.0h,1573.10,0.00,0.0h / 0.0h,0.00,0.00
+4,蔡总,0.0h / 0.0h,0.00,0.00,0.7h / 0.0h,8967.31,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_小燕.md

@@ -0,0 +1,196 @@
+# 助教详情：小燕（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础268.3h，附加23.0h；客户流水¥77172.28，充值归因¥0.00；头部客户(12月)Top3：葛先生、小燕、梅。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_希希.csv

@@ -0,0 +1,47 @@
+助教详情：希希（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础6.6h，附加0.0h；客户流水¥3367.56，充值归因¥0.00；头部客户(12月)Top3：陈腾鑫、歌神、郭先生。
+
+一、基础课业绩
+说明：评价：基础6.6h，附加0.0h；客户流水¥3367.56，充值归因¥0.00；头部客户(12月)Top3：陈腾鑫、歌神、郭先生。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,5.02h,19,-39.39h,-31.51h
+11月,1.58h,27,-52.89h,-37.30h
+12月,0.00h,,0.00h,0.00h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,0.00h,,0.00h,0.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,3086.34,18,-26193.19,-27563.49
+11月,281.22,27,-43973.17,-33045.10
+12月,0.00,,0.00,0.00
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,陈腾鑫,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,1.0h / 0.0h,359.28,0.00
+2,歌神,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,4.0h / 0.0h,2727.06,0.00
+3,郭先生,0.0h / 0.0h,0.00,0.00,1.6h / 0.0h,281.22,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_希希.md

@@ -0,0 +1,196 @@
+# 助教详情：希希（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础6.6h，附加0.0h；客户流水¥3367.56，充值归因¥0.00；头部客户(12月)Top3：陈腾鑫、歌神、郭先生。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_年糕.csv

@@ -0,0 +1,63 @@
+助教详情：年糕（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础82.6h，附加7.0h；客户流水¥59830.16，充值归因¥9000.00；头部客户(12月)Top3：葛先生、明哥、蔡总。
+
+一、基础课业绩
+说明：评价：基础82.6h，附加7.0h；客户流水¥59830.16，充值归因¥9000.00；头部客户(12月)Top3：葛先生、明哥、蔡总。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,21.23h,14,-23.18h,-15.29h
+11月,35.80h,15,-18.68h,-3.08h
+12月,25.62h,14,-19.19h,-15.48h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,7.00h,4,-1.80h,0.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,26289.89,12,-2989.64,-4359.94
+11月,15696.08,20,-28558.31,-17630.24
+12月,17844.19,9,-2886.34,4203.54
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,9000.00,10,-5466.67,-4000.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,葛先生,2.6h / 0.0h,5551.79,0.00,6.7h / 0.0h,3777.09,0.00,0.0h / 0.0h,0.00,0.00
+2,明哥,4.6h / 0.0h,4190.45,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+3,蔡总,6.0h / 0.0h,2130.39,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+4,叶先生,2.6h / 0.0h,1558.40,0.00,1.8h / 0.0h,711.79,0.00,5.3h / 0.0h,3607.02,3000.00
+5,君姐,2.2h / 0.0h,1414.23,0.00,6.0h / 0.0h,1864.86,0.00,0.0h / 0.0h,0.00,0.00
+6,林先生,0.9h / 0.0h,1369.51,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+7,张先生,2.5h / 0.0h,540.07,0.00,4.2h / 0.0h,1007.52,0.00,2.1h / 0.0h,596.24,0.00
+8,潘先生,2.0h / 0.0h,516.93,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+9,常总,1.7h / 0.0h,460.52,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+10,小燕,0.5h / 0.0h,111.90,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+11,谢俊,0.0h / 0.0h,0.00,0.00,4.5h / 0.0h,794.53,0.00,0.0h / 0.0h,0.00,0.00
+12,歌神,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,2.6h / 0.0h,1929.19,0.00
+13,夏,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,1.2h / 0.0h,4670.88,0.00
+14,轩哥,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,7.2h / 0.0h,12643.27,3000.00
+15,罗先生,0.0h / 0.0h,0.00,0.00,7.6h / 0.0h,2400.67,0.00,0.0h / 0.0h,0.00,0.00
+16,吕先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,2.8h / 0.0h,2843.29,3000.00
+17,万先生,0.0h / 0.0h,0.00,0.00,0.0h / 7.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+18,胡先生,0.0h / 0.0h,0.00,0.00,0.8h / 0.0h,2354.39,0.00,0.0h / 0.0h,0.00,0.00
+19,李先生,0.0h / 0.0h,0.00,0.00,4.2h / 0.0h,2785.23,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_年糕.md

@@ -0,0 +1,196 @@
+# 助教详情：年糕（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础82.6h，附加7.0h；客户流水¥59830.16，充值归因¥9000.00；头部客户(12月)Top3：葛先生、明哥、蔡总。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_悦悦.csv

@@ -0,0 +1,46 @@
+助教详情：悦悦（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础2.3h，附加0.0h；客户流水¥2970.96，充值归因¥3000.00；头部客户(12月)Top3：小宇、轩哥。
+
+一、基础课业绩
+说明：评价：基础2.3h，附加0.0h；客户流水¥2970.96，充值归因¥3000.00；头部客户(12月)Top3：小宇、轩哥。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,2.30h,20,-42.11h,-34.23h
+11月,0.00h,,0.00h,0.00h
+12月,0.00h,,0.00h,0.00h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,0.00h,,0.00h,0.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,2970.96,19,-26308.57,-27678.87
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,3000.00,13,-11466.67,-10000.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,小宇,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,0.4h / 0.0h,78.52,0.00
+2,轩哥,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,1.9h / 0.0h,2892.44,3000.00

etl_billiards/docs/table_2025-12-19/助教详情_悦悦.md

@@ -0,0 +1,196 @@
+# 助教详情：悦悦（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础2.3h，附加0.0h；客户流水¥2970.96，充值归因¥3000.00；头部客户(12月)Top3：小宇、轩哥。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_柚子.csv

@@ -0,0 +1,56 @@
+助教详情：柚子（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础35.4h，附加10.0h；客户流水¥23234.98，充值归因¥3000.00；头部客户(12月)Top3：陈先生、羊、葛先生。
+
+一、基础课业绩
+说明：评价：基础35.4h，附加10.0h；客户流水¥23234.98，充值归因¥3000.00；头部客户(12月)Top3：陈先生、羊、葛先生。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,35.40h,16,-19.08h,-3.48h
+12月,0.00h,,0.00h,0.00h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,10.00h,3,1.20h,3.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,23234.98,18,-21019.41,-10091.34
+12月,0.00,,0.00,0.00
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,3000.00,8,-4769.23,-6000.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,陈先生,0.0h / 0.0h,0.00,0.00,4.4h / 0.0h,1150.83,0.00,0.0h / 0.0h,0.00,0.00
+2,羊,0.0h / 0.0h,0.00,0.00,1.6h / 0.0h,429.33,0.00,0.0h / 0.0h,0.00,0.00
+3,葛先生,0.0h / 0.0h,0.00,0.00,3.4h / 0.0h,3052.28,0.00,0.0h / 0.0h,0.00,0.00
+4,陈腾鑫,0.0h / 0.0h,0.00,0.00,0.8h / 0.0h,472.03,0.00,0.0h / 0.0h,0.00,0.00
+5,张先生,0.0h / 0.0h,0.00,0.00,3.2h / 0.0h,941.37,0.00,0.0h / 0.0h,0.00,0.00
+6,夏,0.0h / 0.0h,0.00,0.00,0.2h / 7.0h,2991.13,0.00,0.0h / 0.0h,0.00,0.00
+7,轩哥,0.0h / 0.0h,0.00,0.00,0.2h / 0.0h,4484.68,0.00,0.0h / 0.0h,0.00,0.00
+8,胡先生,0.0h / 0.0h,0.00,0.00,1.0h / 0.0h,5688.29,3000.00,0.0h / 0.0h,0.00,0.00
+9,李先生,0.0h / 0.0h,0.00,0.00,7.8h / 0.0h,1712.78,0.00,0.0h / 0.0h,0.00,0.00
+10,牛先生,0.0h / 0.0h,0.00,0.00,10.4h / 0.0h,1887.48,0.00,0.0h / 0.0h,0.00,0.00
+11,阿亮,0.0h / 0.0h,0.00,0.00,2.4h / 0.0h,424.78,0.00,0.0h / 0.0h,0.00,0.00
+12,万先生,0.0h / 0.0h,0.00,0.00,0.0h / 3.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_柚子.md

@@ -0,0 +1,196 @@
+# 助教详情：柚子（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础35.4h，附加10.0h；客户流水¥23234.98，充值归因¥3000.00；头部客户(12月)Top3：陈先生、羊、葛先生。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_梦梦.csv

@@ -0,0 +1,51 @@
+助教详情：梦梦（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础35.7h，附加1.0h；客户流水¥45991.89，充值归因¥0.00；头部客户(12月)Top3：葛先生、阿亮、蔡总。
+
+一、基础课业绩
+说明：评价：基础35.7h，附加1.0h；客户流水¥45991.89，充值归因¥0.00；头部客户(12月)Top3：葛先生、阿亮、蔡总。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,19.60h,22,-34.88h,-19.28h
+12月,16.08h,18,-28.73h,-25.01h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,1.00h,7,-7.80h,-6.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,39768.09,11,-4486.30,6441.77
+12月,6223.80,18,-14506.73,-7416.84
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,葛先生,2.9h / 0.0h,5551.79,0.00,4.4h / 0.0h,1884.66,0.00,0.0h / 0.0h,0.00,0.00
+2,阿亮,2.7h / 0.0h,495.46,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+3,蔡总,10.5h / 0.0h,176.55,0.00,7.0h / 1.0h,34274.25,0.00,0.0h / 0.0h,0.00,0.00
+4,轩哥,0.0h / 0.0h,0.00,0.00,1.0h / 0.0h,363.25,0.00,0.0h / 0.0h,0.00,0.00
+5,羊,0.0h / 0.0h,0.00,0.00,0.6h / 0.0h,104.77,0.00,0.0h / 0.0h,0.00,0.00
+6,林先生,0.0h / 0.0h,0.00,0.00,4.3h / 0.0h,2461.85,0.00,0.0h / 0.0h,0.00,0.00
+7,张先生,0.0h / 0.0h,0.00,0.00,2.3h / 0.0h,679.31,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_梦梦.md

@@ -0,0 +1,196 @@
+# 助教详情：梦梦（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础35.7h，附加1.0h；客户流水¥45991.89，充值归因¥0.00；头部客户(12月)Top3：葛先生、阿亮、蔡总。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_楚楚.csv

@@ -0,0 +1,44 @@
+助教详情：楚楚（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础0.0h，附加0.0h；客户流水¥0.00，充值归因¥0.00；头部客户(12月)Top3：无。
+
+一、基础课业绩
+说明：评价：基础0.0h，附加0.0h；客户流水¥0.00，充值归因¥0.00；头部客户(12月)Top3：无。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,0.00h,,0.00h,0.00h
+12月,0.00h,,0.00h,0.00h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,0.00h,,0.00h,0.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)

etl_billiards/docs/table_2025-12-19/助教详情_楚楚.md

@@ -0,0 +1,196 @@
+# 助教详情：楚楚（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础0.0h，附加0.0h；客户流水¥0.00，充值归因¥0.00；头部客户(12月)Top3：无。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_欣怡.csv

@@ -0,0 +1,52 @@
+助教详情：欣怡（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础29.8h，附加0.0h；客户流水¥9981.98，充值归因¥7000.00；头部客户(12月)Top3：老宋、张先生、轩哥。
+
+一、基础课业绩
+说明：评价：基础29.8h，附加0.0h；客户流水¥9981.98，充值归因¥7000.00；头部客户(12月)Top3：老宋、张先生、轩哥。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,19.50h,15,-24.91h,-17.03h
+11月,10.33h,24,-44.14h,-28.55h
+12月,0.00h,,0.00h,0.00h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,0.00h,,0.00h,0.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,4824.69,17,-24454.84,-25825.14
+11月,5157.29,24,-39097.10,-28169.03
+12月,0.00,,0.00,0.00
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,3000.00,13,-11466.67,-10000.00
+11月,4000.00,7,-3769.23,-5000.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,老宋,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,0.5h / 0.0h,75.99,0.00
+2,张先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,2.4h / 0.0h,570.76,0.00
+3,轩哥,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,1.4h / 0.0h,408.24,0.00
+4,羊,0.0h / 0.0h,0.00,0.00,2.0h / 0.0h,364.70,0.00,0.0h / 0.0h,0.00,0.00
+5,小熊,0.0h / 0.0h,0.00,0.00,7.8h / 0.0h,2438.20,4000.00,0.0h / 0.0h,0.00,0.00
+6,陶,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,4.9h / 0.0h,904.50,0.00
+7,T,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,10.3h / 0.0h,2865.20,3000.00
+8,胡先生,0.0h / 0.0h,0.00,0.00,0.4h / 0.0h,2354.39,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_欣怡.md

@@ -0,0 +1,196 @@
+# 助教详情：欣怡（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础29.8h，附加0.0h；客户流水¥9981.98，充值归因¥7000.00；头部客户(12月)Top3：老宋、张先生、轩哥。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_泡芙.csv

@@ -0,0 +1,48 @@
+助教详情：泡芙（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础21.4h，附加3.0h；客户流水¥8323.03，充值归因¥0.00；头部客户(12月)Top3：夏、艾宇民、羊。
+
+一、基础课业绩
+说明：评价：基础21.4h，附加3.0h；客户流水¥8323.03，充值归因¥0.00；头部客户(12月)Top3：夏、艾宇民、羊。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,21.38h,21,-33.09h,-17.50h
+12月,0.00h,,0.00h,0.00h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,3.00h,5,-5.80h,-4.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,8323.03,23,-35931.36,-25003.29
+12月,0.00,,0.00,0.00
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,夏,0.0h / 0.0h,0.00,0.00,2.4h / 0.0h,501.87,0.00,0.0h / 0.0h,0.00,0.00
+2,艾宇民,0.0h / 0.0h,0.00,0.00,17.5h / 3.0h,5037.44,0.00,0.0h / 0.0h,0.00,0.00
+3,羊,0.0h / 0.0h,0.00,0.00,0.8h / 0.0h,429.33,0.00,0.0h / 0.0h,0.00,0.00
+4,胡先生,0.0h / 0.0h,0.00,0.00,0.8h / 0.0h,2354.39,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_泡芙.md

@@ -0,0 +1,196 @@
+# 助教详情：泡芙（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础21.4h，附加3.0h；客户流水¥8323.03，充值归因¥0.00；头部客户(12月)Top3：夏、艾宇民、羊。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_涛涛.csv

@@ -0,0 +1,60 @@
+助教详情：涛涛（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础168.6h，附加3.0h；客户流水¥166962.41，充值归因¥26000.00；头部客户(12月)Top3：蔡总、轩哥、葛先生。
+
+一、基础课业绩
+说明：评价：基础168.6h，附加3.0h；客户流水¥166962.41，充值归因¥26000.00；头部客户(12月)Top3：蔡总、轩哥、葛先生。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,50.13h,8,5.72h,13.61h
+11月,74.40h,7,19.92h,35.52h
+12月,44.08h,10,-0.73h,2.99h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,3.00h,5,-5.80h,-4.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,35940.84,8,6661.31,5291.01
+11月,88677.55,4,44423.16,55351.23
+12月,42344.02,4,21613.49,28703.38
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,21000.00,5,6533.33,8000.00
+11月,5000.00,6,-2769.23,-4000.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,蔡总,15.1h / 0.0h,18914.63,0.00,28.4h / 0.0h,57785.21,0.00,0.0h / 0.0h,0.00,0.00
+2,轩哥,10.8h / 0.0h,14490.25,0.00,21.8h / 0.0h,22186.78,0.00,40.1h / 0.0h,30415.18,15000.00
+3,葛先生,5.8h / 0.0h,5551.79,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+4,君姐,2.2h / 0.0h,1414.23,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+5,候,3.1h / 0.0h,563.70,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+6,阿亮,2.9h / 0.0h,538.07,0.00,1.7h / 3.0h,300.53,0.00,0.0h / 0.0h,0.00,0.00
+7,罗先生,2.3h / 0.0h,524.41,0.00,2.3h / 0.0h,844.76,0.00,6.0h / 0.0h,1773.80,3000.00
+8,张先生,1.9h / 0.0h,346.94,0.00,1.8h / 0.0h,325.27,0.00,0.0h / 0.0h,0.00,0.00
+9,江先生,0.0h / 0.0h,0.00,0.00,1.8h / 0.0h,2095.41,5000.00,0.0h / 0.0h,0.00,0.00
+10,叶先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,3.1h / 0.0h,3607.02,3000.00
+11,吴生,0.0h / 0.0h,0.00,0.00,2.2h / 0.0h,537.96,0.00,0.0h / 0.0h,0.00,0.00
+12,陈淑涛,0.0h / 0.0h,0.00,0.00,4.4h / 0.0h,1176.77,0.00,0.0h / 0.0h,0.00,0.00
+13,李先生,0.0h / 0.0h,0.00,0.00,2.0h / 0.0h,779.48,0.00,0.0h / 0.0h,0.00,0.00
+14,明哥,0.0h / 0.0h,0.00,0.00,5.4h / 0.0h,2258.14,0.00,0.0h / 0.0h,0.00,0.00
+15,冯先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,0.9h / 0.0h,144.84,0.00
+16,胡先生,0.0h / 0.0h,0.00,0.00,2.7h / 0.0h,387.24,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_涛涛.md

@@ -0,0 +1,196 @@
+# 助教详情：涛涛（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础168.6h，附加3.0h；客户流水¥166962.41，充值归因¥26000.00；头部客户(12月)Top3：蔡总、轩哥、葛先生。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_球球.csv

@@ -0,0 +1,74 @@
+助教详情：球球（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础172.5h，附加41.0h；客户流水¥95647.82，充值归因¥24000.00；头部客户(12月)Top3：葛先生、周周、T。
+
+一、基础课业绩
+说明：评价：基础172.5h，附加41.0h；客户流水¥95647.82，充值归因¥24000.00；头部客户(12月)Top3：葛先生、周周、T。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,57.45h,6,13.04h,20.92h
+11月,66.50h,11,12.02h,27.62h
+12月,48.58h,9,3.77h,7.49h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,11.00h,1,4.43h,3.00h
+11月,25.00h,1,16.20h,18.00h
+12月,5.00h,6,-5.27h,-3.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,33923.75,10,4644.22,3273.92
+11月,41907.39,10,-2347.00,8581.07
+12月,19816.68,8,-913.85,6176.04
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,15000.00,8,533.33,2000.00
+11月,9000.00,4,1230.77,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,葛先生,1.1h / 0.0h,5551.79,0.00,0.9h / 0.0h,3052.28,0.00,0.0h / 0.0h,0.00,0.00
+2,周周,15.2h / 0.0h,4161.75,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+3,T,10.7h / 0.0h,3327.26,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+4,蔡总,8.5h / 0.0h,2130.39,0.00,3.4h / 0.0h,6196.43,0.00,0.0h / 0.0h,0.00,0.00
+5,罗先生,2.3h / 0.0h,1584.22,0.00,0.0h / 0.0h,0.00,0.00,1.4h / 0.0h,514.68,0.00
+6,候,5.3h / 0.0h,926.59,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+7,张先生,3.0h / 0.0h,876.46,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+8,李先生,0.0h / 0.0h,703.83,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+9,大G,1.8h / 0.0h,467.10,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+10,黄先生,0.6h / 5.0h,87.29,0.00,4.4h / 18.0h,828.62,0.00,11.0h / 9.0h,1608.13,0.00
+11,邓飛,0.0h / 0.0h,0.00,0.00,0.8h / 0.0h,925.47,1000.00,0.0h / 0.0h,0.00,0.00
+12,罗超杰,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,0.8h / 0.0h,176.14,0.00
+13,陈腾鑫,0.0h / 0.0h,0.00,0.00,25.1h / 0.0h,5150.83,0.00,6.1h / 2.0h,980.14,0.00
+14,叶先生,0.0h / 0.0h,0.00,0.00,6.7h / 0.0h,3013.86,0.00,0.0h / 0.0h,0.00,0.00
+15,歌神,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,2.6h / 0.0h,1929.19,0.00
+16,羊,0.0h / 0.0h,0.00,0.00,2.2h / 0.0h,383.40,0.00,3.1h / 0.0h,3620.75,0.00
+17,夏,0.0h / 0.0h,0.00,0.00,2.0h / 4.0h,370.12,0.00,2.5h / 0.0h,455.12,0.00
+18,轩哥,0.0h / 0.0h,0.00,0.00,11.2h / 0.0h,12063.93,5000.00,9.1h / 0.0h,9207.48,6000.00
+19,罗超,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,2.9h / 0.0h,2385.30,3000.00
+20,小熊,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,4.4h / 0.0h,709.18,0.00
+21,陈德韩,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,1.8h / 0.0h,5714.01,0.00
+22,刘哥,0.0h / 0.0h,0.00,0.00,2.0h / 0.0h,2982.34,0.00,0.0h / 0.0h,0.00,0.00
+23,桂先生,0.0h / 0.0h,0.00,0.00,2.6h / 0.0h,826.02,0.00,0.0h / 0.0h,0.00,0.00
+24,万先生,0.0h / 0.0h,0.00,0.00,0.0h / 3.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+25,吕先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,2.8h / 0.0h,2843.29,3000.00
+26,胡先生,0.0h / 0.0h,0.00,0.00,3.9h / 0.0h,5688.29,3000.00,0.0h / 0.0h,0.00,0.00
+27,江先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,2.7h / 0.0h,2740.35,3000.00
+28,小宇,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,1.5h / 0.0h,318.98,0.00
+29,陶,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,4.7h / 0.0h,721.01,0.00
+30,小燕,0.0h / 0.0h,0.00,0.00,1.4h / 0.0h,425.80,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_球球.md

@@ -0,0 +1,196 @@
+# 助教详情：球球（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础172.5h，附加41.0h；客户流水¥95647.82，充值归因¥24000.00；头部客户(12月)Top3：葛先生、周周、T。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_瑶瑶.csv

@@ -0,0 +1,49 @@
+助教详情：瑶瑶（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础53.7h，附加3.0h；客户流水¥72216.91，充值归因¥0.00；头部客户(12月)Top3：蔡总、轩哥、陈世。
+
+一、基础课业绩
+说明：评价：基础53.7h，附加3.0h；客户流水¥72216.91，充值归因¥0.00；头部客户(12月)Top3：蔡总、轩哥、陈世。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,34.25h,18,-20.23h,-4.63h
+12月,19.48h,15,-25.33h,-21.61h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,3.00h,5,-5.80h,-4.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,65924.36,8,21669.97,32598.04
+12月,6292.55,17,-14437.98,-7348.10
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,蔡总,19.5h / 0.0h,6292.55,0.00,20.3h / 2.0h,53735.58,0.00,0.0h / 0.0h,0.00,0.00
+2,轩哥,0.0h / 0.0h,0.00,0.00,9.3h / 1.0h,8137.22,0.00,0.0h / 0.0h,0.00,0.00
+3,陈世,0.0h / 0.0h,0.00,0.00,0.8h / 0.0h,139.15,0.00,0.0h / 0.0h,0.00,0.00
+4,游,0.0h / 0.0h,0.00,0.00,2.0h / 0.0h,3544.42,0.00,0.0h / 0.0h,0.00,0.00
+5,林先生,0.0h / 0.0h,0.00,0.00,1.9h / 0.0h,367.99,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_瑶瑶.md

@@ -0,0 +1,196 @@
+# 助教详情：瑶瑶（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础53.7h，附加3.0h；客户流水¥72216.91，充值归因¥0.00；头部客户(12月)Top3：蔡总、轩哥、陈世。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_璇子.csv

@@ -0,0 +1,56 @@
+助教详情：璇子（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础358.9h，附加32.0h；客户流水¥301070.23，充值归因¥44700.00；头部客户(12月)Top3：轩哥、蔡总、江先生。
+
+一、基础课业绩
+说明：评价：基础358.9h，附加32.0h；客户流水¥301070.23，充值归因¥44700.00；头部客户(12月)Top3：轩哥、蔡总、江先生。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,120.20h,2,75.79h,83.67h
+11月,147.92h,2,93.44h,109.03h
+12月,90.75h,3,45.94h,49.66h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,8.00h,3,1.43h,0.00h
+11月,10.00h,3,1.20h,3.00h
+12月,14.00h,4,3.73h,6.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,80804.14,1,51524.61,50154.31
+11月,154486.83,1,110232.44,121160.51
+12月,65779.26,1,45048.73,52138.62
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,34700.00,1,20233.33,21700.00
+11月,10000.00,3,2230.77,1000.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,轩哥,37.7h / 0.0h,28122.15,0.00,50.3h / 0.0h,46514.85,0.00,64.6h / 0.0h,53866.23,23000.00
+2,蔡总,18.3h / 0.0h,21216.97,0.00,49.9h / 0.0h,84757.28,5000.00,0.0h / 0.0h,0.00,0.00
+3,江先生,17.7h / 14.0h,10018.73,0.00,29.6h / 10.0h,14700.83,5000.00,15.2h / 8.0h,5637.50,3000.00
+4,林先生,9.6h / 0.0h,3351.61,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+5,罗先生,5.2h / 0.0h,1655.57,0.00,5.1h / 0.0h,1718.68,0.00,2.8h / 0.0h,1087.21,0.00
+6,君姐,2.2h / 0.0h,1414.23,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+7,游,0.0h / 0.0h,0.00,0.00,3.1h / 0.0h,3544.42,0.00,0.0h / 0.0h,0.00,0.00
+8,羊,0.0h / 0.0h,0.00,0.00,5.0h / 0.0h,1017.12,0.00,0.0h / 0.0h,0.00,0.00
+9,张先生,0.0h / 0.0h,0.00,0.00,4.9h / 0.0h,2233.65,0.00,5.4h / 0.0h,3211.91,0.00
+10,夏,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,7.6h / 0.0h,6452.71,0.00
+11,罗超,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,2.6h / 0.0h,2385.30,3000.00
+12,叶总,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,22.0h / 0.0h,8163.28,5700.00

etl_billiards/docs/table_2025-12-19/助教详情_璇子.md

@@ -0,0 +1,196 @@
+# 助教详情：璇子（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础358.9h，附加32.0h；客户流水¥301070.23，充值归因¥44700.00；头部客户(12月)Top3：轩哥、蔡总、江先生。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_素素.csv

@@ -0,0 +1,64 @@
+助教详情：素素（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础85.5h，附加10.0h；客户流水¥55755.41，充值归因¥7000.00；头部客户(12月)Top3：叶先生、周先生、轩哥。
+
+一、基础课业绩
+说明：评价：基础85.5h，附加10.0h；客户流水¥55755.41，充值归因¥7000.00；头部客户(12月)Top3：叶先生、周先生、轩哥。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,40.45h,10,-3.96h,3.92h
+11月,35.03h,17,-19.44h,-3.85h
+12月,9.98h,19,-34.83h,-31.11h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,10.00h,3,1.20h,3.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,34135.89,9,4856.36,3486.06
+11月,18707.30,19,-25547.09,-14619.02
+12月,2912.22,19,-17818.31,-10728.42
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,6000.00,11,-8466.67,-7000.00
+11月,1000.00,9,-6769.23,-8000.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,叶先生,3.0h / 0.0h,1558.40,0.00,7.0h / 3.0h,3725.65,0.00,15.1h / 0.0h,8807.20,3000.00
+2,周先生,6.2h / 0.0h,1065.70,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+3,轩哥,0.8h / 0.0h,288.12,0.00,0.0h / 0.0h,0.00,0.00,3.6h / 0.0h,8972.50,0.00
+4,邓飛,0.0h / 0.0h,0.00,0.00,0.8h / 0.0h,925.47,1000.00,0.0h / 0.0h,0.00,0.00
+5,罗超杰,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,1.9h / 0.0h,396.94,0.00
+6,陈腾鑫,0.0h / 0.0h,0.00,0.00,0.4h / 0.0h,152.18,0.00,2.1h / 0.0h,542.06,0.00
+7,谢俊,0.0h / 0.0h,0.00,0.00,2.4h / 0.0h,417.75,0.00,0.0h / 0.0h,0.00,0.00
+8,罗先生,0.0h / 0.0h,0.00,0.00,4.7h / 0.0h,861.57,0.00,0.0h / 0.0h,0.00,0.00
+9,葛先生,0.0h / 0.0h,0.00,0.00,9.1h / 0.0h,4668.26,0.00,0.0h / 0.0h,0.00,0.00
+10,歌神,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,3.9h / 0.0h,2789.21,0.00
+11,羊,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,4.3h / 0.0h,3620.75,0.00
+12,张先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,1.1h / 0.0h,449.93,0.00
+13,都先生,0.0h / 0.0h,0.00,0.00,1.6h / 0.0h,269.64,0.00,0.0h / 0.0h,0.00,0.00
+14,夏,0.0h / 0.0h,0.00,0.00,0.8h / 4.0h,2991.13,0.00,0.0h / 0.0h,0.00,0.00
+15,罗先生,0.0h / 0.0h,0.00,0.00,4.3h / 0.0h,1856.86,0.00,0.0h / 0.0h,0.00,0.00
+16,陈德韩,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,5.5h / 0.0h,5714.01,0.00
+17,吕先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,2.8h / 0.0h,2843.29,3000.00
+18,万先生,0.0h / 0.0h,0.00,0.00,0.0h / 3.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+19,阿亮,0.0h / 0.0h,0.00,0.00,3.1h / 0.0h,484.40,0.00,0.0h / 0.0h,0.00,0.00
+20,胡先生,0.0h / 0.0h,0.00,0.00,0.8h / 0.0h,2354.39,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_素素.md

@@ -0,0 +1,196 @@
+# 助教详情：素素（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础85.5h，附加10.0h；客户流水¥55755.41，充值归因¥7000.00；头部客户(12月)Top3：叶先生、周先生、轩哥。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_苏苏.csv

@@ -0,0 +1,58 @@
+助教详情：苏苏（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础82.6h，附加9.0h；客户流水¥33952.79，充值归因¥6000.00；头部客户(12月)Top3：罗先生、T、林先生。
+
+一、基础课业绩
+说明：评价：基础82.6h，附加9.0h；客户流水¥33952.79，充值归因¥6000.00；头部客户(12月)Top3：罗先生、T、林先生。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,25.15h,12,-19.26h,-11.38h
+11月,13.52h,23,-40.96h,-25.37h
+12月,43.90h,11,-0.91h,2.81h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,8.00h,3,1.43h,0.00h
+11月,0.00h,,0.00h,0.00h
+12月,1.00h,9,-9.27h,-7.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,11236.84,15,-18042.69,-19412.99
+11月,10254.59,22,-33999.80,-23071.73
+12月,12461.36,14,-8269.17,-1179.28
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,6000.00,11,-8466.67,-7000.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,罗先生,19.6h / 0.0h,6061.87,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+2,T,4.5h / 0.0h,1429.89,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+3,林先生,4.9h / 0.0h,1369.86,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+4,黄先生,8.1h / 1.0h,1357.90,0.00,4.0h / 0.0h,674.36,0.00,11.2h / 4.0h,1960.96,0.00
+5,张先生,3.9h / 0.0h,1163.40,0.00,0.0h / 0.0h,0.00,0.00,6.4h / 0.0h,3604.02,0.00
+6,葛先生,2.5h / 0.0h,1004.77,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+7,候,0.4h / 0.0h,73.67,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+8,罗超,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,2.6h / 0.0h,2385.30,3000.00
+9,罗超杰,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,2.1h / 0.0h,443.27,0.00
+10,蔡总,0.0h / 0.0h,0.00,0.00,4.7h / 0.0h,6557.92,0.00,0.0h / 0.0h,0.00,0.00
+11,吕先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,2.8h / 0.0h,2843.29,3000.00
+12,林先生,0.0h / 0.0h,0.00,0.00,3.1h / 0.0h,2398.29,0.00,0.0h / 0.0h,0.00,0.00
+13,邓飛,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 4.0h,0.00,0.00
+14,昌哥,0.0h / 0.0h,0.00,0.00,1.6h / 0.0h,624.02,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_苏苏.md

@@ -0,0 +1,196 @@
+# 助教详情：苏苏（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础82.6h，附加9.0h；客户流水¥33952.79，充值归因¥6000.00；头部客户(12月)Top3：罗先生、T、林先生。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_西子.csv

@@ -0,0 +1,45 @@
+助教详情：西子（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础1.8h，附加0.0h；客户流水¥303.51，充值归因¥0.00；头部客户(12月)Top3：张先生。
+
+一、基础课业绩
+说明：评价：基础1.8h，附加0.0h；客户流水¥303.51，充值归因¥0.00；头部客户(12月)Top3：张先生。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,1.82h,26,-52.66h,-37.07h
+12月,0.00h,,0.00h,0.00h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,0.00h,,0.00h,0.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,303.51,26,-43950.88,-33022.81
+12月,0.00,,0.00,0.00
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,张先生,0.0h / 0.0h,0.00,0.00,1.8h / 0.0h,303.51,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_西子.md

@@ -0,0 +1,196 @@
+# 助教详情：西子（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础1.8h，附加0.0h；客户流水¥303.51，充值归因¥0.00；头部客户(12月)Top3：张先生。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_阿清.csv

@@ -0,0 +1,55 @@
+助教详情：阿清（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础139.9h，附加0.0h；客户流水¥53805.04，充值归因¥0.00；头部客户(12月)Top3：陈腾鑫、葛先生、梅。
+
+一、基础课业绩
+说明：评价：基础139.9h，附加0.0h；客户流水¥53805.04，充值归因¥0.00；头部客户(12月)Top3：陈腾鑫、葛先生、梅。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,73.48h,8,19.01h,34.60h
+12月,66.45h,5,21.64h,25.36h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,0.00h,,0.00h,0.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,37302.04,13,-6952.35,3975.72
+12月,16503.00,10,-4227.53,2862.36
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,陈腾鑫,34.9h / 0.0h,7428.12,0.00,44.4h / 0.0h,10029.58,0.00,0.0h / 0.0h,0.00,0.00
+2,葛先生,7.9h / 0.0h,2831.40,0.00,19.7h / 0.0h,7141.04,0.00,0.0h / 0.0h,0.00,0.00
+3,梅,7.5h / 0.0h,2086.99,0.00,3.3h / 0.0h,1573.10,0.00,0.0h / 0.0h,0.00,0.00
+4,张先生,7.2h / 0.0h,1794.40,0.00,1.0h / 0.0h,679.31,0.00,0.0h / 0.0h,0.00,0.00
+5,清,3.0h / 0.0h,1128.06,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+6,候,4.2h / 0.0h,773.51,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+7,常总,1.7h / 0.0h,460.52,0.00,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00
+8,轩哥,0.0h / 0.0h,0.00,0.00,1.0h / 0.0h,10078.85,0.00,0.0h / 0.0h,0.00,0.00
+9,黄先生,0.0h / 0.0h,0.00,0.00,0.7h / 0.0h,382.08,0.00,0.0h / 0.0h,0.00,0.00
+10,蔡总,0.0h / 0.0h,0.00,0.00,1.5h / 0.0h,6557.92,0.00,0.0h / 0.0h,0.00,0.00
+11,小燕,0.0h / 0.0h,0.00,0.00,2.0h / 0.0h,860.16,0.00,0.0h / 0.0h,0.00,0.00

etl_billiards/docs/table_2025-12-19/助教详情_阿清.md

@@ -0,0 +1,196 @@
+# 助教详情：阿清（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础139.9h，附加0.0h；客户流水¥53805.04，充值归因¥0.00；头部客户(12月)Top3：陈腾鑫、葛先生、梅。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/助教详情_饭团.csv

@@ -0,0 +1,46 @@
+助教详情：饭团（2025年10-12月）
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。评价：基础16.0h，附加0.0h；客户流水¥7955.28，充值归因¥0.00；头部客户(12月)Top3：轩哥、张先生。
+
+一、基础课业绩
+说明：评价：基础16.0h，附加0.0h；客户流水¥7955.28，充值归因¥0.00；头部客户(12月)Top3：轩哥、张先生。
+
+月份,基础课业绩,基础课业绩,基础课业绩,基础课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,16.00h,16,-28.41h,-20.53h
+11月,0.00h,,0.00h,0.00h
+12月,0.00h,,0.00h,0.00h
+
+二、附加课业绩
+说明：附加课=order_assistant_type=2。
+
+月份,附加课业绩,附加课业绩,附加课业绩,附加课业绩
+月份,小时数,排名,平均值差值小时数,中位数值差值小时数
+10月,0.00h,,0.00h,0.00h
+11月,0.00h,,0.00h,0.00h
+12月,0.00h,,0.00h,0.00h
+
+三、客户消费业绩
+说明：订单台费+助教+商品应付金额全额计入订单内助教。
+
+月份,客户消费业绩,客户消费业绩,客户消费业绩,客户消费业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,7955.28,16,-21324.25,-22694.55
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+四、客户充值业绩
+说明：充值命中消费窗口±30分钟且有助教则归因；全额复制。
+
+月份,客户充值业绩,客户充值业绩,客户充值业绩,客户充值业绩
+月份,合计元,排名,平均值差值元,中位数值差值元
+10月,0.00,,0.00,0.00
+11月,0.00,,0.00,0.00
+12月,0.00,,0.00,0.00
+
+五、头部客户（按12月消费业绩排序，Top100）
+说明：基础/附加课时=基础h/附加h。
+
+排名,客户名称,12月,12月,12月,11月,11月,11月,10月,10月,10月
+排名,客户名称,基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元),基础/附加课时,消费业绩(元),客户充值(元)
+1,轩哥,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,6.9h / 0.0h,5089.50,0.00
+2,张先生,0.0h / 0.0h,0.00,0.00,0.0h / 0.0h,0.00,0.00,9.1h / 0.0h,2865.78,0.00

etl_billiards/docs/table_2025-12-19/助教详情_饭团.md

@@ -0,0 +1,196 @@
+# 助教详情：饭团（2025年10-12月）
+## 思考过程
+按模板拆分5部分输出；月度排名采用dense_rank；均值/中位数在当月该指标>0助教集合上计算。
+
+## 查询说明
+本表包含5个部分：基础课业绩、附加课业绩、客户消费业绩、客户充值业绩、头部客户情况。均值/中位数差值对比集合为当月该指标>0的助教。充值/客户流水多助教与多订单命中均按全额复制计入，故汇总可能大于门店总额。
+评价：基础16.0h，附加0.0h；客户流水¥7955.28，充值归因¥0.00；头部客户(12月)Top3：轩哥、张先生。
+
+## SQL
+
+### 服务时长（助教-客户-月份）
+```sql
+with raw as (
+  select
+    asl.nickname as assistant,
+    asl.tenant_member_id as member_id,
+    case when asl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when asl.start_use_time >= '2025-11-01 00:00:00+08'::timestamptz and asl.start_use_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz and asl.start_use_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    asl.order_assistant_type,
+    asl.income_seconds
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+    and asl.tenant_member_id is not null and asl.tenant_member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(case when order_assistant_type=1 then income_seconds else 0 end)/3600.0 as base_hours,
+  sum(case when order_assistant_type=2 then income_seconds else 0 end)/3600.0 as extra_hours
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 客户流水（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_amount as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    bo.order_start_time,
+    bo.order_end_time,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_amount a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+raw as (
+  select
+    ao.assistant,
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  join assistant_orders ao on ao.order_settle_id=o.order_settle_id
+  where o.member_id is not null and o.member_id<>0
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(order_amount) as revenue_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```
+
+### 充值归因（助教-客户-月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as table_start_time,
+    max(tfl.ledger_end_time) as table_end_time
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_time as (
+  select
+    asl.order_settle_id,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id=asl.order_settle_id
+  where asl.site_id=%(site_id)s and coalesce(asl.is_delete,0)=0
+  group by asl.order_settle_id
+),
+order_windows as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.table_start_time, coalesce(at.assistant_start_time, bo.table_start_time)) as win_start,
+    greatest(bo.table_end_time, coalesce(at.assistant_end_time, bo.table_end_time)) as win_end
+  from base_orders bo
+  left join assistant_time at on at.order_settle_id=bo.order_settle_id
+  where bo.member_id is not null and bo.member_id<>0
+),
+assistant_orders as (
+  select distinct order_settle_id, nickname as assistant
+  from billiards_dwd.dwd_assistant_service_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz
+    and start_use_time <  %(window_end)s::timestamptz
+),
+recharge_pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id=p.relate_id
+  where p.site_id=%(site_id)s
+    and p.relate_type=5
+    and p.pay_status=2
+    and p.pay_amount>0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+),
+matched as (
+  select
+    rp.pay_time,
+    ow.order_settle_id,
+    ow.member_id,
+    rp.pay_amount
+  from recharge_pay rp
+  join order_windows ow
+    on ow.member_id=rp.member_id
+   and rp.pay_time >= ow.win_start - interval '30 minutes'
+   and rp.pay_time <= ow.win_end   + interval '30 minutes'
+),
+raw as (
+  select
+    ao.assistant,
+    m.member_id,
+    case when m.pay_time >= '2025-10-01 00:00:00+08'::timestamptz and m.pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when m.pay_time >= '2025-11-01 00:00:00+08'::timestamptz and m.pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when m.pay_time >= '2025-12-01 00:00:00+08'::timestamptz and m.pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    m.pay_amount
+  from matched m
+  join assistant_orders ao on ao.order_settle_id=m.order_settle_id
+)
+select
+  assistant,
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_amount
+from raw
+where month_key is not null
+group by assistant, member_id, month_key;
+```

etl_billiards/docs/table_2025-12-19/客户_Top100_2025年10-12月_分表.csv

@@ -0,0 +1,105 @@
+2025年10-12月 客户消费能力Top100（分表）
+与总表同口径；分表仅计算12月喜爱助教Top5，并展示10-12月各月消费/充值。
+
+排名,客户名称,电话号码,12月,12月,12月,11月,11月,10月,10月
+排名,客户名称,电话号码,喜爱助教昵称,消费(元),充值(元),消费(元),充值(元),消费(元),充值(元)
+    1,轩哥,188****7530,七七(65.8h)、璇子(37.7h)、涛涛(10.8h)、Amy(8.8h)、小柔(8.4h),29170.33,0.00,51318.11,13000.00,72647.61,53000.00
+    2,蔡总,159****8893,小柔(30.2h)、七七(26.7h)、瑶瑶(19.5h)、璇子(18.3h)、涛涛(15.1h),34327.09,0.00,115512.37,5000.00,0.00,0.00
+    3,葛先生,138****8071,小燕(104.8h)、阿清(7.9h)、涛涛(5.9h)、梦梦(2.9h)、年糕(2.6h),23237.78,0.00,24991.55,0.00,0.00,0.00
+    4,罗先生,139****6996,佳怡(73.8h)、苏苏(19.6h)、璇子(5.3h)、周周(2.6h)、球球(2.3h),17515.75,0.00,16177.44,0.00,12880.69,10000.00
+    5,陈腾鑫,178****8218,阿清(34.9h)、小侯(12.1h)、千千(4.9h)、小柔(1.7h),10701.83,0.00,19428.27,1000.00,9945.66,8000.00
+    6,江先生,188****4838,璇子(38.7h)、婉婉(4.7h)、小柔(3.2h)、七七(3.2h)、Amy(2.2h),10050.26,0.00,14700.83,5000.00,5721.32,6000.00
+    7,小燕,178****1334,小燕(85.7h)、Amy(1.0h)、年糕(0.5h),17306.66,0.00,12582.23,0.00,0.00,0.00
+    8,张先生,139****8852,小侯(14.5h)、千千(10.0h)、阿清(7.2h)、苏苏(3.9h)、乔西(3.3h),7533.67,0.00,8692.62,0.00,9878.04,0.00
+    9,曾巧明,186****1488,,6331.51,0.00,8821.59,0.00,9677.06,0.00
+    10,黄生,136****9719,,4181.24,0.00,6353.05,0.00,7871.94,0.00
+    11,夏,191****2851,,0.00,0.00,4318.15,6000.00,12281.44,11000.00
+    12,叶先生,138****9539,素素(3.1h)、年糕(2.6h),1658.40,0.00,3725.65,0.00,9107.20,6307.00
+    13,T,180****9962,球球(10.7h)、周周(10.1h)、乔西(6.9h)、苏苏(4.5h)、小侯(3.9h),6777.66,0.00,2490.70,0.00,4382.70,3000.00
+    14,林先生,133****1070,七七(14.0h)、璇子(9.6h)、苏苏(4.9h)、乔西(4.6h)、佳怡(4.5h),9468.05,0.00,3469.81,0.00,0.00,0.00
+    15,曾丹烨,139****3242,,2485.41,0.00,5168.56,0.00,4702.36,3000.00
+    16,胡先生,186****3391,佳怡(3.0h),522.45,0.00,10898.42,8000.00,0.00,0.00
+    17,艾宇民,150****9958,,1179.56,0.00,7185.78,0.00,2397.73,0.00
+    18,谢俊,186****5198,,969.66,0.00,4819.27,0.00,4193.93,0.00
+    19,周周,198****8725,周周(40.8h)、球球(15.2h)、佳怡(13.2h),8905.19,0.00,0.00,0.00,0.00,0.00
+    20,叶总,137****3287,,0.00,0.00,0.00,0.00,8163.28,5700.00
+    21,羊,187****5094,,0.00,0.00,4534.79,16000.00,3620.75,48000.00
+    22,游,172****6666,佳怡(3.2h)、周周(2.4h)、千千(2.2h)、QQ(1.2h),2795.90,0.00,5001.47,0.00,0.00,0.00
+    23,桂先生,166****7275,,606.24,0.00,4788.94,0.00,2283.11,985.00
+    24,明哥,166****0999,婉婉(26.7h)、小柔(25.7h)、年糕(4.6h)、周周(0.4h)、Amy(0.1h),5322.90,0.00,2258.14,0.00,0.00,0.00
+    25,陈德韩,134****7864,,20.00,0.00,0.00,0.00,7282.92,5000.00
+    26,小熊,139****0145,,0.00,0.00,5213.64,4000.00,1922.24,1000.00
+    27,黄先生,135****3507,苏苏(9.6h)、千千(6.2h),2506.04,0.00,1926.28,0.00,2620.40,3000.00
+    28,吕先生,155****0663,,0.00,0.00,0.00,0.00,7012.05,6000.00
+    29,郑先生,159****4331,小敌(4.9h),1614.93,0.00,4823.14,0.00,0.00,0.00
+    30,歌神,188****2164,,0.00,0.00,0.00,0.00,5877.65,5000.00
+    31,罗先生,139****9222,婉婉(7.5h)、小柔(4.5h),0.00,0.00,4687.61,0.00,1143.87,3000.00
+    32,大G,186****4598,周周(14.9h)、佳怡(9.5h)、球球(1.9h),3591.07,0.00,1783.61,0.00,0.00,0.00
+    33,林先生,137****8785,,480.91,0.00,4486.47,0.00,0.00,0.00
+    34,梅,136****4552,千千(10.8h)、阿清(7.5h)、小侯(3.3h),2483.11,0.00,2107.81,0.00,0.00,0.00
+    35,李先生,134****4343,小侯(9.3h)、球球(0.0h),2395.25,0.00,2084.52,0.00,0.00,0.00
+    36,陶,189****2151,,0.00,0.00,0.00,0.00,4096.07,6000.00
+    37,陈先生,159****2829,,0.00,0.00,3717.58,3000.00,100.00,0.00
+    38,候,131****0323,球球(5.3h)、阿清(4.3h)、小侯(3.4h)、涛涛(3.1h)、乔西(2.9h),3765.88,0.00,0.00,0.00,0.00,0.00
+    39,黄先生,158****2109,球球(8.1h)、QQ(4.5h),300.15,0.00,1660.64,0.00,1608.13,3000.00
+    40,孟紫龙,176****3741,,2888.95,0.00,603.78,0.00,0.00,0.00
+    41,罗超,137****0990,,0.00,0.00,0.00,0.00,3398.81,3000.00
+    42,君姐,166****4594,璇子(2.3h)、涛涛(2.2h)、年糕(2.2h),1414.23,0.00,1947.72,0.00,0.00,0.00
+    43,吴生,136****3341,,368.03,0.00,1721.17,0.00,1164.50,0.00
+    44,刘哥,135****0020,婉婉(0.3h),54.46,0.00,2982.34,0.00,0.00,0.00
+    45,李先生,131****4000,,0.00,0.00,2997.53,3000.00,0.00,0.00
+    46,阿亮,159****2628,涛涛(2.9h)、梦梦(2.7h),1133.53,0.00,1802.73,0.00,0.00,0.00
+    47,李先生,186****8308,,0.00,0.00,2785.23,0.00,0.00,0.00
+    48,周先生,173****7775,,0.00,0.00,0.00,0.00,2726.01,0.00
+    49,林先生,159****0021,,0.00,0.00,2690.52,0.00,0.00,0.00
+    50,周先生,193****6822,千千(8.6h)、素素(6.2h),2643.64,0.00,0.00,0.00,0.00,0.00
+    51,林先生,188****0332,,0.00,0.00,2612.01,0.00,0.00,0.00
+    52,牛先生,152****5159,,0.00,0.00,2187.48,0.00,0.00,0.00
+    53,孟紫龙（该会员已注销）,176****37411,,0.00,0.00,300.00,0.00,1831.72,0.00
+    54,罗超杰,137****8012,,0.00,0.00,423.70,0.00,1160.35,0.00
+    55,汪先生,139****6339,,0.00,0.00,0.00,0.00,1390.07,1000.00
+    56,桂先生（该会员已注销）,166****72751,,0.00,0.00,0.00,0.00,1370.81,0.00
+    57,方先生,153****3185,,0.00,0.00,414.57,0.00,898.73,1000.00
+    58,陈淑涛,132****5485,,0.00,0.00,1276.77,0.00,0.00,0.00
+    59,清,130****3087,阿清(3.0h)、千千(3.0h)、小侯(3.0h),1128.06,0.00,0.00,0.00,0.00,0.00
+    60,amy,137****8221,Amy(4.8h),1105.90,0.00,0.00,0.00,0.00,0.00
+    61,曾先生,133****1235,,193.90,0.00,700.09,0.00,97.35,0.00
+    62,昌哥,137****1229,小柔(1.8h),318.40,0.00,624.02,0.00,0.00,0.00
+    63,李,131****9882,,0.00,0.00,320.99,0.00,606.45,0.00
+    64,邓飛,136****9597,,0.00,0.00,925.47,1000.00,0.00,1345.00
+    65,陈世,134****1938,,0.00,0.00,502.07,0.00,419.02,0.00
+    66,婉婉,183****2742,,0.00,0.00,242.81,0.00,624.68,1000.00
+    67,贺斌,150****0885,,0.00,0.00,108.29,0.00,662.27,0.00
+    68,陈泽斌,132****6060,,100.00,0.00,500.00,0.00,100.00,0.00
+    69,林总,138****1180,,0.00,0.00,684.44,0.00,0.00,0.00
+    70,卢广贤,186****6220,,128.86,0.00,288.51,0.00,263.07,0.00
+    71,王龙,186****8011,,669.58,0.00,0.00,0.00,0.00,0.00
+    72,王先生,185****1125,,183.75,0.00,416.93,0.00,0.00,0.00
+    73,潘先生,186****0511,年糕(2.0h),564.93,0.00,0.00,0.00,0.00,0.00
+    74,老宋,138****4554,,0.00,0.00,465.98,0.00,75.99,0.00
+    75,郭先生,156****5001,,0.00,0.00,281.22,0.00,237.13,0.00
+    76,孙启明,137****6325,,0.00,0.00,200.00,0.00,300.00,0.00
+    77,王先生,136****0168,,100.00,0.00,400.00,0.00,0.00,0.00
+    78,黎先生,133****0983,,0.00,0.00,0.00,0.00,470.19,0.00
+    79,常总,185****7188,阿清(1.7h)、年糕(1.7h),460.52,0.00,0.00,0.00,0.00,0.00
+    80,张丹逸,136****6637,,100.00,0.00,0.00,0.00,339.36,0.00
+    81,陈先生,186****8238,,0.00,0.00,416.17,0.00,0.00,0.00
+    82,陈先生,138****3964,,100.00,0.00,0.00,0.00,300.00,0.00
+    83,张先生,136****4528,,298.35,0.00,100.00,0.00,0.00,0.00
+    84,小宇,187****8077,,0.00,0.00,0.00,0.00,397.50,0.00
+    85,黄先生,191****8219,,0.00,0.00,364.33,0.00,0.00,0.00
+    86,刘女士,177****7538,,0.00,0.00,0.00,0.00,362.58,0.00
+    87,魏先生,137****6862,,319.39,0.00,0.00,0.00,0.00,0.00
+    88,杜先生,188****4705,,0.00,0.00,207.47,0.00,100.00,0.00
+    89,刘先生,137****2930,,300.00,0.00,0.00,0.00,0.00,0.00
+    90,陈先生,133****6117,,0.00,0.00,300.00,0.00,0.00,0.00
+    91,潘先生,176****7964,,300.00,0.00,0.00,0.00,0.00,0.00
+    92,钟智豪,188****2803,,0.00,0.00,274.34,0.00,0.00,0.00
+    93,都先生,138****7796,,0.00,0.00,269.64,0.00,0.00,0.00
+    94,林志铭,135****4233,,0.00,0.00,267.96,0.00,0.00,0.00
+    95,方先生,158****6447,,148.00,0.00,100.00,0.00,0.00,0.00
+    96,李先生,176****5124,,0.00,0.00,0.00,0.00,244.00,0.00
+    97,钟先生,132****3438,,0.00,0.00,0.00,0.00,239.37,0.00
+    98,杨,130****5960,,232.00,0.00,0.00,0.00,0.00,0.00
+    99,黄国磊,131****3045,,100.00,0.00,100.00,0.00,25.33,0.00
+    100,周先生,159****9997,,100.00,0.00,100.00,0.00,0.00,0.00

etl_billiards/docs/table_2025-12-19/客户_Top100_2025年10-12月_分表.md

@@ -0,0 +1,199 @@
+# 2025年10-12月 客户消费能力Top100（分表）
+## 思考过程
+以台费订单为基准汇总三类明细，满足应付金额口径，并输出Top100客户的消费/充值/助教偏好。按你的要求，先生成评价为空的版本，再在脚本末尾回填评价。
+
+## 查询说明
+与总表同口径；分表仅计算12月喜爱助教Top5，并展示10-12月各月消费/充值。
+
+## SQL
+
+### Top100（按消费总额）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount,
+    sum(tfl.real_table_use_seconds) as table_use_seconds
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_info as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.order_start_time, coalesce(a.assistant_start_time, bo.order_start_time)) as order_start_time,
+    greatest(bo.order_end_time, coalesce(a.assistant_end_time, bo.order_end_time)) as order_end_time,
+    bo.table_use_seconds,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_info a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+select
+  o.member_id,
+  sum(o.order_amount) as consume_total,
+  count(*) as order_cnt
+from orders o
+where o.member_id is not null and o.member_id <> 0
+group by o.member_id
+order by consume_total desc
+limit 100;
+```
+
+### 按月消费汇总
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount,
+    sum(tfl.real_table_use_seconds) as table_use_seconds
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_info as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.order_start_time, coalesce(a.assistant_start_time, bo.order_start_time)) as order_start_time,
+    greatest(bo.order_end_time, coalesce(a.assistant_end_time, bo.order_end_time)) as order_end_time,
+    bo.table_use_seconds,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_info a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, x as (
+  select
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  where o.member_id is not null and o.member_id <> 0
+)
+select
+  member_id,
+  month_key,
+  sum(order_amount) as consume_sum
+from x
+where month_key is not null
+group by member_id, month_key;
+```
+
+### 按月充值汇总
+```sql
+with pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id = p.relate_id
+  where p.site_id = %(site_id)s
+    and p.relate_type = 5
+    and p.pay_status = 2
+    and p.pay_amount > 0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+)
+, x as (
+  select
+    member_id,
+    case when pay_time >= '2025-10-01 00:00:00+08'::timestamptz and pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when pay_time >= '2025-11-01 00:00:00+08'::timestamptz and pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when pay_time >= '2025-12-01 00:00:00+08'::timestamptz and pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    pay_amount
+  from pay
+)
+select
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_sum
+from x
+where month_key is not null
+group by member_id, month_key;
+```
+
+### 喜爱助教Top5（仅12月）
+```sql
+with x as (
+  select
+    asl.tenant_member_id as member_id,
+    asl.nickname as assistant_nickname,
+    sum(case when asl.order_assistant_type=1 then asl.income_seconds else asl.income_seconds*1.5 end) / 3600.0 as weighted_hours
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0)=0
+    and asl.tenant_member_id is not null and asl.tenant_member_id <> 0
+    and asl.start_use_time >= '2025-12-01 00:00:00+08'::timestamptz
+    and asl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by asl.tenant_member_id, asl.nickname
+),
+ranked as (
+  select *, row_number() over(partition by member_id order by weighted_hours desc) as rn
+  from x
+)
+select
+  member_id,
+  string_agg(assistant_nickname || '(' || to_char(round(weighted_hours::numeric, 1), 'FM999999990.0') || 'h)', '、' order by weighted_hours desc) as fav5
+from ranked
+where rn <= 5
+group by member_id;
+```

etl_billiards/docs/table_2025-12-19/客户_Top100_2025年10-12月_总表.csv

@@ -0,0 +1,105 @@
+2025年10-12月 客户消费能力Top100（总表）
+"消费=台费(dwd_table_fee_log.ledger_amount)+助教(dwd_assistant_service_log.ledger_amount)+商品(dwd_store_goods_sale.ledger_amount)，均为应付金额（不扣优惠），以台费订单为基准串联；充值=充值支付流水(dwd_payment.relate_type=5, pay_status=2, pay_amount>0)按支付时间切月；储值卡未使用金额=当前有效储值卡余额之和(dim_member_card_account.balance, member_card_type_name='储值卡')；喜爱助教=基础课时长+附加课时长*1.5（income_seconds换算小时），总表按10-12月汇总Top5。"
+
+排名,客户名称,电话号码,10月-12月,10月-12月,10月-12月,当前,评价
+排名,客户名称,电话号码,喜爱助教昵称,总消费(元),总充值(元),储值卡未使用金额(元),评价
+1,轩哥,188****7530,璇子(152.6h)、七七(129.6h)、涛涛(72.7h)、小敌(54.3h)、佳怡(50.6h),153136.05,66000.00,6050.80,订单：53单，平均单次¥2889.36；打球：279.0h，平均单次5.3h；偏好：666(54.4%)、发财(13.7%)、包厢(6.5%)、A区(5.9%)；时间：到店均值16:43 中位19:21；离店均值22:31 中位次日01:03；商品：荷花双中支×56(¥3808.00)、细和天下×21(¥2625.00)、100  和成天下×13(¥1495.00)、农夫山泉苏打水×211(¥1266.00)、红牛×101(¥1010.00)、双中支中华×14(¥1008.00)（商品合计¥26812.00 占比17.5%）；综合：10-12月到店消费40天/53次，到店周期中位#1，消费排名#1，在店时长#3；12月到店消费8天/9次，到店周期中位#1，消费排名#1，在店时长#7；最近到店2025-12-18；趋势：10-12月到店频次下降，建议重点唤醒；围客与客户运营建议：重点维护包厢/团建需求，提前锁档与套餐化；商品贡献高，可做常购商品补货提醒与组合促销
+2,蔡总,159****8893,小柔(97.0h)、七七(72.4h)、璇子(68.2h)、涛涛(43.5h)、瑶瑶(42.8h),149839.46,5000.00,2016.18,订单：23单，平均单次¥6514.76；打球：180.4h，平均单次7.8h；偏好：发财(79.2%)、666(6.5%)、A区(4.3%)、补时长(2.8%)；时间：到店均值18:05 中位20:12；离店均值次日02:58 中位次日04:48；商品：钻石荷花×100(¥4685.00)、细和天下×30(¥3750.00)、粗和天下×24(¥3000.00)、100  和成天下×20(¥2300.00)、双中支中华×31(¥2232.00)、荷花双中支×29(¥2008.00)（商品合计¥37968.00 占比25.3%）；综合：10-12月到店消费22天/23次，到店周期中位#2，消费排名#2，在店时长#8；12月到店消费8天/8次，到店周期中位#1，消费排名#2，在店时长#10；最近到店2025-12-15；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：商品贡献高，可做常购商品补货提醒与组合促销
+3,葛先生,138****8071,小燕(158.3h)、阿清(27.6h)、千千(14.4h)、周周(11.0h)、年糕(9.4h),48229.33,0.00,9011.19,订单：42单，平均单次¥1148.32；打球：180.6h，平均单次4.3h；偏好：TV台(43.8%)、A区(21.2%)、包厢(20.6%)、S区/斯诺克(6.7%)；时间：到店均值14:01 中位19:40；离店均值18:19 中位21:54；商品：百威235毫升×72(¥1080.00)、卡士×33(¥726.00)、风花雪月×36(¥576.00)、细荷花×7(¥370.00)、蜂蜜水×36(¥360.00)、东方树叶×39(¥312.00)（商品合计¥5432.00 占比11.3%）；综合：10-12月到店消费25天/42次，到店周期中位#1，消费排名#3，在店时长#7；12月到店消费11天/24次，到店周期中位#1，消费排名#3，在店时长#2；最近到店2025-12-19；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：重点维护包厢/团建需求，提前锁档与套餐化；商品贡献高，可做常购商品补货提醒与组合促销
+4,罗先生,139****6996,佳怡(205.2h)、苏苏(19.6h)、璇子(13.2h)、周周(11.6h)、涛涛(10.5h),46573.88,10000.00,585.35,订单：86单，平均单次¥541.56；打球：212.7h，平均单次2.5h；偏好：TV台(39.2%)、C区(31.8%)、M7(7.8%)、麻将(5.8%)；时间：到店均值15:24 中位18:18；离店均值18:05 中位20:27；商品：百威235毫升×94(¥1410.00)、东方树叶×39(¥312.00)、细荷花×6(¥300.00)、荷花双中支×4(¥276.00)、中支芙蓉王×6(¥228.00)、哇哈哈矿泉水×38(¥190.00)（商品合计¥4390.00 占比9.4%）；综合：10-12月到店消费59天/86次，到店周期中位#1，消费排名#4，在店时长#5；12月到店消费17天/25次，到店周期中位#1，消费排名#4，在店时长#5；最近到店2025-12-18；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：商品贡献高，可做常购商品补货提醒与组合促销
+5,陈腾鑫,178****8218,阿清(79.2h)、佳怡(57.7h)、球球(34.2h)、小侯(19.1h)、千千(7.7h),40075.76,9000.00,0.00,订单：93单，平均单次¥430.92；打球：207.0h，平均单次2.2h；偏好：麻将(25.0%)、C区(18.7%)、TV台(13.6%)、包厢(11.4%)；时间：到店均值17:33 中位20:09；离店均值19:45 中位22:28；商品：百威235毫升×27(¥405.00)、东方树叶×50(¥400.00)、哇哈哈矿泉水×41(¥205.00)、荷花双中支×2(¥144.00)、地道肠×27(¥135.00)、哈啤×12(¥120.00)（商品合计¥2468.00 占比6.2%）；综合：10-12月到店消费49天/93次，到店周期中位#1，消费排名#5，在店时长#6；12月到店消费11天/20次，到店周期中位#1，消费排名#5，在店时长#8；最近到店2025-12-13；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：关注是否为临时充值型，建议引导储值梯度与权益；重点维护包厢/团建需求，提前锁档与套餐化；商品贡献高，可做常购商品补货提醒与组合促销
+6,江先生,188****4838,璇子(110.5h)、婉婉(16.8h)、小柔(12.0h)、Amy(6.7h)、周周(4.3h),30472.41,11000.00,0.00,订单：24单，平均单次¥1269.68；打球：65.6h，平均单次2.7h；偏好：包厢(44.5%)、888(15.3%)、M8(12.8%)、B区(10.3%)；时间：到店均值16:05 中位19:47；离店均值18:53 中位21:24；商品：百威235毫升×480(¥7200.00)、风花雪月×28(¥448.00)、钻石荷花×7(¥330.00)、东方树叶×11(¥88.00)、双中支中华×1(¥72.00)、荷花双中支×1(¥68.00)（商品合计¥8943.00 占比29.3%）；综合：10-12月到店消费21天/24次，到店周期中位#3，消费排名#6，在店时长#14；12月到店消费5天/6次，到店周期中位#3，消费排名#6，在店时长#18；最近到店2025-12-13；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：关注是否为临时充值型，建议引导储值梯度与权益；重点维护包厢/团建需求，提前锁档与套餐化；商品贡献高，可做常购商品补货提醒与组合促销
+7,小燕,178****1334,小燕(140.1h)、阿清(2.0h)、球球(1.4h)、Amy(1.0h)、年糕(0.5h),29888.89,0.00,1802.64,订单：62单，平均单次¥482.08；打球：148.0h，平均单次2.4h；偏好：A区(55.2%)、麻将(20.7%)、S区/斯诺克(14.7%)、补时长(4.3%)；时间：到店均值15:22 中位19:05；离店均值17:50 中位21:43；商品：东方树叶×28(¥224.00)、双中支中华×2(¥144.00)、钻石荷花×2(¥90.00)、荷花双中支×1(¥68.00)、蜂蜜水×6(¥60.00)、百威235毫升×3(¥45.00)（商品合计¥863.00 占比2.9%）；综合：10-12月到店消费27天/62次，到店周期中位#1，消费排名#7，在店时长#10；12月到店消费15天/30次，到店周期中位#1，消费排名#7，在店时长#3；最近到店2025-12-18；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+8,张先生,139****8852,七七(25.1h)、小侯(20.2h)、千千(14.4h)、周周(10.4h)、璇子(10.3h),26104.33,0.00,1942.41,订单：46单，平均单次¥567.49；打球：105.7h，平均单次2.3h；偏好：C区(76.3%)、666(8.8%)、麻将(5.7%)、补时长(4.8%)；时间：到店均值18:43 中位18:47；离店均值21:25 中位21:12；商品：东方树叶×28(¥224.00)、哇哈哈矿泉水×41(¥205.00)、钻石荷花×3(¥135.00)、细和天下×1(¥125.00)、双中支中华×1(¥72.00)、农夫山泉苏打水×11(¥66.00)（商品合计¥1205.00 占比4.6%）；综合：10-12月到店消费40天/46次，到店周期中位#1，消费排名#8，在店时长#13；12月到店消费13天/13次，到店周期中位#1，消费排名#8，在店时长#13；最近到店2025-12-18；趋势：10-12月到店频次下降，建议重点唤醒；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+9,曾巧明,186****1488,,24830.16,0.00,0.00,订单：59单，平均单次¥420.85；打球：404.3h，平均单次6.9h；偏好：B区(66.4%)、C区(32.8%)、A区(0.8%)；时间：到店均值17:01 中位16:13；离店均值23:52 中位次日00:12；商品：跨越贵烟×1(¥30.00)、地道肠×4(¥20.00)、东方树叶×2(¥16.00)、可乐×2(¥10.00)、哇哈哈矿泉水×2(¥10.00)（商品合计¥86.00 占比0.3%）；综合：10-12月到店消费54天/59次，到店周期中位#1，消费排名#9，在店时长#1；12月到店消费15天/16次，到店周期中位#1，消费排名#9，在店时长#1；最近到店2025-12-18；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+10,黄生,136****9719,,18406.23,0.00,0.00,订单：36单，平均单次¥511.28；打球：298.2h，平均单次8.3h；偏好：B区(64.4%)、C区(35.6%)；时间：到店均值13:13 中位12:53；离店均值21:30 中位22:04；商品：地道肠×8(¥40.00)、可乐×2(¥10.00)（商品合计¥50.00 占比0.3%）；综合：10-12月到店消费36天/36次，到店周期中位#1，消费排名#10，在店时长#2；12月到店消费7天/7次，到店周期中位#3，消费排名#10，在店时长#4；最近到店2025-12-18；趋势：10-12月到店频次下降，建议重点唤醒；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+11,夏,191****2851,婉婉(29.7h)、柚子(10.7h)、球球(10.6h)、佳怡(7.8h)、璇子(7.6h),16599.59,17000.00,0.00,订单：10单，平均单次¥1659.96；打球：40.5h，平均单次4.1h；偏好：888(72.1%)、TV台(22.1%)、包厢(5.8%)、补时长(0.0%)；时间：到店均值18:28 中位20:36；离店均值22:43 中位次日01:10；商品：百威235毫升×247(¥3705.00)、细荷花×5(¥250.00)、荷花双中支×3(¥204.00)、清洁费150×1(¥150.00)、三只松鼠开心果×4(¥120.00)、鸡翅三个一份×6(¥108.00)（商品合计¥5162.90 占比31.1%）；综合：10-12月到店消费9天/10次，到店周期中位#5，消费排名#11，在店时长#19；12月到店消费0天/0次，到店周期中位—，消费排名#11，在店时长—；最近到店2025-11-06；趋势：10-12月到店频次下降，建议重点唤醒；围客与客户运营建议：关注是否为临时充值型，建议引导储值梯度与权益；重点维护包厢/团建需求，提前锁档与套餐化；商品贡献高，可做常购商品补货提醒与组合促销
+12,叶先生,138****9539,素素(29.7h)、年糕(9.7h)、Amy(7.4h)、球球(6.7h)、婉婉(3.7h),14491.25,6307.00,0.00,订单：8单，平均单次¥1811.41；打球：27.2h，平均单次3.4h；偏好：B区(30.9%)、888(24.0%)、C区(17.8%)、包厢(16.3%)；时间：到店均值18:38 中位20:59；离店均值次日00:04 中位次日02:12；商品：蓝妹×96(¥1728.00)、钻石荷花×9(¥405.00)、50  和成天下×6(¥390.00)、鸡翅三个一份×14(¥252.00)、100  和成天下×2(¥230.00)、透明袋无穷鸡翅×10(¥200.00)（商品合计¥4880.00 占比33.7%）；综合：10-12月到店消费8天/8次，到店周期中位#2，消费排名#12，在店时长#27；12月到店消费2天/2次，到店周期中位#1，消费排名#12，在店时长#37；最近到店2025-12-06；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：关注是否为临时充值型，建议引导储值梯度与权益；重点维护包厢/团建需求，提前锁档与套餐化；商品贡献高，可做常购商品补货提醒与组合促销
+13,T,180****9962,佳怡(15.0h)、周周(14.7h)、乔西(11.2h)、球球(10.7h)、欣怡(10.3h),13651.06,3000.00,849.68,订单：16单，平均单次¥853.19；打球：55.6h，平均单次3.5h；偏好：麻将(50.8%)、M8(17.0%)、包厢(16.0%)、A区(9.2%)；时间：到店均值18:05 中位19:34；离店均值21:23 中位22:01；商品：荷花双中支×4(¥288.00)、钻石荷花×4(¥190.00)、软荷花×3(¥174.00)、哇哈哈矿泉水×26(¥130.00)、红牛×10(¥100.00)、跨越贵烟×3(¥86.00)（商品合计¥1689.00 占比12.4%）；综合：10-12月到店消费12天/16次，到店周期中位#1，消费排名#13，在店时长#16；12月到店消费5天/8次，到店周期中位#1，消费排名#13，在店时长#14；最近到店2025-12-17；趋势：12月为高峰月，具备加深运营空间；围客与客户运营建议：重点维护包厢/团建需求，提前锁档与套餐化
+14,林先生,133****1070,七七(19.4h)、佳怡(14.9h)、璇子(9.6h)、乔西(9.6h)、苏苏(4.9h),12937.86,0.00,158.79,订单：14单，平均单次¥924.13；打球：44.4h，平均单次3.2h；偏好：麻将(33.6%)、M8(31.5%)、M7(21.5%)、补时长(11.4%)；时间：到店均值12:18 中位17:34；离店均值15:50 中位次日01:15；商品：荷花双中支×5(¥360.00)、细荷花×5(¥260.00)、细和天下×2(¥250.00)、钻石荷花×5(¥250.00)、双中支中华×3(¥216.00)、粗和天下×1(¥125.00)（商品合计¥2246.00 占比17.4%）；综合：10-12月到店消费13天/14次，到店周期中位#1，消费排名#14，在店时长#17；12月到店消费10天/11次，到店周期中位#1，消费排名#14，在店时长#12；最近到店2025-12-13；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导；商品贡献高，可做常购商品补货提醒与组合促销
+15,曾丹烨,139****3242,,12356.33,3000.00,6639.14,订单：54单，平均单次¥228.82；打球：256.7h，平均单次4.8h；偏好：麻将(100.0%)；时间：到店均值16:05 中位17:32；离店均值20:29 中位22:06；商品：合味道泡面×1(¥12.00)、地道肠×2(¥10.00)、东鹏特饮×1(¥7.00)、喜之郎果冻×1(¥5.00)（商品合计¥34.00 占比0.3%）；综合：10-12月到店消费44天/54次，到店周期中位#1，消费排名#15，在店时长#4；12月到店消费9天/11次，到店周期中位#2，消费排名#15，在店时长#6；最近到店2025-12-16；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+16,胡先生,186****3391,佳怡(29.4h)、七七(4.2h)、球球(3.9h)、婉婉(3.9h)、涛涛(2.7h),11420.87,8000.00,0.00,订单：9单，平均单次¥1268.99；打球：39.8h，平均单次4.4h；偏好：包厢(44.2%)、A区(23.1%)、888(11.9%)、麻将(10.7%)；时间：到店均值13:36 中位17:09；离店均值17:14 中位20:36；商品：百威235毫升×70(¥1050.00)、钻石荷花×5(¥225.00)、鸡翅三个一份×4(¥72.00)、蜂蜜水×7(¥70.00)、50  和成天下×1(¥65.00)、软荷花×1(¥58.00)（商品合计¥1900.00 占比16.6%）；综合：10-12月到店消费5天/9次，到店周期中位#6，消费排名#16，在店时长#20；12月到店消费1天/1次，到店周期中位—，消费排名#16，在店时长#41；最近到店2025-12-01；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：关注是否为临时充值型，建议引导储值梯度与权益；重点维护包厢/团建需求，提前锁档与套餐化
+17,艾宇民,150****9958,泡芙(22.0h)、小侯(7.6h),10763.07,0.00,0.00,订单：44单，平均单次¥244.62；打球：112.4h，平均单次2.6h；偏好：B区(82.1%)、C区(12.7%)、包厢(5.2%)；时间：到店均值16:20 中位16:06；离店均值18:50 中位18:03；商品：百威235毫升×30(¥450.00)、地道肠×13(¥65.00)、细荷花×1(¥50.00)、中支芙蓉王×1(¥38.00)、乖媳妇山椒泡爪×1(¥25.00)、哇哈哈矿泉水×5(¥25.00)（商品合计¥808.00 占比7.5%）；综合：10-12月到店消费41天/44次，到店周期中位#1，消费排名#17，在店时长#12；12月到店消费8天/10次，到店周期中位#2，消费排名#17，在店时长#16；最近到店2025-12-17；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：重点维护包厢/团建需求，提前锁档与套餐化
+18,谢俊,186****5198,年糕(4.5h)、素素(2.4h),9982.86,0.00,0.00,订单：46单，平均单次¥217.02；打球：159.5h，平均单次3.5h；偏好：B区(100.0%)；时间：到店均值19:31 中位20:12；离店均值22:59 中位23:48；商品：钻石荷花×1(¥45.00)（商品合计¥45.00 占比0.5%）；综合：10-12月到店消费43天/46次，到店周期中位#1，消费排名#18，在店时长#9；12月到店消费4天/6次，到店周期中位#5，消费排名#18，在店时长#19；最近到店2025-12-17；趋势：10-12月到店频次下降，建议重点唤醒；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+19,周周,198****8725,周周(70.8h)、球球(15.2h)、佳怡(13.2h),8905.19,0.00,0.00,订单：10单，平均单次¥890.52；打球：35.9h，平均单次3.6h；偏好：麻将(59.3%)、M7(19.6%)、补时长(19.6%)、A区(1.5%)；时间：到店均值11:40 中位14:30；离店均值17:04 中位次日00:34；商品：哇哈哈AD钙奶×13(¥104.00)、跨越贵烟×3(¥90.00)、细荷花×1(¥55.00)、红利群×2(¥52.00)、钻石荷花×1(¥50.00)、哇哈哈矿泉水×7(¥35.00)（商品合计¥685.00 占比7.7%）；综合：10-12月到店消费8天/10次，到店周期中位#1，消费排名#19，在店时长#22；12月到店消费8天/10次，到店周期中位#1，消费排名#19，在店时长#11；最近到店2025-12-17；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+20,叶总,137****3287,璇子(22.0h)、小柔(3.8h)、七七(3.0h),8163.28,5700.00,0.00,订单：5单，平均单次¥1632.66；打球：23.4h，平均单次4.7h；偏好：包厢(53.5%)、888(25.5%)、麻将(12.9%)、B区(8.2%)；时间：到店均值21:36 中位21:10；离店均值次日02:18 中位次日02:08；商品：百威235毫升×132(¥1980.00)、荷花双中支×3(¥204.00)、软荷花×1(¥58.00)、跨越贵烟×2(¥56.00)、细荷花×1(¥50.00)、钻石荷花×1(¥45.00)（商品合计¥2487.00 占比30.5%）；综合：10-12月到店消费5天/5次，到店周期中位#2，消费排名#20，在店时长#31；12月到店消费0天/0次，到店周期中位—，消费排名#20，在店时长—；最近到店2025-10-27；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：关注是否为临时充值型，建议引导储值梯度与权益；重点维护包厢/团建需求，提前锁档与套餐化；商品贡献高，可做常购商品补货提醒与组合促销
+21,羊,187****5094,球球(5.3h)、璇子(5.0h)、素素(4.3h)、婉婉(4.3h)、Amy(4.1h),8155.54,64000.00,0.00,订单：13单，平均单次¥627.35；打球：35.7h，平均单次2.7h；偏好：麻将(64.8%)、TV台(14.7%)、B区(10.3%)、888(9.7%)；时间：到店均值18:13 中位19:52；离店均值21:16 中位22:36；商品：蓝妹×54(¥972.00)、钻石荷花×4(¥180.00)、中支芙蓉王×2(¥76.00)、无穷烤小腿×3(¥60.00)、跨越贵烟×2(¥56.00)、鸡翅三个一份×3(¥54.00)（商品合计¥2072.00 占比25.4%）；综合：10-12月到店消费10天/13次，到店周期中位#1，消费排名#21，在店时长#23；12月到店消费0天/0次，到店周期中位—，消费排名#21，在店时长—；最近到店2025-11-23；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：关注是否为临时充值型，建议引导储值梯度与权益；商品贡献高，可做常购商品补货提醒与组合促销
+22,游,172****6666,佳怡(13.5h)、小敌(3.6h)、璇子(3.1h)、周周(2.4h)、千千(2.2h),7797.37,0.00,0.00,订单：13单，平均单次¥599.80；打球：24.1h，平均单次1.9h；偏好：包厢(52.1%)、666(31.7%)、补时长(8.5%)、S区/斯诺克(7.7%)；时间：到店均值19:13 中位18:51；离店均值21:14 中位20:36；商品：双中支中华×3(¥216.00)、红牛×17(¥170.00)、跨越贵烟×5(¥140.00)、荷花双中支×2(¥136.00)、50枸杞槟榔×2(¥130.00)、100  和成天下×1(¥115.00)（商品合计¥1406.00 占比18.0%）；综合：10-12月到店消费9天/13次，到店周期中位#2，消费排名#22，在店时长#29；12月到店消费6天/9次，到店周期中位#1，消费排名#22，在店时长#27；最近到店2025-12-13；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：重点维护包厢/团建需求，提前锁档与套餐化
+23,桂先生,166****7275,球球(2.6h),7678.29,985.00,0.00,订单：21单，平均单次¥365.63；打球：117.4h，平均单次5.6h；偏好：B区(79.3%)、C区(20.7%)；时间：到店均值18:37 中位19:01；离店均值次日00:12 中位次日00:13；商品：地道肠×17(¥85.00)、哇哈哈矿泉水×15(¥75.00)、可乐×13(¥65.00)、红牛×4(¥40.00)、蜂蜜水×3(¥30.00)、东方树叶×3(¥24.00)（商品合计¥373.00 占比4.9%）；综合：10-12月到店消费21天/21次，到店周期中位#2，消费排名#23，在店时长#11；12月到店消费2天/2次，到店周期中位#9，消费排名#23，在店时长#26；最近到店2025-12-16；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：关注是否为临时充值型，建议引导储值梯度与权益
+24,明哥,166****0999,小柔(31.6h)、婉婉(26.7h)、涛涛(5.4h)、年糕(4.6h)、周周(0.4h),7581.04,0.00,954.64,订单：4单，平均单次¥1895.26；打球：21.1h，平均单次5.3h；偏好：包厢(43.0%)、A区(26.0%)、补时长(23.8%)、C区(5.2%)；时间：到店均值10:22 中位10:15；离店均值13:40 中位13:49；商品：百威235毫升×78(¥1170.00)、硬中华×2(¥100.00)、地道肠×9(¥45.00)、东方树叶×4(¥32.00)、鱼蛋×6(¥30.00)、蜂蜜水×2(¥20.00)（商品合计¥1447.00 占比19.1%）；综合：10-12月到店消费4天/4次，到店周期中位#5，消费排名#24，在店时长#35；12月到店消费3天/3次，到店周期中位#5，消费排名#24，在店时长#20；最近到店2025-12-10；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：重点维护包厢/团建需求，提前锁档与套餐化
+25,陈德韩,134****7864,乔西(9.1h)、素素(5.5h)、奈千(4.4h)、周周(4.3h)、小敌(4.0h),7302.92,5000.00,20.11,订单：4单，平均单次¥1825.73；打球：19.6h，平均单次4.9h；偏好：包厢(66.6%)、888(21.3%)、麻将(12.1%)、补时长(0.0%)；时间：到店均值18:53 中位18:48；离店均值21:23 中位21:56；商品：百威235毫升×85(¥1275.00)、卡士×9(¥198.00)、无穷爱辣烤鸡爪×3(¥60.00)、鸡翅三个一份×2(¥36.00)、鱼蛋×6(¥30.00)、鱼豆腐×2(¥30.00)（商品合计¥1898.00 占比26.0%）；综合：10-12月到店消费4天/4次，到店周期中位#15，消费排名#25，在店时长#38；12月到店消费1天/1次，到店周期中位—，消费排名#25，在店时长—；最近到店2025-12-01；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导；重点维护包厢/团建需求，提前锁档与套餐化
+26,小熊,139****0145,佳怡(13.2h)、周周(10.7h)、欣怡(7.9h)、球球(4.4h)、七七(1.3h),7135.88,5000.00,0.00,订单：9单，平均单次¥792.88；打球：32.7h，平均单次3.6h；偏好：麻将(82.3%)、包厢(9.2%)、TV台(8.5%)、补时长(0.0%)；时间：到店均值14:59 中位18:25；离店均值19:25 中位23:36；商品：软荷花×3(¥174.00)、100  和成天下×1(¥115.00)、中支芙蓉王×3(¥114.00)、钻石荷花×2(¥90.00)、炫赫门小南京×3(¥78.00)、荷花双中支×1(¥68.00)（商品合计¥1223.00 占比17.1%）；综合：10-12月到店消费6天/9次，到店周期中位#2，消费排名#26，在店时长#25；12月到店消费0天/0次，到店周期中位—，消费排名#26，在店时长—；最近到店2025-11-05；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：关注是否为临时充值型，建议引导储值梯度与权益；重点维护包厢/团建需求，提前锁档与套餐化
+27,黄先生,135****3507,苏苏(30.8h)、千千(13.1h)、乔西(1.6h)、阿清(0.7h),7052.72,3000.00,1639.43,订单：22单，平均单次¥320.58；打球：39.4h，平均单次1.8h；偏好：B区(80.5%)、A区(17.0%)、C区(2.2%)、补时长(0.4%)；时间：到店均值18:32 中位19:27；离店均值20:30 中位21:28；商品：东方树叶×6(¥48.00)、钻石荷花×1(¥45.00)、哇哈哈AD钙奶×4(¥32.00)、东鹏特饮×3(¥21.00)、无穷烤小腿×1(¥20.00)、综合蔬果干×1(¥15.00)（商品合计¥225.00 占比3.2%）；综合：10-12月到店消费20天/22次，到店周期中位#3，消费排名#27，在店时长#21；12月到店消费7天/7次，到店周期中位#2，消费排名#27，在店时长#22；最近到店2025-12-16；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+28,吕先生,155****0663,小柔(7.5h)、周周(3.7h)、素素(2.8h)、婉婉(2.8h)、苏苏(2.8h),7012.05,6000.00,0.00,订单：4单，平均单次¥1753.01；打球：17.0h，平均单次4.2h；偏好：666(66.5%)、888(18.2%)、麻将(15.3%)；时间：到店均值14:20 中位18:05；离店均值18:34 中位21:28；商品：蓝妹×35(¥630.00)、软中华×3(¥210.00)、红牛×17(¥170.00)、软荷花×2(¥116.00)、鸡翅三个一份×4(¥72.00)、地道肠×10(¥50.00)（商品合计¥1819.00 占比25.9%）；综合：10-12月到店消费4天/4次，到店周期中位#1，消费排名#28，在店时长#41；12月到店消费0天/0次，到店周期中位—，消费排名#28，在店时长—；最近到店2025-10-20；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：关注是否为临时充值型，建议引导储值梯度与权益
+29,郑先生,159****4331,小敌(32.1h),6438.07,0.00,0.00,订单：16单，平均单次¥402.38；打球：42.2h，平均单次2.6h；偏好：C区(63.7%)、补时长(14.4%)、B区(12.9%)、A区(9.0%)；时间：到店均值15:29 中位20:24；离店均值19:22 中位次日00:03；商品：农夫山泉苏打水×8(¥48.00)、蜂蜜水×4(¥40.00)、地道肠×7(¥35.00)、一次性手套×5(¥10.00)、东方树叶×1(¥8.00)、哇哈哈矿泉水×1(¥5.00)（商品合计¥156.00 占比2.4%）；综合：10-12月到店消费13天/16次，到店周期中位#1，消费排名#29，在店时长#18；12月到店消费5天/7次，到店周期中位#1，消费排名#29，在店时长#25；最近到店2025-12-12；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+30,歌神,188****2164,婉婉(6.7h)、希希(4.0h)、素素(3.9h)、小柔(3.8h)、年糕(2.6h),5877.65,5000.00,0.00,订单：4单，平均单次¥1469.41；打球：12.8h，平均单次3.2h；偏好：888(43.0%)、包厢(41.0%)、麻将(16.0%)；时间：到店均值22:20 中位22:39；离店均值次日01:34 中位次日01:19；商品：百威235毫升×90(¥1350.00)、鸡翅三个一份×5(¥90.00)、焦糖瓜子×3(¥45.00)、鱼蛋×9(¥45.00)、三只松鼠开心果×1(¥30.00)、地道肠×6(¥30.00)（商品合计¥1687.00 占比28.7%）；综合：10-12月到店消费4天/4次，到店周期中位#4，消费排名#30，在店时长#46；12月到店消费0天/0次，到店周期中位—，消费排名#30，在店时长—；最近到店2025-10-24；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：关注是否为临时充值型，建议引导储值梯度与权益；重点维护包厢/团建需求，提前锁档与套餐化
+31,罗先生,139****9222,婉婉(18.8h)、年糕(7.7h)、小柔(4.5h)、素素(4.3h)、姜姜(1.5h),5831.48,3000.00,310.31,订单：5单，平均单次¥1166.30；打球：20.0h，平均单次4.0h；偏好：包厢(72.1%)、888(17.8%)、补时长(10.2%)；时间：到店均值16:46 中位21:04；离店均值次日01:10 中位次日00:07；商品：科罗娜啤酒275ml×26(¥468.00)、蓝妹×18(¥324.00)、鸡翅三个一份×2(¥36.00)、三只松鼠开心果×1(¥30.00)、地道肠×5(¥25.00)、焦糖瓜子×1(¥15.00)（商品合计¥924.00 占比15.8%）；综合：10-12月到店消费5天/5次，到店周期中位#14，消费排名#31，在店时长#37；12月到店消费0天/0次，到店周期中位—，消费排名#31，在店时长—；最近到店2025-11-27；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导；重点维护包厢/团建需求，提前锁档与套餐化
+32,大G,186****4598,周周(17.7h)、佳怡(9.5h)、球球(1.9h),5374.68,0.00,0.00,订单：5单，平均单次¥1074.94；打球：21.6h，平均单次4.3h；偏好：麻将(47.3%)、M7(33.2%)、A区(19.4%)、补时长(0.1%)；时间：到店均值17:16 中位20:28；离店均值21:48 中位次日00:04；商品：荷花双中支×2(¥144.00)、哇哈哈AD钙奶×9(¥72.00)、无穷烤小腿×3(¥60.00)、芙蓉王×2(¥56.00)、红利群×2(¥52.00)、蜂蜜水×4(¥40.00)（商品合计¥748.00 占比13.9%）；综合：10-12月到店消费5天/5次，到店周期中位#1，消费排名#32，在店时长#34；12月到店消费4天/4次，到店周期中位#1，消费排名#32，在店时长#24；最近到店2025-12-07；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+33,林先生,137****8785,小侯(9.7h)、梦梦(4.3h)、千千(2.7h)、瑶瑶(1.9h)、周周(0.7h),4967.38,0.00,0.00,订单：7单，平均单次¥709.63；打球：18.7h，平均单次2.7h；偏好：C区(76.5%)、包厢(23.4%)、补时长(0.0%)；时间：到店均值17:58 中位19:48；离店均值20:49 中位23:39；商品：百威235毫升×30(¥450.00)、50枸杞槟榔×1(¥65.00)、东方树叶×5(¥40.00)、蜂蜜水×4(¥40.00)、哇哈哈矿泉水×7(¥35.00)、酒鬼花生×4(¥32.00)（商品合计¥887.00 占比17.9%）；综合：10-12月到店消费6天/7次，到店周期中位#3，消费排名#33，在店时长#40；12月到店消费2天/2次，到店周期中位#4，消费排名#33，在店时长#34；最近到店2025-12-06；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：重点维护包厢/团建需求，提前锁档与套餐化
+34,梅,136****4552,千千(16.6h)、阿清(10.8h)、小侯(4.7h)、小燕(3.8h),4590.92,0.00,2050.00,订单：8单，平均单次¥573.86；打球：20.4h，平均单次2.5h；偏好：麻将(77.0%)、C区(12.1%)、补时长(10.8%)；时间：到店均值14:47 中位20:54；离店均值20:02 中位23:31；商品：硬中华×1(¥50.00)、轻上椰子水×2(¥24.00)、哇哈哈矿泉水×2(¥10.00)（商品合计¥84.00 占比1.8%）；综合：10-12月到店消费7天/8次，到店周期中位#5，消费排名#34，在店时长#36；12月到店消费5天/5次，到店周期中位#5，消费排名#34，在店时长#23；最近到店2025-12-19；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+35,李先生,134****4343,小侯(9.3h)、柚子(7.8h)、球球(0.0h),4479.77,0.00,2433.01,订单：8单，平均单次¥559.97；打球：32.1h，平均单次4.0h；偏好：C区(73.4%)、包厢(26.6%)；时间：到店均值19:34 中位20:19；离店均值23:35 中位23:13；商品：热水可续杯×13(¥39.00)、农夫山泉苏打水×5(¥30.00)、水果脆×1(¥30.00)、跨越贵烟×1(¥30.00)、哇哈哈矿泉水×4(¥20.00)、水溶C×1(¥8.00)（商品合计¥185.00 占比4.1%）；综合：10-12月到店消费8天/8次，到店周期中位#3，消费排名#35，在店时长#26；12月到店消费5天/5次，到店周期中位#3，消费排名#35，在店时长#17；最近到店2025-12-18；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：重点维护包厢/团建需求，提前锁档与套餐化
+36,陶,189****2151,欣怡(4.9h)、球球(4.7h)、周周(4.4h)、七七(4.2h)、佳怡(4.1h),4096.07,6000.00,8.82,订单：7单，平均单次¥585.15；打球：21.7h，平均单次3.1h；偏好：麻将(61.0%)、A区(13.3%)、包厢(9.7%)、C区(8.1%)；时间：到店均值20:49 中位21:49；离店均值23:55 中位次日01:19；商品：荷花双中支×1(¥68.00)、红牛×5(¥50.00)、轻上椰子水×4(¥48.00)、哇哈哈矿泉水×8(¥40.00)、软玉溪×1(¥28.00)、炫赫门小南京×1(¥26.00)（商品合计¥399.00 占比9.7%）；综合：10-12月到店消费6天/7次，到店周期中位#2，消费排名#36，在店时长#33；12月到店消费0天/0次，到店周期中位—，消费排名#36，在店时长—；最近到店2025-10-26；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导；重点维护包厢/团建需求，提前锁档与套餐化
+37,陈先生,159****2829,佳怡(7.6h)、奈千(5.1h)、柚子(4.4h)、乔西(4.2h)、Amy(4.2h),3817.58,3000.00,0.55,订单：3单，平均单次¥1272.53；打球：7.6h，平均单次2.5h；偏好：麻将(54.5%)、TV台(45.4%)、补时长(0.1%)、A区(0.0%)；时间：到店均值13:21 中位19:06；离店均值18:07 中位次日00:40；商品：百威235毫升×10(¥150.00)、钻石荷花×1(¥45.00)、跨越贵烟×1(¥28.00)、无穷爱辣烤鸡爪×1(¥20.00)、白桦树汁×1(¥12.00)、椰汁×1(¥8.00)（商品合计¥277.00 占比7.3%）；综合：10-12月到店消费3天/3次，到店周期中位#19，消费排名#37，在店时长#57；12月到店消费0天/0次，到店周期中位—，消费排名#37，在店时长—；最近到店2025-11-07；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导
+38,候,131****0323,球球(5.3h)、阿清(4.3h)、小侯(3.4h)、涛涛(3.1h)、乔西(2.9h),3765.88,0.00,0.00,订单：8单，平均单次¥470.74；打球：22.0h，平均单次2.7h；偏好：TV台(66.3%)、A区(28.3%)、C区(5.4%)；时间：到店均值20:08 中位19:43；离店均值22:53 中位22:54；商品：冰红茶×8(¥48.00)、地道肠×4(¥20.00)、鱼蛋×4(¥20.00)、可乐×2(¥10.00)、哇米诺豆奶×1(¥10.00)、红牛×1(¥10.00)（商品合计¥161.00 占比4.3%）；综合：10-12月到店消费5天/8次，到店周期中位#1，消费排名#38，在店时长#32；12月到店消费5天/8次，到店周期中位#1，消费排名#38，在店时长#15；最近到店2025-12-17；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+39,黄先生,158****2109,球球(64.0h)、QQ(6.9h),3568.92,3000.00,420.01,订单：10单，平均单次¥356.89；打球：23.4h，平均单次2.3h；偏好：A区(60.1%)、包厢(27.0%)、补时长(12.9%)；时间：到店均值04:58 中位01:43；离店均值06:06 中位03:20；商品：蓝妹×4(¥72.00)、跨越贵烟×1(¥28.00)、火腿肠×2(¥10.00)、东方树叶×1(¥8.00)、卫龙魔芋爽×1(¥8.00)、农夫山泉苏打水×1(¥6.00)（商品合计¥142.00 占比4.0%）；综合：10-12月到店消费8天/10次，到店周期中位#4，消费排名#39，在店时长#30；12月到店消费3天/4次，到店周期中位#8，消费排名#39，在店时长#45；最近到店2025-12-17；趋势：12月为高峰月，具备加深运营空间；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导；重点维护包厢/团建需求，提前锁档与套餐化
+40,孟紫龙,176****3741,,3492.73,0.00,0.00,订单：13单，平均单次¥268.67；打球：57.6h，平均单次4.4h；偏好：B区(100.0%)；时间：到店均值18:44 中位18:50；离店均值23:10 中位23:35；商品：哇哈哈矿泉水×9(¥45.00)、东方树叶×5(¥40.00)、地道肠×7(¥35.00)、可乐×2(¥10.00)、维他柠檬茶×1(¥8.00)、脉动×1(¥8.00)（商品合计¥153.00 占比4.4%）；综合：10-12月到店消费12天/13次，到店周期中位#1，消费排名#40，在店时长#15；12月到店消费10天/11次，到店周期中位#1，消费排名#40，在店时长#9；最近到店2025-12-18；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+41,罗超,137****0990,球球(2.9h)、璇子(2.7h)、苏苏(2.6h)、奈千(2.5h),3398.81,3000.00,101.19,订单：2单，平均单次¥1699.40；打球：7.3h，平均单次3.7h；偏好：888(78.8%)、麻将(21.2%)；时间：到店均值20:57 中位20:57；离店均值次日00:37 中位次日00:37；商品：科罗娜啤酒275ml×49(¥882.00)、荷花双中支×1(¥68.00)、百威235毫升×3(¥45.00)、掼蛋扑克×4(¥32.00)、跨越贵烟×1(¥28.00)、东方树叶×3(¥24.00)（商品合计¥1115.00 占比32.8%）；综合：10-12月到店消费2天/2次，到店周期中位#16，消费排名#41，在店时长#59；12月到店消费0天/0次，到店周期中位—，消费排名#41，在店时长—；最近到店2025-10-24；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导
+42,君姐,166****4594,年糕(8.2h)、婉婉(4.3h)、璇子(2.3h)、涛涛(2.2h),3361.95,0.00,272.15,订单：3单，平均单次¥1120.65；打球：12.5h，平均单次4.2h；偏好：666(30.5%)、M8(25.3%)、M7(17.7%)、包厢(13.9%)；时间：到店均值15:38 中位16:21；离店均值20:33 中位21:15；商品：哇哈哈矿泉水×20(¥100.00)、卡士×2(¥44.00)、东方树叶×5(¥40.00)、麻将房茶位费×1(¥40.00)、小果盘×1(¥37.90)、软玉溪×1(¥28.00)（商品合计¥319.90 占比9.5%）；综合：10-12月到店消费3天/3次，到店周期中位#5，消费排名#42，在店时长#47；12月到店消费1天/1次，到店周期中位—，消费排名#42，在店时长#31；最近到店2025-12-02；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导；重点维护包厢/团建需求，提前锁档与套餐化
+43,吴生,136****3341,小柔(2.8h)、涛涛(2.2h),3253.70,0.00,3680.65,订单：15单，平均单次¥216.91；打球：34.3h，平均单次2.3h；偏好：C区(86.1%)、B区(7.0%)、包厢(6.7%)、补时长(0.1%)；时间：到店均值19:01 中位18:43；离店均值21:14 中位21:25；商品：哇哈哈矿泉水×18(¥90.00)、双中支中华×1(¥72.00)、阿萨姆×1(¥8.00)（商品合计¥170.00 占比5.2%）；综合：10-12月到店消费12天/15次，到店周期中位#6，消费排名#43，在店时长#24；12月到店消费2天/3次，到店周期中位#5，消费排名#43，在店时长#35；最近到店2025-12-16；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：重点维护包厢/团建需求，提前锁档与套餐化
+44,刘哥,135****0020,婉婉(5.3h)、球球(2.1h),3036.80,0.00,371.51,订单：2单，平均单次¥1518.40；打球：5.0h，平均单次2.5h；偏好：包厢(93.4%)、S区/斯诺克(6.6%)、补时长(0.0%)；时间：到店均值11:33 中位11:33；离店均值12:15 中位12:15；商品：百威235毫升×79(¥1185.00)、中支芙蓉王×2(¥76.00)、地道肠×6(¥30.00)、鱼蛋×6(¥30.00)、红利群×1(¥26.00)、红烧牛肉面×2(¥24.00)（商品合计¥1417.00 占比46.7%）；综合：10-12月到店消费2天/2次，到店周期中位#20，消费排名#44，在店时长#66；12月到店消费1天/1次，到店周期中位—，消费排名#44，在店时长#49；最近到店2025-12-17；趋势：12月为高峰月，具备加深运营空间；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导；重点维护包厢/团建需求，提前锁档与套餐化
+45,李先生,131****4000,小敌(11.8h)、涛涛(2.1h)、Amy(1.1h),2997.53,3000.00,563.47,订单：4单，平均单次¥749.38；打球：11.8h，平均单次3.0h；偏好：包厢(100.0%)；时间：到店均值16:24 中位20:21；离店均值19:21 中位23:11；商品：轻上椰子水×10(¥120.00)、科罗娜啤酒275ml×6(¥108.00)、跨越贵烟×1(¥28.00)、地道肠×2(¥10.00)、海之言×1(¥8.00)、茶兀×1(¥8.00)（商品合计¥298.00 占比9.9%）；综合：10-12月到店消费3天/4次，到店周期中位#5，消费排名#45，在店时长#48；12月到店消费0天/0次，到店周期中位—，消费排名#45，在店时长—；最近到店2025-11-07；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：重点维护包厢/团建需求，提前锁档与套餐化
+46,阿亮,159****2628,涛涛(9.1h)、素素(3.1h)、奈千(2.9h)、梦梦(2.7h)、柚子(2.4h),2936.26,0.00,612.33,订单：8单，平均单次¥367.03；打球：16.8h，平均单次2.1h；偏好：B区(65.3%)、S区/斯诺克(34.5%)、补时长(0.2%)；时间：到店均值21:06 中位20:32；离店均值23:26 中位23:29；商品：哇哈哈矿泉水×4(¥20.00)、香飘飘果汁茶×1(¥10.00)、茶兀×1(¥8.00)（商品合计¥38.00 占比1.3%）；综合：10-12月到店消费6天/8次，到店周期中位#1，消费排名#46，在店时长#43；12月到店消费2天/3次，到店周期中位#1，消费排名#46，在店时长#29；最近到店2025-12-04；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+47,李先生,186****8308,小侯(5.6h)、年糕(4.2h),2785.23,0.00,0.00,订单：1单，平均单次¥2785.23；打球：8.3h，平均单次8.3h；偏好：包厢(100.0%)；时间：到店均值23:15 中位23:15；离店均值次日07:32 中位次日07:32；商品：清洁费150×1(¥150.00)、卡士×1(¥22.00)、合味道泡面×1(¥12.00)、轻上椰子水×1(¥12.00)、雪碧×2(¥10.00)、哇哈哈矿泉水×1(¥5.00)（商品合计¥216.00 占比7.8%）；综合：10-12月到店消费1天/1次，到店周期中位—，消费排名#47，在店时长#55；12月到店消费0天/0次，到店周期中位—，消费排名#47，在店时长—；最近到店2025-11-28；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：重点维护包厢/团建需求，提前锁档与套餐化
+48,周先生,173****7775,乔西(7.5h)、小敌(6.6h),2726.01,0.00,0.00,订单：1单，平均单次¥2726.01；打球：7.5h，平均单次7.5h；偏好：666(100.0%)；时间：到店均值15:40 中位15:40；离店均值23:11 中位23:11；商品：哇哈哈矿泉水×1(¥5.00)（商品合计¥5.00 占比0.2%）；综合：10-12月到店消费1天/1次，到店周期中位—，消费排名#48，在店时长#58；12月到店消费0天/0次，到店周期中位—，消费排名#48，在店时长—；最近到店2025-10-25；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+49,林先生,159****0021,婉婉(12.0h)、小柔(3.2h)、苏苏(3.1h),2690.52,0.00,49.48,订单：2单，平均单次¥1345.26；打球：6.0h，平均单次3.0h；偏好：包厢(66.7%)、A区(33.3%)；时间：到店均值10:13 中位10:13；离店均值12:14 中位12:14；商品：百威235毫升×24(¥360.00)、中支芙蓉王×2(¥76.00)、绿茶×4(¥24.00)、无穷爱辣烤鸡爪×1(¥20.00)、透明袋无穷鸡翅×1(¥20.00)、益达×1(¥18.00)（商品合计¥597.00 占比22.2%）；综合：10-12月到店消费2天/2次，到店周期中位#6，消费排名#49，在店时长#62；12月到店消费0天/0次，到店周期中位—，消费排名#49，在店时长—；最近到店2025-11-25；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导；重点维护包厢/团建需求，提前锁档与套餐化
+50,周先生,193****6822,千千(8.6h)、素素(6.2h),2643.64,0.00,2092.01,订单：3单，平均单次¥881.21；打球：14.9h，平均单次5.0h；偏好：C区(100.0%)；时间：到店均值21:06 中位22:21；离店均值次日02:04 中位次日02:07；商品：哇哈哈矿泉水×5(¥25.00)、卡士×1(¥22.00)、东方树叶×2(¥16.00)、王老吉×1(¥8.00)、美汁源果粒橙×1(¥8.00)、一次性拖鞋×1(¥5.00)（商品合计¥89.00 占比3.4%）；综合：10-12月到店消费3天/3次，到店周期中位#4，消费排名#50，在店时长#44；12月到店消费3天/3次，到店周期中位#4，消费排名#50，在店时长#21；最近到店2025-12-13；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+51,林先生,188****0332,周周(4.0h)、小敌(1.9h),2612.01,0.00,0.00,订单：8单，平均单次¥326.50；打球：16.9h，平均单次2.1h；偏好：666(25.1%)、麻将(22.3%)、M7(20.7%)、S区/斯诺克(12.8%)；时间：到店均值14:55 中位17:06；离店均值17:17 中位19:13；商品：钻石荷花×3(¥135.00)、50枸杞槟榔×1(¥65.00)、跨越贵烟×1(¥28.00)、哇哈哈矿泉水×5(¥25.00)、红牛×2(¥20.00)、蜂蜜水×2(¥20.00)（商品合计¥330.00 占比12.6%）；综合：10-12月到店消费6天/8次，到店周期中位#1，消费排名#51，在店时长#42；12月到店消费0天/0次，到店周期中位—，消费排名#51，在店时长—；最近到店2025-11-28；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+52,牛先生,152****5159,柚子(10.4h),2187.48,0.00,0.00,订单：5单，平均单次¥437.50；打球：10.4h，平均单次2.1h；偏好：C区(99.8%)、补时长(0.2%)、A区(0.0%)、B区(0.0%)；时间：到店均值19:00 中位19:37；离店均值21:29 中位20:52；商品：地道肠×6(¥30.00)、东方树叶×2(¥16.00)、美汁源果粒橙×2(¥16.00)（商品合计¥62.00 占比2.8%）；综合：10-12月到店消费3天/5次，到店周期中位#1，消费排名#52，在店时长#51；12月到店消费0天/0次，到店周期中位—，消费排名#52，在店时长—；最近到店2025-11-24；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+53,孟紫龙（该会员已注销）,176****37411,,2131.72,0.00,0.00,订单：7单，平均单次¥304.53；打球：26.3h，平均单次3.8h；偏好：C区(99.9%)、补时长(0.1%)；时间：到店均值19:17 中位19:34；离店均值23:32 中位23:56；商品：可乐×5(¥25.00)、东鹏特饮×2(¥14.00)、鱼蛋×1(¥5.00)、一次性手套×1(¥2.00)（商品合计¥46.00 占比2.2%）；综合：10-12月到店消费7天/7次，到店周期中位#4，消费排名#53，在店时长#28；12月到店消费0天/0次，到店周期中位—，消费排名#53，在店时长—；最近到店2025-11-28；趋势：10-12月到店频次下降，建议重点唤醒；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+54,罗超杰,137****8012,苏苏(2.1h)、素素(1.9h)、周周(1.2h)、球球(0.8h),1584.05,0.00,3037.00,订单：8单，平均单次¥198.01；打球：8.4h，平均单次1.0h；偏好：包厢(74.8%)、B区(14.0%)、A区(10.9%)、补时长(0.2%)；时间：到店均值18:55 中位22:56；离店均值20:06 中位23:23；商品：雪碧×4(¥20.00)、可乐×3(¥15.00)、维他柠檬茶×1(¥8.00)、打火机×1(¥2.00)（商品合计¥45.00 占比2.8%）；综合：10-12月到店消费6天/8次，到店周期中位#3，消费排名#54，在店时长#54；12月到店消费0天/0次，到店周期中位—，消费排名#54，在店时长—；最近到店2025-11-05；趋势：10-12月到店频次下降，建议重点唤醒；围客与客户运营建议：重点维护包厢/团建需求，提前锁档与套餐化
+55,汪先生,139****6339,姜姜(9.6h),1390.07,1000.00,11.99,订单：4单，平均单次¥347.52；打球：5.1h，平均单次1.3h；偏好：包厢(82.3%)、A区(17.4%)、补时长(0.4%)；时间：到店均值21:16 中位22:01；离店均值23:26 中位23:25；商品：轻上椰子水×1(¥12.00)、三得利×1(¥8.00)、巧乐兹伊利×1(¥8.00)（商品合计¥28.00 占比2.0%）；综合：10-12月到店消费3天/4次，到店周期中位#1，消费排名#55，在店时长#65；12月到店消费0天/0次，到店周期中位—，消费排名#55，在店时长—；最近到店2025-10-09；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导；重点维护包厢/团建需求，提前锁档与套餐化
+56,桂先生（该会员已注销）,166****72751,,1370.81,0.00,0.00,订单：5单，平均单次¥274.16；打球：19.0h，平均单次3.8h；偏好：C区(100.0%)；时间：到店均值19:20 中位19:37；离店均值23:08 中位23:57；商品：红牛×4(¥40.00)、东鹏特饮×2(¥14.00)、可乐×2(¥10.00)、地道肠×2(¥10.00)、普通扑克×1(¥5.00)（商品合计¥79.00 占比5.8%）；综合：10-12月到店消费5天/5次，到店周期中位#2，消费排名#56，在店时长#39；12月到店消费0天/0次，到店周期中位—，消费排名#56，在店时长—；最近到店2025-10-16；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+57,方先生,153****3185,乔西(3.2h),1313.30,1000.00,0.00,订单：3单，平均单次¥437.77；打球：10.2h，平均单次3.4h；偏好：麻将(69.0%)、666(31.0%)；时间：到店均值17:32 中位17:19；离店均值21:01 中位20:30；商品：农夫山泉苏打水×5(¥30.00)、掼蛋扑克×2(¥16.00)、可乐×3(¥15.00)、普通扑克×2(¥10.00)、红牛×1(¥10.00)、椰汁×1(¥8.00)（商品合计¥100.00 占比7.6%）；综合：10-12月到店消费3天/3次，到店周期中位#10，消费排名#57，在店时长#52；12月到店消费0天/0次，到店周期中位—，消费排名#57，在店时长—；最近到店2025-11-17；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：关注是否为临时充值型，建议引导储值梯度与权益
+58,陈淑涛,132****5485,涛涛(4.4h)、乔西(3.4h),1276.77,0.00,0.00,订单：3单，平均单次¥425.59；打球：4.4h，平均单次1.5h；偏好：麻将(77.1%)、A区(22.9%)、补时长(0.0%)；时间：到店均值12:52 中位09:40；离店均值14:39 中位10:40；商品：细荷花×1(¥50.00)、轻上椰子水×1(¥12.00)、哇哈哈矿泉水×2(¥10.00)、火鸡面×1(¥10.00)、茶兀×1(¥8.00)、火腿肠×1(¥5.00)（商品合计¥99.00 占比7.8%）；综合：10-12月到店消费2天/3次，到店周期中位#7，消费排名#58，在店时长#69；12月到店消费0天/0次，到店周期中位—，消费排名#58，在店时长—；最近到店2025-11-22；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+59,清,130****3087,小侯(3.0h)、阿清(3.0h)、千千(3.0h),1128.06,0.00,1944.76,订单：1单，平均单次¥1128.06；打球：3.0h，平均单次3.0h；偏好：麻将(100.0%)；时间：到店均值21:18 中位21:18；离店均值次日02:00 中位次日02:00；综合：10-12月到店消费1天/1次，到店周期中位—，消费排名#59，在店时长#73；12月到店消费1天/1次，到店周期中位—，消费排名#59，在店时长#38；最近到店2025-12-08；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+60,amy,137****8221,Amy(4.8h),1105.90,0.00,0.38,订单：1单，平均单次¥1105.90；打球：4.8h，平均单次4.8h；偏好：麻将(100.0%)；时间：到店均值22:47 中位22:47；离店均值次日03:37 中位次日03:37；商品：跨越贵烟×2(¥60.00)、掼蛋扑克×2(¥16.00)、热水可续杯×4(¥12.00)、可乐×2(¥10.00)、蜂蜜水×1(¥10.00)（商品合计¥108.00 占比9.8%）；综合：10-12月到店消费1天/1次，到店周期中位—，消费排名#60，在店时长#67；12月到店消费1天/1次，到店周期中位—，消费排名#60，在店时长#32；最近到店2025-12-06；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导
+61,曾先生,133****1235,奈千(1.0h),991.34,0.00,303.19,订单：7单，平均单次¥141.62；打球：13.0h，平均单次1.9h；偏好：S区/斯诺克(63.2%)、A区(36.8%)；时间：到店均值13:58 中位12:19；离店均值15:49 中位14:10；商品：咖啡代购×3(¥45.00)、东方树叶×3(¥24.00)、屈臣氏苏打水×1(¥8.00)、阿萨姆×1(¥8.00)、绿茶×1(¥6.00)、哇哈哈矿泉水×1(¥5.00)（商品合计¥98.00 占比9.9%）；综合：10-12月到店消费7天/7次，到店周期中位#6，消费排名#61，在店时长#45；12月到店消费2天/2次，到店周期中位#6，消费排名#61，在店时长#36；最近到店2025-12-10；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导
+62,昌哥,137****1229,Amy(2.0h)、小柔(1.8h)、苏苏(1.7h),942.42,0.00,2471.98,订单：2单，平均单次¥471.21；打球：4.5h，平均单次2.3h；偏好：A区(100.0%)；时间：到店均值20:05 中位20:05；离店均值22:21 中位22:21；商品：农夫山泉苏打水×3(¥18.00)、麻辣王子×1(¥12.00)、哇哈哈矿泉水×2(¥10.00)、营养快线×1(¥8.00)、阿萨姆×1(¥8.00)（商品合计¥56.00 占比5.9%）；综合：10-12月到店消费2天/2次，到店周期中位#15，消费排名#62，在店时长#68；12月到店消费1天/1次，到店周期中位—，消费排名#62，在店时长#44；最近到店2025-12-07；趋势：12月为高峰月，具备加深运营空间；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+63,李,131****9882,,927.44,0.00,0.00,订单：11单，平均单次¥84.31；打球：11.1h，平均单次1.0h；偏好：B区(68.7%)、A区(30.9%)、补时长(0.4%)；时间：到店均值13:13 中位13:14；离店均值14:13 中位14:00；商品：哇哈哈矿泉水×4(¥20.00)（商品合计¥20.00 占比2.2%）；综合：10-12月到店消费6天/11次，到店周期中位#8，消费排名#63，在店时长#49；12月到店消费0天/0次，到店周期中位—，消费排名#63，在店时长—；最近到店2025-11-24；趋势：10-12月到店频次下降，建议重点唤醒；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+64,邓飛,136****9597,婉婉(6.8h)、苏苏(6.0h)、小敌(4.5h)、小柔(1.2h)、球球(0.8h),925.47,2345.00,74.77,订单：1单，平均单次¥925.47；打球：1.9h，平均单次1.9h；偏好：包厢(100.0%)；时间：到店均值01:56 中位01:56；离店均值03:48 中位03:48；商品：百威235毫升×12(¥180.00)、卡士×1(¥22.00)、农夫山泉苏打水×3(¥18.00)、一次性拖鞋×1(¥5.00)、哇哈哈矿泉水×1(¥5.00)（商品合计¥230.00 占比24.9%）；综合：10-12月到店消费1天/1次，到店周期中位—，消费排名#64，在店时长#85；12月到店消费0天/0次，到店周期中位—，消费排名#64，在店时长—；最近到店2025-11-09；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导；重点维护包厢/团建需求，提前锁档与套餐化
+65,陈世,134****1938,瑶瑶(0.8h),921.09,0.00,0.00,订单：6单，平均单次¥153.52；打球：6.5h，平均单次1.1h；偏好：C区(49.7%)、S区/斯诺克(49.6%)、补时长(0.7%)；时间：到店均值20:42 中位21:08；离店均值22:16 中位22:24；综合：10-12月到店消费3天/6次，到店周期中位#24，消费排名#65，在店时长#61；12月到店消费0天/0次，到店周期中位—，消费排名#65，在店时长—；最近到店2025-11-20；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+66,婉婉,183****2742,婉婉(4.3h),867.49,1000.00,514.08,订单：2单，平均单次¥433.74；打球：4.4h，平均单次2.2h；偏好：B区(36.5%)、A区(35.9%)、888(27.6%)；时间：到店均值15:41 中位15:41；离店均值18:01 中位18:01；商品：旺仔牛奶×2(¥20.00)、红牛×1(¥10.00)、王老吉×1(¥8.00)、哇哈哈矿泉水×1(¥5.00)、喜之郎果冻×1(¥5.00)（商品合计¥48.00 占比5.5%）；综合：10-12月到店消费2天/2次，到店周期中位#23，消费排名#66，在店时长#70；12月到店消费0天/0次，到店周期中位—，消费排名#66，在店时长—；最近到店2025-11-21；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+67,贺斌,150****0885,佳怡(1.0h),770.56,0.00,0.00,订单：4单，平均单次¥192.64；打球：9.6h，平均单次2.4h；偏好：A区(50.1%)、C区(49.9%)；时间：到店均值17:30 中位17:55；离店均值19:53 中位19:54；商品：细荷花×1(¥50.00)、可乐×5(¥25.00)、哇哈哈矿泉水×2(¥10.00)、地道肠×2(¥10.00)、东方树叶×1(¥8.00)（商品合计¥103.00 占比13.4%）；综合：10-12月到店消费3天/4次，到店周期中位#4，消费排名#67，在店时长#53；12月到店消费0天/0次，到店周期中位—，消费排名#67，在店时长—；最近到店2025-11-03；趋势：10-12月到店频次下降，建议重点唤醒；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+68,陈泽斌,132****6060,,700.00,0.00,0.00,订单：6单，平均单次¥116.67；打球：0.1h，平均单次0.0h；偏好：补时长(100.0%)；时间：到店均值17:51 中位20:40；离店均值18:50 中位21:39；综合：10-12月到店消费6天/6次，到店周期中位#10，消费排名#68，在店时长#97；12月到店消费1天/1次，到店周期中位—，消费排名#68，在店时长#51；最近到店2025-12-11；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+69,林总,138****1180,周周(2.2h)、婉婉(0.9h),684.44,0.00,16729.21,订单：2单，平均单次¥342.22；打球：3.7h，平均单次1.9h；偏好：包厢(100.0%)；时间：到店均值18:28 中位18:28；离店均值20:20 中位20:20；商品：东方树叶×1(¥8.00)、哇哈哈AD钙奶×1(¥8.00)（商品合计¥16.00 占比2.3%）；综合：10-12月到店消费2天/2次，到店周期中位#2，消费排名#69，在店时长#72；12月到店消费0天/0次，到店周期中位—，消费排名#69，在店时长—；最近到店2025-11-25；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：储值余额充足，可提供包厢/团建档期与专属权益；重点维护包厢/团建需求，提前锁档与套餐化
+70,卢广贤,186****6220,,680.44,0.00,0.00,订单：4单，平均单次¥170.11；打球：10.4h，平均单次2.6h；偏好：B区(62.8%)、C区(37.2%)；时间：到店均值18:15 中位18:45；离店均值20:51 中位20:41；商品：轻上椰子水×1(¥12.00)、三得利×1(¥8.00)、东鹏特饮×1(¥7.00)、冰红茶×1(¥6.00)、普通扑克×1(¥5.00)（商品合计¥38.00 占比5.6%）；综合：10-12月到店消费4天/4次，到店周期中位#18，消费排名#70，在店时长#50；12月到店消费1天/1次，到店周期中位—，消费排名#70，在店时长#46；最近到店2025-12-13；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+71,王龙,186****8011,,669.58,0.00,95.20,订单：4单，平均单次¥167.40；打球：7.8h，平均单次2.0h；偏好：麻将(99.7%)、补时长(0.3%)；时间：到店均值20:56 中位21:02；离店均值23:28 中位23:28；商品：麻将房茶位费×2(¥80.00)、东方树叶×2(¥16.00)（商品合计¥96.00 占比14.3%）；综合：10-12月到店消费2天/4次，到店周期中位#9，消费排名#71，在店时长#56；12月到店消费2天/4次，到店周期中位#7，消费排名#71，在店时长#28；最近到店2025-12-12；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导
+72,王先生,185****1125,,600.68,0.00,0.00,订单：4单，平均单次¥150.17；打球：5.9h，平均单次1.5h；偏好：S区/斯诺克(99.4%)、补时长(0.6%)；时间：到店均值18:13 中位18:31；离店均值20:09 中位19:30；商品：哇哈哈矿泉水×1(¥5.00)（商品合计¥5.00 占比0.8%）；综合：10-12月到店消费4天/4次，到店周期中位#6，消费排名#72，在店时长#63；12月到店消费1天/1次，到店周期中位—，消费排名#72，在店时长#43；最近到店2025-12-02；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+73,潘先生,186****0511,年糕(2.0h),564.93,0.00,0.00,订单：2单，平均单次¥282.46；打球：3.0h，平均单次1.5h；偏好：666(66.7%)、A区(33.3%)；时间：到店均值18:28 中位18:28；离店均值19:29 中位19:29；商品：农夫山泉苏打水×1(¥6.00)（商品合计¥6.00 占比1.1%）；综合：10-12月到店消费2天/2次，到店周期中位#11，消费排名#73，在店时长#75；12月到店消费2天/2次，到店周期中位#8，消费排名#73，在店时长#40；最近到店2025-12-13；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+74,老宋,138****4554,婉婉(1.0h)、欣怡(0.5h),541.97,0.00,2126.14,订单：2单，平均单次¥270.98；打球：1.5h，平均单次0.7h；偏好：888(66.9%)、A区(33.1%)；时间：到店均值19:24 中位19:24；离店均值20:02 中位20:02；商品：百威235毫升×12(¥180.00)（商品合计¥180.00 占比33.2%）；综合：10-12月到店消费2天/2次，到店周期中位#13，消费排名#74，在店时长#91；12月到店消费0天/0次，到店周期中位—，消费排名#74，在店时长—；最近到店2025-11-07；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+75,郭先生,156****5001,希希(1.6h),518.35,0.00,2.57,订单：2单，平均单次¥259.18；打球：7.1h，平均单次3.5h；偏好：A区(100.0%)；时间：到店均值20:53 中位20:53；离店均值次日00:32 中位次日00:32；商品：可乐×2(¥10.00)、普通扑克×2(¥10.00)、一次性手套×2(¥4.00)（商品合计¥24.00 占比4.6%）；综合：10-12月到店消费2天/2次，到店周期中位#7，消费排名#75，在店时长#60；12月到店消费0天/0次，到店周期中位—，消费排名#75，在店时长—；最近到店2025-11-01；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导
+76,孙启明,137****6325,,500.00,0.00,0.00,订单：4单，平均单次¥125.00；打球：2.0h，平均单次0.5h；偏好：补时长(100.0%)；时间：到店均值10:50 中位09:58；离店均值11:49 中位10:57；综合：10-12月到店消费4天/4次，到店周期中位#16，消费排名#76，在店时长#83；12月到店消费0天/0次，到店周期中位—，消费排名#76，在店时长—；最近到店2025-11-19；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+77,王先生,136****0168,,500.00,0.00,0.00,订单：5单，平均单次¥100.00；打球：0.1h，平均单次0.0h；偏好：补时长(100.0%)；时间：到店均值19:45 中位19:22；离店均值20:44 中位20:21；综合：10-12月到店消费4天/5次，到店周期中位#5，消费排名#77，在店时长#98；12月到店消费1天/1次，到店周期中位—，消费排名#77，在店时长#53；最近到店2025-12-03；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+78,黎先生,133****0983,奈千(1.7h)、婉婉(0.8h),470.19,0.00,0.00,订单：1单，平均单次¥470.19；打球：1.7h，平均单次1.7h；偏好：包厢(100.0%)；时间：到店均值19:17 中位19:17；离店均值21:01 中位21:01；商品：王老吉×2(¥16.00)、普通茶位×1(¥10.00)、香飘飘果汁茶×1(¥10.00)（商品合计¥36.00 占比7.7%）；综合：10-12月到店消费1天/1次，到店周期中位—，消费排名#78，在店时长#89；12月到店消费0天/0次，到店周期中位—，消费排名#78，在店时长—；最近到店2025-10-10；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：重点维护包厢/团建需求，提前锁档与套餐化
+79,常总,185****7188,年糕(1.7h)、阿清(1.7h),460.52,0.00,5000.00,订单：1单，平均单次¥460.52；打球：1.8h，平均单次1.8h；偏好：A区(100.0%)；时间：到店均值19:13 中位19:13；离店均值20:58 中位20:58；商品：哇哈哈矿泉水×2(¥10.00)、脉动×1(¥8.00)（商品合计¥18.00 占比3.9%）；综合：10-12月到店消费1天/1次，到店周期中位—，消费排名#79，在店时长#88；12月到店消费1天/1次，到店周期中位—，消费排名#79，在店时长#47；最近到店2025-12-14；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+80,张丹逸,136****6637,,439.36,0.00,0.00,订单：4单，平均单次¥109.84；打球：2.4h，平均单次0.6h；偏好：B区(99.0%)、补时长(1.0%)；时间：到店均值11:15 中位10:29；离店均值12:36 中位11:29；综合：10-12月到店消费4天/4次，到店周期中位#1，消费排名#80，在店时长#80；12月到店消费1天/1次，到店周期中位—，消费排名#80，在店时长#55；最近到店2025-12-03；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+81,陈先生,186****8238,乔西(1.8h),416.17,0.00,497.83,订单：1单，平均单次¥416.17；打球：1.8h，平均单次1.8h；偏好：包厢(100.0%)；时间：到店均值21:15 中位21:15；离店均值23:04 中位23:04；商品：普通茶位×3(¥30.00)（商品合计¥30.00 占比7.2%）；综合：10-12月到店消费1天/1次，到店周期中位—，消费排名#81，在店时长#86；12月到店消费0天/0次，到店周期中位—，消费排名#81，在店时长—；最近到店2025-11-20；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导；重点维护包厢/团建需求，提前锁档与套餐化
+82,陈先生,138****3964,,400.00,0.00,0.00,订单：3单，平均单次¥133.33；打球：2.5h，平均单次0.8h；偏好：补时长(100.0%)；时间：到店均值15:17 中位21:41；离店均值15:57 中位22:40；综合：10-12月到店消费3天/3次，到店周期中位#24，消费排名#82，在店时长#79；12月到店消费1天/1次，到店周期中位—，消费排名#82，在店时长#52；最近到店2025-12-04；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+83,张先生,136****4528,,398.35,0.00,0.00,订单：3单，平均单次¥132.78；打球：2.9h，平均单次1.0h；偏好：S区/斯诺克(98.8%)、补时长(1.2%)；时间：到店均值21:05 中位20:51；离店均值22:41 中位22:13；商品：农夫山泉苏打水×1(¥6.00)（商品合计¥6.00 占比1.5%）；综合：10-12月到店消费3天/3次，到店周期中位#15，消费排名#83，在店时长#76；12月到店消费2天/2次，到店周期中位#10，消费排名#83，在店时长#42；最近到店2025-12-17；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+84,小宇,187****8077,球球(1.5h)、悦悦(0.4h),397.50,0.00,936.07,订单：2单，平均单次¥198.75；打球：2.1h，平均单次1.0h；偏好：包厢(80.9%)、TV台(19.1%)；时间：到店均值18:07 中位18:07；离店均值19:09 中位19:09；商品：哇哈哈AD钙奶×1(¥8.00)、农夫山泉苏打水×1(¥6.00)、哇哈哈矿泉水×1(¥5.00)（商品合计¥19.00 占比4.8%）；综合：10-12月到店消费2天/2次，到店周期中位#9，消费排名#84，在店时长#82；12月到店消费0天/0次，到店周期中位—，消费排名#84，在店时长—；最近到店2025-10-13；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：重点维护包厢/团建需求，提前锁档与套餐化
+85,黄先生,191****8219,,364.33,0.00,0.00,订单：4单，平均单次¥91.08；打球：1.2h，平均单次0.3h；偏好：B区(95.4%)、补时长(4.6%)；时间：到店均值22:45 中位22:43；离店均值23:46 中位23:45；综合：10-12月到店消费3天/4次，到店周期中位#4，消费排名#85，在店时长#93；12月到店消费0天/0次，到店周期中位—，消费排名#85，在店时长—；最近到店2025-11-29；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+86,刘女士,177****7538,,362.58,0.00,0.00,订单：4单，平均单次¥90.64；打球：2.8h，平均单次0.7h；偏好：B区(99.3%)、补时长(0.7%)；时间：到店均值15:53 中位15:44；离店均值17:04 中位17:04；综合：10-12月到店消费1天/4次，到店周期中位—，消费排名#86，在店时长#77；12月到店消费0天/0次，到店周期中位—，消费排名#86，在店时长—；最近到店2025-10-14；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+87,魏先生,137****6862,,319.39,0.00,84.51,订单：1单，平均单次¥319.39；打球：5.4h，平均单次5.4h；偏好：A区(66.1%)、S区/斯诺克(33.9%)、B区(0.0%)；时间：到店均值18:44 中位18:44；离店均值次日00:08 中位次日00:08；商品：维他柠檬茶×2(¥16.00)、东方树叶×1(¥8.00)（商品合计¥24.00 占比7.5%）；综合：10-12月到店消费1天/1次，到店周期中位—，消费排名#87，在店时长#64；12月到店消费1天/1次，到店周期中位—，消费排名#87，在店时长#30；最近到店2025-12-05；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导
+88,杜先生,188****4705,,307.47,0.00,0.00,订单：3单，平均单次¥102.49；打球：2.1h，平均单次0.7h；偏好：A区(98.3%)、补时长(1.7%)；时间：到店均值21:25 中位21:44；离店均值22:45 中位22:43；商品：哇哈哈矿泉水×2(¥10.00)（商品合计¥10.00 占比3.3%）；综合：10-12月到店消费2天/3次，到店周期中位#21，消费排名#88，在店时长#81；12月到店消费0天/0次，到店周期中位—，消费排名#88，在店时长—；最近到店2025-11-02；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+89,刘先生,137****2930,,300.00,0.00,1.27,订单：1单，平均单次¥300.00；打球：0.0h，平均单次0.0h；偏好：补时长(100.0%)；时间：到店均值23:47 中位23:47；离店均值次日00:48 中位次日00:48；综合：10-12月到店消费1天/1次，到店周期中位—，消费排名#89，在店时长#100；12月到店消费1天/1次，到店周期中位—，消费排名#89，在店时长#54；最近到店2025-12-07；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导
+90,陈先生,133****6117,,300.00,0.00,0.00,订单：2单，平均单次¥150.00；打球：0.0h，平均单次0.0h；偏好：补时长(100.0%)；时间：到店均值20:50 中位20:50；离店均值21:49 中位21:49；综合：10-12月到店消费2天/2次，到店周期中位#2，消费排名#90，在店时长#99；12月到店消费0天/0次，到店周期中位—，消费排名#90，在店时长—；最近到店2025-11-21；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+91,潘先生,176****7964,,300.00,0.00,0.00,订单：1单，平均单次¥300.00；打球：3.0h，平均单次3.0h；偏好：补时长(100.0%)；时间：到店均值00:03 中位00:03；离店均值00:03 中位00:03；综合：10-12月到店消费1天/1次，到店周期中位—，消费排名#91，在店时长#74；12月到店消费1天/1次，到店周期中位—，消费排名#91，在店时长#39；最近到店2025-12-19；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+92,钟智豪,188****2803,小侯(1.8h),274.34,0.00,0.00,订单：1单，平均单次¥274.34；打球：1.8h，平均单次1.8h；偏好：A区(100.0%)；时间：到店均值23:50 中位23:50；离店均值次日01:36 中位次日01:36；综合：10-12月到店消费1天/1次，到店周期中位—，消费排名#92，在店时长#87；12月到店消费0天/0次，到店周期中位—，消费排名#92，在店时长—；最近到店2025-11-28；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+93,都先生,138****7796,素素(1.6h),269.64,0.00,0.00,订单：1单，平均单次¥269.64；打球：1.6h，平均单次1.6h；偏好：B区(100.0%)；时间：到店均值15:17 中位15:17；离店均值16:56 中位16:56；商品：东方树叶×1(¥8.00)、哇哈哈AD钙奶×1(¥8.00)（商品合计¥16.00 占比5.9%）；综合：10-12月到店消费1天/1次，到店周期中位—，消费排名#93，在店时长#90；12月到店消费0天/0次，到店周期中位—，消费排名#93，在店时长—；最近到店2025-11-13；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+94,林志铭,135****4233,小侯(1.2h),267.96,0.00,795.66,订单：3单，平均单次¥89.32；打球：2.0h，平均单次0.7h；偏好：TV台(75.2%)、S区/斯诺克(24.8%)；时间：到店均值19:45 中位20:54；离店均值20:24 中位21:40；综合：10-12月到店消费2天/3次，到店周期中位#12，消费排名#94，在店时长#84；12月到店消费0天/0次，到店周期中位—，消费排名#94，在店时长—；最近到店2025-11-30；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+95,方先生,158****6447,,248.00,0.00,0.00,订单：3单，平均单次¥82.67；打球：1.0h，平均单次0.3h；偏好：A区(97.2%)、补时长(2.8%)；时间：到店均值15:35 中位22:11；离店均值16:15 中位23:11；综合：10-12月到店消费3天/3次，到店周期中位#17，消费排名#95，在店时长#94；12月到店消费2天/2次，到店周期中位#2，消费排名#95，在店时长#48；最近到店2025-12-18；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+96,李先生,176****5124,,244.00,0.00,0.00,订单：3单，平均单次¥81.33；打球：0.9h，平均单次0.3h；偏好：A区(97.8%)、补时长(2.2%)；时间：到店均值20:59 中位21:28；离店均值21:39 中位21:28；综合：10-12月到店消费3天/3次，到店周期中位#5，消费排名#96，在店时长#95；12月到店消费0天/0次，到店周期中位—，消费排名#96，在店时长—；最近到店2025-10-13；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+97,钟先生,132****3438,,239.37,0.00,0.00,订单：1单，平均单次¥239.37；打球：2.6h，平均单次2.6h；偏好：麻将(100.0%)；时间：到店均值19:32 中位19:32；离店均值22:10 中位22:10；商品：50枸杞槟榔×1(¥65.00)、地道肠×3(¥15.00)、哇哈哈矿泉水×2(¥10.00)、蜂蜜水×1(¥10.00)、屈臣氏苏打水×1(¥8.00)、芬达×1(¥5.00)（商品合计¥113.00 占比47.2%）；综合：10-12月到店消费1天/1次，到店周期中位—，消费排名#97，在店时长#78；12月到店消费0天/0次，到店周期中位—，消费排名#97，在店时长—；最近到店2025-10-20；趋势：到店频次波动，建议按常用时段做稳定触达；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性
+98,杨,130****5960,,232.00,0.00,287.50,订单：2单，平均单次¥116.00；打球：4.0h，平均单次2.0h；偏好：B区(100.0%)；时间：到店均值18:09 中位18:09；离店均值20:11 中位20:11；综合：10-12月到店消费2天/2次，到店周期中位#2，消费排名#98，在店时长#71；12月到店消费2天/2次，到店周期中位#2，消费排名#98，在店时长#33；最近到店2025-12-05；趋势：10-12月到店频次上升，12月更活跃；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导
+99,黄国磊,131****3045,,225.33,0.00,0.22,订单：3单，平均单次¥75.11；打球：0.1h，平均单次0.0h；偏好：A区(85.5%)、补时长(14.5%)；时间：到店均值17:56 中位18:02；离店均值18:36 中位18:19；商品：蜂蜜水×2(¥20.00)（商品合计¥20.00 占比8.9%）；综合：10-12月到店消费3天/3次，到店周期中位#22，消费排名#99，在店时长#96；12月到店消费1天/1次，到店周期中位—，消费排名#99，在店时长#55；最近到店2025-12-12；趋势：12月为高峰月，具备加深运营空间；围客与客户运营建议：储值余额偏低，建议在其常用时段做补能引导
+100,周先生,159****9997,,200.00,0.00,0.00,订单：2单，平均单次¥100.00；打球：1.2h，平均单次0.6h；偏好：补时长(100.0%)；时间：到店均值02:54 中位02:54；离店均值03:19 中位03:19；综合：10-12月到店消费2天/2次，到店周期中位#2，消费排名#100，在店时长#92；12月到店消费1天/1次，到店周期中位—，消费排名#100，在店时长#50；最近到店2025-12-01；趋势：12月为高峰月，具备加深运营空间；围客与客户运营建议：保持常用时段的稳定服务供给，提升复购粘性

etl_billiards/docs/table_2025-12-19/客户_Top100_2025年10-12月_总表.md

@@ -0,0 +1,459 @@
+# 2025年10-12月 客户消费能力Top100（总表）
+## 思考过程
+以台费订单为基准汇总三类明细，满足应付金额口径，并输出Top100客户的消费/充值/助教偏好。按你的要求，先生成评价为空的版本，再在脚本末尾回填评价。
+
+## 查询说明
+消费=台费(dwd_table_fee_log.ledger_amount)+助教(dwd_assistant_service_log.ledger_amount)+商品(dwd_store_goods_sale.ledger_amount)，均为应付金额（不扣优惠），以台费订单为基准串联；充值=充值支付流水(dwd_payment.relate_type=5, pay_status=2, pay_amount>0)按支付时间切月；储值卡未使用金额=当前有效储值卡余额之和(dim_member_card_account.balance, member_card_type_name='储值卡')；喜爱助教=基础课时长+附加课时长*1.5（income_seconds换算小时），总表按10-12月汇总Top5。
+
+## SQL
+
+### Top100（按消费总额）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount,
+    sum(tfl.real_table_use_seconds) as table_use_seconds
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_info as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.order_start_time, coalesce(a.assistant_start_time, bo.order_start_time)) as order_start_time,
+    greatest(bo.order_end_time, coalesce(a.assistant_end_time, bo.order_end_time)) as order_end_time,
+    bo.table_use_seconds,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_info a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+select
+  o.member_id,
+  sum(o.order_amount) as consume_total,
+  count(*) as order_cnt
+from orders o
+where o.member_id is not null and o.member_id <> 0
+group by o.member_id
+order by consume_total desc
+limit 100;
+```
+
+### 按月消费汇总
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount,
+    sum(tfl.real_table_use_seconds) as table_use_seconds
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_info as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.order_start_time, coalesce(a.assistant_start_time, bo.order_start_time)) as order_start_time,
+    greatest(bo.order_end_time, coalesce(a.assistant_end_time, bo.order_end_time)) as order_end_time,
+    bo.table_use_seconds,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_info a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, x as (
+  select
+    o.member_id,
+    case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    o.order_amount
+  from orders o
+  where o.member_id is not null and o.member_id <> 0
+)
+select
+  member_id,
+  month_key,
+  sum(order_amount) as consume_sum
+from x
+where month_key is not null
+group by member_id, month_key;
+```
+
+### 按月充值汇总
+```sql
+with pay as (
+  select
+    p.pay_time,
+    r.member_id,
+    p.pay_amount
+  from billiards_dwd.dwd_payment p
+  join billiards_dwd.dwd_recharge_order r on r.recharge_order_id = p.relate_id
+  where p.site_id = %(site_id)s
+    and p.relate_type = 5
+    and p.pay_status = 2
+    and p.pay_amount > 0
+    and p.pay_time >= %(window_start)s::timestamptz
+    and p.pay_time <  %(window_end)s::timestamptz
+)
+, x as (
+  select
+    member_id,
+    case when pay_time >= '2025-10-01 00:00:00+08'::timestamptz and pay_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when pay_time >= '2025-11-01 00:00:00+08'::timestamptz and pay_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when pay_time >= '2025-12-01 00:00:00+08'::timestamptz and pay_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    pay_amount
+  from pay
+)
+select
+  member_id,
+  month_key,
+  sum(pay_amount) as recharge_sum
+from x
+where month_key is not null
+group by member_id, month_key;
+```
+
+### 储值卡未使用金额（当前余额汇总）
+```sql
+select
+  tenant_member_id as member_id,
+  sum(balance) as stored_value_balance
+from billiards_dwd.dim_member_card_account
+where scd2_is_current=1
+  and coalesce(is_delete,0)=0
+  and member_card_type_name='储值卡'
+  and tenant_member_id = any(%(ids)s)
+group by tenant_member_id;
+```
+
+### 喜爱助教Top5（10-12月）
+```sql
+with x as (
+  select
+    asl.tenant_member_id as member_id,
+    asl.nickname as assistant_nickname,
+    sum(case when asl.order_assistant_type=1 then asl.income_seconds else asl.income_seconds*1.5 end) / 3600.0 as weighted_hours
+  from billiards_dwd.dwd_assistant_service_log asl
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0)=0
+    and asl.tenant_member_id is not null and asl.tenant_member_id <> 0
+    and asl.start_use_time >= %(window_start)s::timestamptz
+    and asl.start_use_time <  %(window_end)s::timestamptz
+  group by asl.tenant_member_id, asl.nickname
+),
+ranked as (
+  select *, row_number() over(partition by member_id order by weighted_hours desc) as rn
+  from x
+)
+select
+  member_id,
+  string_agg(assistant_nickname || '(' || to_char(round(weighted_hours::numeric, 1), 'FM999999990.0') || 'h)', '、' order by weighted_hours desc) as fav5
+from ranked
+where rn <= 5
+group by member_id;
+```
+
+### 评价画像：订单/时长/到店日期
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount,
+    sum(tfl.real_table_use_seconds) as table_use_seconds
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_info as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.order_start_time, coalesce(a.assistant_start_time, bo.order_start_time)) as order_start_time,
+    greatest(bo.order_end_time, coalesce(a.assistant_end_time, bo.order_end_time)) as order_end_time,
+    bo.table_use_seconds,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_info a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+select
+  o.member_id,
+  count(*) as orders,
+  avg(o.order_amount) as avg_order,
+  sum(o.table_use_seconds)/3600.0 as play_hours,
+  avg(o.table_use_seconds)/3600.0 as avg_play_hours,
+  min((o.order_start_time at time zone 'Asia/Shanghai')::date) as first_visit_day,
+  max((o.order_start_time at time zone 'Asia/Shanghai')::date) as last_visit_day,
+  count(distinct (o.order_start_time at time zone 'Asia/Shanghai')::date) as visit_days,
+  sum(case when o.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then 1 else 0 end) as orders_oct,
+  sum(case when o.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and o.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then 1 else 0 end) as orders_nov,
+  sum(case when o.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and o.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then 1 else 0 end) as orders_dec
+from orders o
+where o.member_id = any(%(ids)s)
+group by o.member_id;
+```
+
+### 评价画像：到店/离店时间（小时）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount,
+    sum(tfl.real_table_use_seconds) as table_use_seconds
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_info as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.order_start_time, coalesce(a.assistant_start_time, bo.order_start_time)) as order_start_time,
+    greatest(bo.order_end_time, coalesce(a.assistant_end_time, bo.order_end_time)) as order_end_time,
+    bo.table_use_seconds,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_info a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+, t as (
+  select
+    o.member_id,
+    extract(hour from (o.order_start_time at time zone 'Asia/Shanghai'))
+      + extract(minute from (o.order_start_time at time zone 'Asia/Shanghai'))/60.0
+      + extract(second from (o.order_start_time at time zone 'Asia/Shanghai'))/3600.0 as arrive_h,
+    extract(hour from (o.order_end_time at time zone 'Asia/Shanghai'))
+      + extract(minute from (o.order_end_time at time zone 'Asia/Shanghai'))/60.0
+      + extract(second from (o.order_end_time at time zone 'Asia/Shanghai'))/3600.0 as leave_h_raw
+  from orders o
+  where o.member_id = any(%(ids)s)
+),
+tt as (
+  select
+    member_id,
+    arrive_h,
+    case when leave_h_raw < arrive_h then leave_h_raw + 24 else leave_h_raw end as leave_h
+  from t
+)
+select
+  member_id,
+  avg(arrive_h) as arrive_avg_h,
+  percentile_cont(0.5) within group (order by arrive_h) as arrive_med_h,
+  avg(leave_h) as leave_avg_h,
+  percentile_cont(0.5) within group (order by leave_h) as leave_med_h
+from tt
+group by member_id;
+```
+
+### 评价画像：球台分区偏好（按时长）
+```sql
+select
+  tfl.member_id,
+  coalesce(tfl.site_table_area_name,'') as site_table_area_name,
+  sum(tfl.real_table_use_seconds)/3600.0 as hours
+from billiards_dwd.dwd_table_fee_log tfl
+where tfl.site_id=%(site_id)s and coalesce(tfl.is_delete,0)=0
+  and tfl.start_use_time >= %(window_start)s::timestamptz and tfl.start_use_time < %(window_end)s::timestamptz
+  and tfl.member_id = any(%(ids)s)
+group by tfl.member_id, site_table_area_name;
+```
+
+### 评价画像：商品明细（名称+数量）
+```sql
+with base_orders as (
+  select order_settle_id, max(member_id) as member_id
+  from billiards_dwd.dwd_table_fee_log
+  where site_id=%(site_id)s and coalesce(is_delete,0)=0
+    and start_use_time >= %(window_start)s::timestamptz and start_use_time < %(window_end)s::timestamptz
+  group by order_settle_id
+)
+select
+  bo.member_id,
+  g.ledger_name,
+  sum(g.ledger_count) as qty,
+  sum(g.ledger_amount) as amount
+from base_orders bo
+join billiards_dwd.dwd_store_goods_sale g on g.order_settle_id = bo.order_settle_id
+where g.site_id=%(site_id)s and coalesce(g.is_delete,0)=0
+  and bo.member_id = any(%(ids)s)
+group by bo.member_id, g.ledger_name;
+```
+
+### 评价画像：到店日期明细（用于周期/近期分析）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    max(tfl.ledger_end_time) as order_end_time,
+    sum(tfl.ledger_amount) as table_amount,
+    sum(tfl.real_table_use_seconds) as table_use_seconds
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= '2025-10-01 00:00:00+08'::timestamptz
+    and tfl.start_use_time <  '2026-01-01 00:00:00+08'::timestamptz
+  group by tfl.order_settle_id
+),
+assistant_info as (
+  select
+    asl.order_settle_id,
+    sum(asl.ledger_amount) as assistant_amount,
+    min(asl.start_use_time) as assistant_start_time,
+    max(asl.last_use_time) as assistant_end_time
+  from billiards_dwd.dwd_assistant_service_log asl
+  join base_orders bo on bo.order_settle_id = asl.order_settle_id
+  where asl.site_id = %(site_id)s
+    and coalesce(asl.is_delete,0) = 0
+  group by asl.order_settle_id
+),
+goods_amount as (
+  select
+    g.order_settle_id,
+    sum(g.ledger_amount) as goods_amount
+  from billiards_dwd.dwd_store_goods_sale g
+  join base_orders bo on bo.order_settle_id = g.order_settle_id
+  where g.site_id = %(site_id)s
+    and coalesce(g.is_delete,0) = 0
+  group by g.order_settle_id
+),
+orders as (
+  select
+    bo.order_settle_id,
+    bo.member_id,
+    least(bo.order_start_time, coalesce(a.assistant_start_time, bo.order_start_time)) as order_start_time,
+    greatest(bo.order_end_time, coalesce(a.assistant_end_time, bo.order_end_time)) as order_end_time,
+    bo.table_use_seconds,
+    coalesce(bo.table_amount,0) + coalesce(a.assistant_amount,0) + coalesce(g.goods_amount,0) as order_amount
+  from base_orders bo
+  left join assistant_info a on a.order_settle_id = bo.order_settle_id
+  left join goods_amount g on g.order_settle_id = bo.order_settle_id
+)
+
+select
+  o.member_id,
+  (o.order_start_time at time zone 'Asia/Shanghai')::date as visit_date,
+  count(*) as orders,
+  sum(o.order_amount) as amount
+from orders o
+where o.member_id = any(%(ids)s)
+group by o.member_id, visit_date
+order by o.member_id, visit_date;
+```

etl_billiards/docs/table_2025-12-19/财务_优惠分布_2025年10-12月.csv

@@ -0,0 +1,119 @@
+2025年10-12月 财务优惠（会员折扣+台费调账）分布
+优惠=会员折扣(dwd_table_fee_log.member_discount_amount)+台费调账(dwd_table_fee_adjust.ledger_amount)，按订单归集后汇总到客户(member_id)，按订单最早开台时间切月；不含团购抵扣等其它优惠。
+
+客户名称,手机号(脱敏),VIP卡/等级,10月,10月,10月,11月,11月,11月,12月,12月,12月,10-12月
+客户名称,手机号(脱敏),VIP卡/等级,会员折扣(元),台费调账(元),合计优惠(元),会员折扣(元),台费调账(元),合计优惠(元),会员折扣(元),台费调账(元),合计优惠(元),合计优惠(元)
+轩哥,188****7530,储值卡,0.00,9212.40,9212.40,0.00,10685.47,10685.47,0.00,5331.76,5331.76,25229.63
+曾巧明,186****1488,年卡,9649.06,0.00,9649.06,8811.59,0.00,8811.59,6283.51,0.00,6283.51,24744.16
+黄生,136****9719,年卡,7821.94,0.00,7821.94,6353.05,0.00,6353.05,4181.24,0.00,4181.24,18356.23
+蔡总,159****8893,台费卡,0.00,0.00,0.00,0.00,12274.20,12274.20,0.00,2587.28,2587.28,14861.48
+罗先生,139****6996,年卡,3315.05,566.48,3881.53,4474.12,460.59,4934.71,2087.14,1320.22,3407.36,12223.60
+陈腾鑫,178****8218,台费卡,0.00,2843.20,2843.20,0.00,4913.52,4913.52,0.00,1935.78,1935.78,9692.50
+谢俊,186****5198,年卡,4193.93,0.00,4193.93,4089.12,0.00,4089.12,969.66,0.00,969.66,9252.71
+桂先生,166****7275,月卡,2175.11,0.00,2175.11,4009.66,301.18,4310.84,0.00,568.24,568.24,7054.19
+张先生,139****8852,年卡,1317.04,1307.13,2624.17,2114.89,168.51,2283.40,2007.34,0.00,2007.34,6914.91
+艾宇民,150****9958,年卡,2342.73,0.00,2342.73,2543.78,545.12,3088.90,1179.56,0.00,1179.56,6611.19
+曾丹烨,139****3242,储值卡,0.00,2348.26,2348.26,0.00,2573.25,2573.25,0.00,1240.20,1240.20,6161.71
+葛先生,138****8071,储值卡,0.00,0.00,0.00,0.00,3548.54,3548.54,0.00,2593.02,2593.02,6141.56
+小燕,178****1334,储值卡,0.00,0.00,0.00,0.00,2250.36,2250.36,0.00,2228.54,2228.54,4478.90
+孟紫龙,176****3741,月卡,0.00,0.00,0.00,576.78,0.00,576.78,2762.95,0.00,2762.95,3339.73
+江先生,188****4838,储值卡,0.00,344.54,344.54,0.00,1454.41,1454.41,0.00,405.54,405.54,2204.49
+郑先生,159****4331,储值卡,0.00,0.00,0.00,0.00,1659.16,1659.16,0.00,463.02,463.02,2122.18
+黄先生,135****3507,台费卡,0.00,631.08,631.08,0.00,631.56,631.56,0.00,810.48,810.48,2073.12
+孟紫龙（该会员已注销）,176****37411,月卡,1785.72,0.00,1785.72,0.00,164.00,164.00,0.00,0.00,0.00,1949.72
+T,180****9962,储值卡,0.00,578.93,578.93,0.00,461.72,461.72,0.00,839.34,839.34,1879.99
+林先生,133****1070,储值卡,0.00,0.00,0.00,0.00,199.95,199.95,0.00,1417.70,1417.70,1617.65
+游,172****6666,储值卡,0.00,0.00,0.00,0.00,921.43,921.43,0.00,600.85,600.85,1522.28
+林先生,137****8785,储值卡,0.00,0.00,0.00,0.00,1114.20,1114.20,0.00,186.45,186.45,1300.65
+桂先生（该会员已注销）,166****72751,月卡,988.04,303.77,1291.81,0.00,0.00,0.00,0.00,0.00,0.00,1291.81
+叶先生,138****9539,储值卡,0.00,703.64,703.64,0.00,341.29,341.29,0.00,99.41,99.41,1144.34
+陈德韩,134****7864,台费卡,0.00,1091.02,1091.02,0.00,0.00,0.00,0.00,0.00,0.00,1091.02
+周周,198****8725,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,1061.95,1061.95,1061.95
+胡先生,186****3391,储值卡,0.00,0.00,0.00,0.00,785.20,785.20,0.00,100.59,100.59,885.79
+小熊,139****0145,台费卡,0.00,297.01,297.01,0.00,530.45,530.45,0.00,0.00,0.00,827.46
+夏,191****2851,储值卡,0.00,659.78,659.78,0.00,137.90,137.90,0.00,0.00,0.00,797.68
+李先生,186****8308,储值卡,0.00,0.00,0.00,0.00,777.27,777.27,0.00,0.00,0.00,777.27
+叶总,137****3287,储值卡,0.00,737.12,737.12,0.00,0.00,0.00,0.00,0.00,0.00,737.12
+羊,187****5094,储值卡,0.00,38.64,38.64,0.00,647.57,647.57,0.00,0.00,0.00,686.21
+黄先生,158****2109,储值卡,0.00,240.00,240.00,0.00,326.90,326.90,0.00,100.06,100.06,666.96
+卢广贤,186****6220,年卡,263.07,0.00,263.07,250.51,0.00,250.51,128.86,0.00,128.86,642.44
+陶,189****2151,储值卡,0.00,604.89,604.89,0.00,0.00,0.00,0.00,0.00,0.00,604.89
+大G,186****4598,台费卡,0.00,0.00,0.00,0.00,245.70,245.70,0.00,356.31,356.31,602.01
+明哥,166****0999,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,547.68,547.68,547.68
+梅,136****4552,储值卡,0.00,0.00,0.00,0.00,99.02,99.02,0.00,372.66,372.66,471.68
+阿亮,159****2628,储值卡,0.00,0.00,0.00,0.00,361.34,361.34,0.00,100.86,100.86,462.20
+牛先生,152****5159,台费卡,0.00,0.00,0.00,0.00,424.31,424.31,0.00,0.00,0.00,424.31
+罗超杰,137****8012,储值卡,0.00,245.95,245.95,0.00,164.54,164.54,0.00,0.00,0.00,410.49
+陈世,134****1938,储值卡,0.00,199.51,199.51,0.00,198.89,198.89,0.00,0.00,0.00,398.40
+曾先生,133****1235,储值卡,0.00,36.67,36.67,0.00,265.71,265.71,0.00,90.45,90.45,392.83
+李,131****9882,储值卡,0.00,254.13,254.13,0.00,138.12,138.12,0.00,0.00,0.00,392.25
+陈泽斌,132****6060,储值卡,0.00,77.00,77.00,0.00,222.00,222.00,0.00,78.00,78.00,377.00
+王先生,136****0168,台费卡,0.00,0.00,0.00,0.00,235.00,235.00,0.00,65.00,65.00,300.00
+林先生,188****0332,储值卡,0.00,0.00,0.00,0.00,252.11,252.11,0.00,0.00,0.00,252.11
+歌神,188****2164,储值卡,0.00,246.28,246.28,0.00,0.00,0.00,0.00,0.00,0.00,246.28
+陈先生,159****2829,台费卡,0.00,21.00,21.00,0.00,218.13,218.13,0.00,0.00,0.00,239.13
+黄先生,191****8219,台费卡,0.00,0.00,0.00,0.00,229.16,229.16,0.00,0.00,0.00,229.16
+汪先生,139****6339,储值卡,0.00,228.07,228.07,0.00,0.00,0.00,0.00,0.00,0.00,228.07
+王龙,186****8011,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,200.78,200.78,200.78
+罗先生,139****9222,储值卡,0.00,0.00,0.00,0.00,199.17,199.17,0.00,0.00,0.00,199.17
+吴生,136****3341,储值卡,0.00,0.00,0.00,0.00,99.23,99.23,0.00,99.92,99.92,199.15
+刘女士,177****7538,储值卡,0.00,193.29,193.29,0.00,0.00,0.00,0.00,0.00,0.00,193.29
+林总,138****1180,储值卡,0.00,0.00,0.00,0.00,183.25,183.25,0.00,0.00,0.00,183.25
+陈淑涛,132****5485,储值卡,0.00,0.00,0.00,0.00,180.86,180.86,0.00,0.00,0.00,180.86
+张丹逸,136****6637,储值卡,0.00,133.68,133.68,0.00,0.00,0.00,0.00,45.00,45.00,178.68
+孙启明,137****6325,台费卡,0.00,90.00,90.00,0.00,85.00,85.00,0.00,0.00,0.00,175.00
+陈先生,138****3964,储值卡,0.00,98.00,98.00,0.00,0.00,0.00,0.00,60.00,60.00,158.00
+杜先生,188****4705,台费卡,0.00,48.00,48.00,0.00,99.73,99.73,0.00,0.00,0.00,147.73
+周先生,193****6822,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,135.65,135.65,135.65
+君姐,166****4594,储值卡,0.00,0.00,0.00,0.00,134.10,134.10,0.00,0.00,0.00,134.10
+昌哥,137****1229,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,122.03,122.03,122.03
+amy,137****8221,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,116.28,116.28,116.28
+杨,130****5960,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,116.00,116.00,116.00
+方先生,153****3185,储值卡,0.00,100.27,100.27,0.00,0.00,0.00,0.00,0.00,0.00,100.27
+方先生,158****6447,台费卡,0.00,0.00,0.00,0.00,72.00,72.00,0.00,28.00,28.00,100.00
+陈先生,133****6117,台费卡,0.00,0.00,0.00,0.00,100.00,100.00,0.00,0.00,0.00,100.00
+周先生,159****9997,台费卡,0.00,0.00,0.00,0.00,61.00,61.00,0.00,37.00,37.00,98.00
+王姐,158****8819,储值卡,0.00,97.40,97.40,0.00,0.00,0.00,0.00,0.00,0.00,97.40
+老宋,138****4554,储值卡,0.00,0.00,0.00,0.00,93.95,93.95,0.00,0.00,0.00,93.95
+李先生,176****5124,储值卡,0.00,90.00,90.00,0.00,0.00,0.00,0.00,0.00,0.00,90.00
+曾先生,159****0492,储值卡,0.00,88.00,88.00,0.00,0.00,0.00,0.00,0.00,0.00,88.00
+张先生,136****4528,台费卡,0.00,0.00,0.00,0.00,37.00,37.00,0.00,45.00,45.00,82.00
+刘先生,137****2930,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,82.00,82.00,82.00
+潘先生,176****7964,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,74.00,74.00,74.00
+王先生,183****9763,台费卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,74.00,74.00,74.00
+王先生,185****1125,台费卡,0.00,0.00,0.00,0.00,73.00,73.00,0.00,0.00,0.00,73.00
+清,130****3087,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,72.82,72.82,72.82
+贺斌,150****0885,活动抵用券,0.00,71.27,71.27,0.00,0.00,0.00,0.00,0.00,0.00,71.27
+李先生,186****9266,储值卡,0.00,0.00,0.00,0.00,69.00,69.00,0.00,0.00,0.00,69.00
+林志铭,135****4233,储值卡,0.00,0.00,0.00,0.00,66.50,66.50,0.00,0.00,0.00,66.50
+候,131****0323,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,62.79,62.79,62.79
+吕先生,155****0663,储值卡,0.00,62.54,62.54,0.00,0.00,0.00,0.00,0.00,0.00,62.54
+梁先生,134****2609,台费卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,59.00,59.00,59.00
+肖先生,156****6427,台费卡,0.00,0.00,0.00,0.00,7.00,7.00,0.00,50.00,50.00,57.00
+婉婉,183****2742,储值卡,0.00,0.00,0.00,0.00,37.47,37.47,0.00,0.00,0.00,37.47
+陈小姐,138****0778,储值卡,0.00,32.49,32.49,0.00,0.00,0.00,0.00,0.00,0.00,32.49
+黄国磊,131****3045,台费卡,0.00,0.30,0.30,0.00,19.00,19.00,0.00,0.00,0.00,19.30
+小宇,187****8077,储值卡,0.00,13.42,13.42,0.00,0.00,0.00,0.00,0.00,0.00,13.42
+胡总,133****3091,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,1.00,1.00,1.00
+李先生,131****4000,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+周先生,173****7775,台费卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+林先生,159****0021,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+魏先生,137****6862,台费卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+都先生,138****7796,活动抵用券,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+陈先生,188****8626,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+冯先生,155****0348,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+常总,185****7188,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+邓飛,136****9597,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+李先生,134****4343,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+谭先生,138****3185,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+黎先生,133****0983,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+刘哥,135****0020,台费卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+孙先生,135****3191,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+钟智豪,188****2803,活动抵用券,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+陈先生,186****8238,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+郭先生,156****5001,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+钟先生,132****3438,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+唐先生,135****0785,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+陈先生,139****0419,储值卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+潘先生,186****0511,台费卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00
+罗超,137****0990,台费卡,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00,0.00

etl_billiards/docs/table_2025-12-19/财务_优惠分布_2025年10-12月.md

@@ -0,0 +1,52 @@
+# 2025年10-12月 财务优惠（会员折扣+台费调账）分布
+## 思考过程
+用台费订单为基准关联调账表，再按客户+月份汇总，输出“谁享受了优惠”及金额分布。
+
+## 查询说明
+优惠=会员折扣(dwd_table_fee_log.member_discount_amount)+台费调账(dwd_table_fee_adjust.ledger_amount)，按订单归集后汇总到客户(member_id)，按订单最早开台时间切月；不含团购抵扣等其它优惠。
+
+## SQL
+
+### 优惠分布（客户+月份）
+```sql
+with base_orders as (
+  select
+    tfl.order_settle_id,
+    max(tfl.member_id) as member_id,
+    min(tfl.start_use_time) as order_start_time,
+    sum(tfl.member_discount_amount) as member_discount_amount
+  from billiards_dwd.dwd_table_fee_log tfl
+  where tfl.site_id = %(site_id)s
+    and coalesce(tfl.is_delete,0) = 0
+    and tfl.start_use_time >= %(window_start)s::timestamptz
+    and tfl.start_use_time <  %(window_end)s::timestamptz
+  group by tfl.order_settle_id
+),
+adjusts as (
+  select
+    tfa.order_settle_id,
+    sum(tfa.ledger_amount) as adjust_amount
+  from billiards_dwd.dwd_table_fee_adjust tfa
+  join base_orders bo on bo.order_settle_id = tfa.order_settle_id
+  where tfa.site_id = %(site_id)s
+    and coalesce(tfa.is_delete,0) = 0
+  group by tfa.order_settle_id
+)
+, x as (
+  select
+    bo.member_id,
+    case when bo.order_start_time >= '2025-10-01 00:00:00+08'::timestamptz and bo.order_start_time < '2025-11-01 00:00:00+08'::timestamptz then '2025-10' when bo.order_start_time >= '2025-11-01 00:00:00+08'::timestamptz and bo.order_start_time < '2025-12-01 00:00:00+08'::timestamptz then '2025-11' when bo.order_start_time >= '2025-12-01 00:00:00+08'::timestamptz and bo.order_start_time < '2026-01-01 00:00:00+08'::timestamptz then '2025-12' else null end as month_key,
+    coalesce(bo.member_discount_amount,0) as member_discount_amount,
+    coalesce(a.adjust_amount,0) as adjust_amount
+  from base_orders bo
+  left join adjusts a on a.order_settle_id = bo.order_settle_id
+)
+select
+  member_id,
+  month_key,
+  sum(member_discount_amount) as member_discount_sum,
+  sum(adjust_amount) as adjust_sum
+from x
+where month_key is not null
+group by member_id, month_key;
+```

etl_billiards/docs/在线抓取，更新ODS ，然后将更新的ODS内容，对应到DWD的更新。.md

@@ -0,0 +1,34 @@
+在线抓取，更新ODS ，然后将更新的ODS内容，对应到DWD的更新。
+
+
+可以按“两段定时”跑：先在线抓取+入库更新 ODS，再跑 DWD_LOAD_FROM_ODS 把新增/变更同步到 DWD。CLI 用 python -m etl_billiards.cli.main。
+
+1) ODS：在线抓取 + 入库（FULL）
+python -m etl_billiards.cli.main ^
+  --pipeline-flow FULL ^
+  --tasks PRODUCTS,TABLES,MEMBERS,ASSISTANTS,PACKAGES_DEF,ORDERS,PAYMENTS,REFUNDS,COUPON_USAGE,INVENTORY_CHANGE,TOPUPS,TABLE_DISCOUNT,ASSISTANT_ABOLISH,LEDGER ^
+  --pg-dsn "%PG_DSN%" ^
+  --store-id %STORE_ID% ^
+  --api-token "%API_TOKEN%"
+（可选）指定落盘目录：加 --fetch-root "D:\etl\json"；美化 JSON：--write-pretty-json
+
+2) DWD：ODS → DWD
+python -m etl_billiards.cli.main ^
+  --pipeline-flow INGEST_ONLY ^
+  --tasks DWD_LOAD_FROM_ODS ^
+  --pg-dsn "%PG_DSN%" ^
+  --store-id %STORE_ID%
+推荐的环境变量
+PG_DSN=postgresql://user:pwd@host:5432/db
+STORE_ID=...
+API_TOKEN=...
+（可选）JSON_FETCH_ROOT=... / FETCH_ROOT=...，LOG_ROOT=...
+如果你希望“一条命令顺序跑完 ODS+DWD”，也可以直接：
+
+python -m etl_billiards.cli.main ^
+  --pipeline-flow FULL ^
+  --tasks PRODUCTS,TABLES,MEMBERS,ASSISTANTS,PACKAGES_DEF,ORDERS,PAYMENTS,REFUNDS,COUPON_USAGE,INVENTORY_CHANGE,TOPUPS,TABLE_DISCOUNT,ASSISTANT_ABOLISH,LEDGER,DWD_LOAD_FROM_ODS ^
+  --pg-dsn "%PG_DSN%" ^
+  --store-id %STORE_ID% ^
+  --api-token "%API_TOKEN%"
+（这会对前半段任务走在线抓取+入库，对 DWD_LOAD_FROM_ODS 只做入库阶段，因为它没有抓取逻辑。）

etl_billiards/etl_billiards/reports/dwd_quality_report.json

@@ -0,0 +1,692 @@
+{
+  "generated_at": "2026-01-16T01:53:35.751172",
+  "tables": [
+    {
+      "dwd_table": "billiards_dwd.dim_site",
+      "ods_table": "billiards_ods.table_fee_transactions",
+      "count": {
+        "dwd": 1,
+        "ods": 17217,
+        "diff": -17216
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dim_site_ex",
+      "ods_table": "billiards_ods.table_fee_transactions",
+      "count": {
+        "dwd": 1,
+        "ods": 17217,
+        "diff": -17216
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dim_table",
+      "ods_table": "billiards_ods.site_tables_master",
+      "count": {
+        "dwd": 74,
+        "ods": 74,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dim_table_ex",
+      "ods_table": "billiards_ods.site_tables_master",
+      "count": {
+        "dwd": 74,
+        "ods": 74,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dim_assistant",
+      "ods_table": "billiards_ods.assistant_accounts_master",
+      "count": {
+        "dwd": 64,
+        "ods": 64,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dim_assistant_ex",
+      "ods_table": "billiards_ods.assistant_accounts_master",
+      "count": {
+        "dwd": 64,
+        "ods": 64,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dim_member",
+      "ods_table": "billiards_ods.member_profiles",
+      "count": {
+        "dwd": 552,
+        "ods": 552,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dim_member_ex",
+      "ods_table": "billiards_ods.member_profiles",
+      "count": {
+        "dwd": 552,
+        "ods": 552,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dim_member_card_account",
+      "ods_table": "billiards_ods.member_stored_value_cards",
+      "count": {
+        "dwd": 938,
+        "ods": 938,
+        "diff": 0
+      },
+      "amounts": [
+        {
+          "column": "balance",
+          "dwd_sum": 406040.3,
+          "ods_sum": 406040.3,
+          "diff": 0.0
+        }
+      ]
+    },
+    {
+      "dwd_table": "billiards_dwd.dim_member_card_account_ex",
+      "ods_table": "billiards_ods.member_stored_value_cards",
+      "count": {
+        "dwd": 938,
+        "ods": 938,
+        "diff": 0
+      },
+      "amounts": [
+        {
+          "column": "deliveryfeededuct",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        }
+      ]
+    },
+    {
+      "dwd_table": "billiards_dwd.dim_tenant_goods",
+      "ods_table": "billiards_ods.tenant_goods_master",
+      "count": {
+        "dwd": 170,
+        "ods": 170,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dim_tenant_goods_ex",
+      "ods_table": "billiards_ods.tenant_goods_master",
+      "count": {
+        "dwd": 170,
+        "ods": 170,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dim_store_goods",
+      "ods_table": "billiards_ods.store_goods_master",
+      "count": {
+        "dwd": 169,
+        "ods": 169,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dim_store_goods_ex",
+      "ods_table": "billiards_ods.store_goods_master",
+      "count": {
+        "dwd": 169,
+        "ods": 169,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dim_goods_category",
+      "ods_table": "billiards_ods.stock_goods_category_tree",
+      "count": {
+        "dwd": 26,
+        "ods": 9,
+        "diff": 17
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dim_groupbuy_package",
+      "ods_table": "billiards_ods.group_buy_packages",
+      "count": {
+        "dwd": 34,
+        "ods": 34,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dim_groupbuy_package_ex",
+      "ods_table": "billiards_ods.group_buy_packages",
+      "count": {
+        "dwd": 34,
+        "ods": 34,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_settlement_head",
+      "ods_table": "billiards_ods.settlement_records",
+      "count": {
+        "dwd": 22055,
+        "ods": 22055,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_settlement_head_ex",
+      "ods_table": "billiards_ods.settlement_records",
+      "count": {
+        "dwd": 22055,
+        "ods": 22055,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_table_fee_log",
+      "ods_table": "billiards_ods.table_fee_transactions",
+      "count": {
+        "dwd": 17217,
+        "ods": 17217,
+        "diff": 0
+      },
+      "amounts": [
+        {
+          "column": "adjust_amount",
+          "dwd_sum": 294914.58,
+          "ods_sum": 294914.58,
+          "diff": 0.0
+        },
+        {
+          "column": "coupon_promotion_amount",
+          "dwd_sum": 896198.51,
+          "ods_sum": 896198.51,
+          "diff": 0.0
+        },
+        {
+          "column": "ledger_amount",
+          "dwd_sum": 1961129.99,
+          "ods_sum": 1961129.99,
+          "diff": 0.0
+        },
+        {
+          "column": "member_discount_amount",
+          "dwd_sum": 186470.94,
+          "ods_sum": 186470.94,
+          "diff": 0.0
+        },
+        {
+          "column": "real_table_charge_money",
+          "dwd_sum": 775239.96,
+          "ods_sum": 775239.96,
+          "diff": 0.0
+        }
+      ]
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_table_fee_log_ex",
+      "ods_table": "billiards_ods.table_fee_transactions",
+      "count": {
+        "dwd": 17217,
+        "ods": 17217,
+        "diff": 0
+      },
+      "amounts": [
+        {
+          "column": "fee_total",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        },
+        {
+          "column": "mgmt_fee",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        },
+        {
+          "column": "service_money",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        },
+        {
+          "column": "used_card_amount",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        }
+      ]
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_table_fee_adjust",
+      "ods_table": "billiards_ods.table_fee_discount_records",
+      "count": {
+        "dwd": 2648,
+        "ods": 2648,
+        "diff": 0
+      },
+      "amounts": [
+        {
+          "column": "ledger_amount",
+          "dwd_sum": 303909.03,
+          "ods_sum": 303909.03,
+          "diff": 0.0
+        }
+      ]
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_table_fee_adjust_ex",
+      "ods_table": "billiards_ods.table_fee_discount_records",
+      "count": {
+        "dwd": 2648,
+        "ods": 2648,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_store_goods_sale",
+      "ods_table": "billiards_ods.store_goods_sales_records",
+      "count": {
+        "dwd": 17563,
+        "ods": 17563,
+        "diff": 0
+      },
+      "amounts": [
+        {
+          "column": "cost_money",
+          "dwd_sum": 3116.75,
+          "ods_sum": 3116.75,
+          "diff": 0.0
+        },
+        {
+          "column": "ledger_amount",
+          "dwd_sum": 373588.1,
+          "ods_sum": 373588.1,
+          "diff": 0.0
+        },
+        {
+          "column": "real_goods_money",
+          "dwd_sum": 351914.9,
+          "ods_sum": 351914.9,
+          "diff": 0.0
+        }
+      ]
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_store_goods_sale_ex",
+      "ods_table": "billiards_ods.store_goods_sales_records",
+      "count": {
+        "dwd": 466588,
+        "ods": 17563,
+        "diff": 449025
+      },
+      "amounts": [
+        {
+          "column": "coupon_deduct_money",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        },
+        {
+          "column": "discount_money",
+          "dwd_sum": 571932.5,
+          "ods_sum": 21673.2,
+          "diff": 550259.3
+        },
+        {
+          "column": "member_discount_amount",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        },
+        {
+          "column": "option_coupon_deduct_money",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        },
+        {
+          "column": "option_member_discount_money",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        },
+        {
+          "column": "point_discount_money",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        },
+        {
+          "column": "point_discount_money_cost",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        },
+        {
+          "column": "push_money",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        }
+      ]
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_assistant_service_log",
+      "ods_table": "billiards_ods.assistant_service_records",
+      "count": {
+        "dwd": 4252,
+        "ods": 4666,
+        "diff": -414
+      },
+      "amounts": [
+        {
+          "column": "coupon_deduct_money",
+          "dwd_sum": 10489.93,
+          "ods_sum": 10879.86,
+          "diff": -389.9300000000003
+        },
+        {
+          "column": "ledger_amount",
+          "dwd_sum": 1336399.55,
+          "ods_sum": 1459374.61,
+          "diff": -122975.06000000006
+        }
+      ]
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_assistant_service_log_ex",
+      "ods_table": "billiards_ods.assistant_service_records",
+      "count": {
+        "dwd": 4666,
+        "ods": 4666,
+        "diff": 0
+      },
+      "amounts": [
+        {
+          "column": "manual_discount_amount",
+          "dwd_sum": 414.17,
+          "ods_sum": 414.17,
+          "diff": 0.0
+        },
+        {
+          "column": "member_discount_amount",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        },
+        {
+          "column": "service_money",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        }
+      ]
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_assistant_trash_event",
+      "ods_table": "billiards_ods.assistant_cancellation_records",
+      "count": {
+        "dwd": 88,
+        "ods": 88,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_assistant_trash_event_ex",
+      "ods_table": "billiards_ods.assistant_cancellation_records",
+      "count": {
+        "dwd": 88,
+        "ods": 88,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_member_balance_change",
+      "ods_table": "billiards_ods.member_balance_changes",
+      "count": {
+        "dwd": 4492,
+        "ods": 4492,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_member_balance_change_ex",
+      "ods_table": "billiards_ods.member_balance_changes",
+      "count": {
+        "dwd": 4492,
+        "ods": 4492,
+        "diff": 0
+      },
+      "amounts": [
+        {
+          "column": "refund_amount",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        }
+      ]
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_groupbuy_redemption",
+      "ods_table": "billiards_ods.group_buy_redemption_records",
+      "count": {
+        "dwd": 10544,
+        "ods": 10544,
+        "diff": 0
+      },
+      "amounts": [
+        {
+          "column": "coupon_money",
+          "dwd_sum": 668774.0,
+          "ods_sum": 668774.0,
+          "diff": 0.0
+        },
+        {
+          "column": "ledger_amount",
+          "dwd_sum": 641028.09,
+          "ods_sum": 641028.09,
+          "diff": 0.0
+        }
+      ]
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_groupbuy_redemption_ex",
+      "ods_table": "billiards_ods.group_buy_redemption_records",
+      "count": {
+        "dwd": 10544,
+        "ods": 10544,
+        "diff": 0
+      },
+      "amounts": [
+        {
+          "column": "assistant_promotion_money",
+          "dwd_sum": 7353.59,
+          "ods_sum": 7353.59,
+          "diff": 0.0
+        },
+        {
+          "column": "assistant_service_promotion_money",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        },
+        {
+          "column": "goods_promotion_money",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        },
+        {
+          "column": "recharge_promotion_money",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        },
+        {
+          "column": "reward_promotion_money",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        },
+        {
+          "column": "table_service_promotion_money",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        }
+      ]
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_platform_coupon_redemption",
+      "ods_table": "billiards_ods.platform_coupon_redemption_records",
+      "count": {
+        "dwd": 16086,
+        "ods": 16086,
+        "diff": 0
+      },
+      "amounts": [
+        {
+          "column": "coupon_money",
+          "dwd_sum": 1040212.0,
+          "ods_sum": 1040212.0,
+          "diff": 0.0
+        }
+      ]
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_platform_coupon_redemption_ex",
+      "ods_table": "billiards_ods.platform_coupon_redemption_records",
+      "count": {
+        "dwd": 16086,
+        "ods": 16086,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_recharge_order",
+      "ods_table": "billiards_ods.recharge_settlements",
+      "count": {
+        "dwd": 431,
+        "ods": 431,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_recharge_order_ex",
+      "ods_table": "billiards_ods.recharge_settlements",
+      "count": {
+        "dwd": 431,
+        "ods": 431,
+        "diff": 0
+      },
+      "amounts": []
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_payment",
+      "ods_table": "billiards_ods.payment_transactions",
+      "count": {
+        "dwd": 21611,
+        "ods": 21611,
+        "diff": 0
+      },
+      "amounts": [
+        {
+          "column": "pay_amount",
+          "dwd_sum": 2073449.0,
+          "ods_sum": 2073449.0,
+          "diff": 0.0
+        }
+      ]
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_refund",
+      "ods_table": "billiards_ods.refund_transactions",
+      "count": {
+        "dwd": 42,
+        "ods": 42,
+        "diff": 0
+      },
+      "amounts": [
+        {
+          "column": "channel_fee",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        },
+        {
+          "column": "pay_amount",
+          "dwd_sum": -68958.0,
+          "ods_sum": -68958.0,
+          "diff": 0.0
+        }
+      ]
+    },
+    {
+      "dwd_table": "billiards_dwd.dwd_refund_ex",
+      "ods_table": "billiards_ods.refund_transactions",
+      "count": {
+        "dwd": 42,
+        "ods": 42,
+        "diff": 0
+      },
+      "amounts": [
+        {
+          "column": "balance_frozen_amount",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        },
+        {
+          "column": "card_frozen_amount",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        },
+        {
+          "column": "refund_amount",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        },
+        {
+          "column": "round_amount",
+          "dwd_sum": 0.0,
+          "ods_sum": 0.0,
+          "diff": 0.0
+        }
+      ]
+    }
+  ],
+  "note": "行数/金额核对，金额字段基于列名包含 amount/money/fee/balance 的数值列自动扫描。"
+}

etl_billiards/fetch-test/README.md

@@ -0,0 +1,25 @@
+# fetch-test
+
+用于放置“接口联调/规则验证”的一次性脚本（不影响主流程）。
+
+## 近期记录 vs 历史记录(Former) 对比
+
+脚本：`fetch-test/compare_recent_former_endpoints.py`
+
+默认对比窗口（end 为次日 00:00:00，与 ETL 窗口一致）：
+- 近期：2025-12-01 ~ 2025-12-15
+- 历史：2025-08-01 ~ 2025-08-15
+
+运行：
+```bash
+cd etl_billiards
+python fetch-test/compare_recent_former_endpoints.py
+```
+
+输出：
+- `etl_billiards/fetch-test/recent_vs_former_report.md`
+- `etl_billiards/fetch-test/recent_vs_former_report.json`
+
+依赖：
+- `.env` 需配置 `API_TOKEN`（或 `FICOO_TOKEN`）与 `STORE_ID`，并保证 `API_BASE` 正确。
+

etl_billiards/fetch-test/compare_recent_former_endpoints.py

@@ -0,0 +1,434 @@
+# -*- coding: utf-8 -*-
+"""
+对比“近期记录”与“历史记录(Former)”接口：
+- 是否能正确响应（HTTP + API code==0）
+- 返回字段（基于 sample records 的 JSON path）是否一致
+
+默认时间窗口（与 ETL 窗口语义一致，end 为次日 00:00:00）：
+- 近期：2025-12-01 ~ 2025-12-15（end=2025-12-16 00:00:00）
+- 历史：2025-08-01 ~ 2025-08-15（end=2025-08-16 00:00:00）
+"""
+
+from __future__ import annotations
+
+import argparse
+import json
+import sys
+from dataclasses import asdict, dataclass
+from datetime import datetime, timedelta
+from pathlib import Path
+from typing import Any, Iterable
+
+from dateutil import parser as dtparser
+from zoneinfo import ZoneInfo
+
+PROJECT_ROOT = Path(__file__).resolve().parents[1]
+if str(PROJECT_ROOT) not in sys.path:
+    sys.path.insert(0, str(PROJECT_ROOT))
+
+from api.client import APIClient
+from api.endpoint_routing import derive_former_endpoint as derive_former_endpoint_shared
+from config.settings import AppConfig
+from models.parsers import TypeParser
+from tasks.ods_json_archive_task import EndpointSpec, OdsJsonArchiveTask
+
+CHINESE_NAMES: dict[str, str] = {
+    "/MemberProfile/GetMemberCardBalanceChange": "会员余额变动",
+    "/AssistantPerformance/GetOrderAssistantDetails": "助教服务记录",
+    "/AssistantPerformance/GetAbolitionAssistant": "助教撤销/作废记录",
+    "/TenantGoods/GetGoodsSalesList": "商品销售记录",
+    "/Site/GetSiteTableUseDetails": "团购核销记录",
+    "/Site/GetSiteTableOrderDetails": "台费订单明细",
+    "/Site/GetTaiFeeAdjustList": "台费调整/优惠记录",
+    "/GoodsStockManage/QueryGoodsOutboundReceipt": "出库单/出库记录",
+    "/Promotion/GetOfflineCouponConsumePageList": "平台券核销记录",
+    "/Order/GetRefundPayLogList": "退款记录",
+    "/Site/GetAllOrderSettleList": "结账记录",
+    "/Site/GetRechargeSettleList": "充值结算记录",
+    "/PayLog/GetPayLogListPage": "支付记录",
+}
+
+
+@dataclass
+class EndpointCheckResult:
+    name_zh: str
+    recent_endpoint: str
+    recent_ok: bool
+    former_endpoint: str
+    former_ok: bool
+    has_schema_diff: str  # "是" | "否" | "未知"
+    diff_detail: str
+    recent_records: int | None = None
+    former_records: int | None = None
+    recent_error: str | None = None
+    former_error: str | None = None
+    extracted_list_key: str | None = None
+
+
+def _reconfigure_stdout_utf8():
+    try:
+        sys.stdout.reconfigure(encoding="utf-8")
+    except Exception:
+        pass
+
+
+def derive_former_endpoint(endpoint: str) -> str | None:
+    # backward compatible wrapper: keep local name but delegate to shared router
+    return derive_former_endpoint_shared(endpoint)
+
+
+def _parse_day_start(d: str, tz: ZoneInfo) -> datetime:
+    dt = dtparser.parse(d)
+    if dt.tzinfo is None:
+        dt = dt.replace(tzinfo=tz)
+    else:
+        dt = dt.astimezone(tz)
+    return dt.replace(hour=0, minute=0, second=0, microsecond=0)
+
+
+def _window_from_dates(start_date: str, end_date_inclusive: str, tz: ZoneInfo) -> tuple[datetime, datetime]:
+    start = _parse_day_start(start_date, tz)
+    end_inclusive = _parse_day_start(end_date_inclusive, tz)
+    end_exclusive = end_inclusive + timedelta(days=1)
+    return start, end_exclusive
+
+
+def _build_window_params(
+    window_style: str,
+    store_id: int,
+    window_start: datetime,
+    window_end: datetime,
+    tz: ZoneInfo,
+) -> dict:
+    if window_style == "none":
+        return {}
+    if window_style == "site":
+        return {"siteId": store_id}
+    if window_style == "range":
+        return {
+            "siteId": store_id,
+            "rangeStartTime": TypeParser.format_timestamp(window_start, tz),
+            "rangeEndTime": TypeParser.format_timestamp(window_end, tz),
+        }
+    if window_style == "pay":
+        return {
+            "siteId": store_id,
+            "StartPayTime": TypeParser.format_timestamp(window_start, tz),
+            "EndPayTime": TypeParser.format_timestamp(window_end, tz),
+        }
+    return {
+        "siteId": store_id,
+        "startTime": TypeParser.format_timestamp(window_start, tz),
+        "endTime": TypeParser.format_timestamp(window_end, tz),
+    }
+
+
+def _extract_records(payload: dict, spec: EndpointSpec) -> tuple[list, str | None]:
+    # 优先使用 spec.list_key；若拿不到数据，再尝试自动推断（None）。
+    records_primary = APIClient._extract_list(payload, spec.data_path, spec.list_key)
+    if records_primary:
+        return records_primary, spec.list_key
+
+    records_fallback = APIClient._extract_list(payload, spec.data_path, None)
+    if records_fallback:
+        return records_fallback, None
+
+    return [], spec.list_key
+
+
+def _walk_paths(obj: Any, prefix: str, out: set[str], max_depth: int, depth: int, sample_list_elems: int):
+    if depth > max_depth:
+        return
+    if isinstance(obj, dict):
+        for k, v in obj.items():
+            if not isinstance(k, str):
+                k = str(k)
+            p = f"{prefix}.{k}" if prefix else k
+            out.add(p)
+            _walk_paths(v, p, out, max_depth, depth + 1, sample_list_elems)
+    elif isinstance(obj, list):
+        p = f"{prefix}[]" if prefix else "[]"
+        out.add(p)
+        for v in obj[:sample_list_elems]:
+            _walk_paths(v, p, out, max_depth, depth + 1, sample_list_elems)
+
+
+def _schema_from_records(records: list, max_records: int, max_depth: int) -> set[str]:
+    paths: set[str] = set()
+    for rec in (records or [])[:max_records]:
+        _walk_paths(rec, "", paths, max_depth=max_depth, depth=0, sample_list_elems=5)
+    return paths
+
+
+def _schema_from_data(payload: dict, data_path: tuple[str, ...], max_depth: int) -> set[str]:
+    cur: Any = payload
+    for k in data_path:
+        if isinstance(cur, dict):
+            cur = cur.get(k)
+        else:
+            cur = None
+        if cur is None:
+            break
+    paths: set[str] = set()
+    _walk_paths(cur, "", paths, max_depth=max_depth, depth=0, sample_list_elems=5)
+    return paths
+
+
+def _cell(text: str) -> str:
+    # markdown table cell escape
+    s = (text or "").replace("|", "\\|").replace("\n", "<br>")
+    return s
+
+
+def _format_diff(recent_paths: set[str], former_paths: set[str], limit: int = 60) -> tuple[str, str]:
+    if recent_paths == former_paths:
+        return "否", ""
+
+    only_recent = sorted(recent_paths - former_paths)
+    only_former = sorted(former_paths - recent_paths)
+
+    parts: list[str] = []
+    if only_recent:
+        truncated = only_recent[:limit]
+        suffix = "" if len(only_recent) <= limit else f" ...(+{len(only_recent) - limit})"
+        parts.append(f"仅近期({len(only_recent)}): " + ", ".join(truncated) + suffix)
+    if only_former:
+        truncated = only_former[:limit]
+        suffix = "" if len(only_former) <= limit else f" ...(+{len(only_former) - limit})"
+        parts.append(f"仅历史({len(only_former)}): " + ", ".join(truncated) + suffix)
+
+    return "是", "\n".join(parts).strip()
+
+
+def _post_first_page(
+    client: APIClient,
+    endpoint: str,
+    params: dict,
+    page_size: int,
+    spec: EndpointSpec,
+) -> tuple[dict, list]:
+    # 只拉取第 1 页，用于“能否响应”与字段对比
+    payload: dict | None = None
+    records: list = []
+    for _, page_records, _, raw in client.iter_paginated(
+        endpoint=endpoint,
+        params=params,
+        page_size=page_size,
+        data_path=spec.data_path,
+        list_key=spec.list_key,
+        page_end=1,
+    ):
+        payload = raw
+        records = page_records or []
+        break
+    return (payload or {}), (records or [])
+
+
+def _load_specs_for_range_only() -> list[EndpointSpec]:
+    # 以 ODS_JSON_ARCHIVE 的 ENDPOINTS 为准，筛选出“可定义时间范围”的接口
+    specs: list[EndpointSpec] = []
+    for spec in OdsJsonArchiveTask.ENDPOINTS:
+        if spec.window_style in ("start_end", "range", "pay"):
+            specs.append(spec)
+    return specs
+
+
+def main() -> int:
+    _reconfigure_stdout_utf8()
+    ap = argparse.ArgumentParser()
+    ap.add_argument("--recent-start", default="2025-12-01")
+    ap.add_argument("--recent-end", default="2025-12-15")
+    ap.add_argument("--former-start", default="2025-08-01")
+    ap.add_argument("--former-end", default="2025-08-15")
+    ap.add_argument("--page-size", type=int, default=50)
+    ap.add_argument("--max-records", type=int, default=50)
+    ap.add_argument("--max-depth", type=int, default=5)
+    ap.add_argument(
+        "--out",
+        default=str(Path(__file__).with_name("recent_vs_former_report.md")),
+        help="输出 markdown 路径",
+    )
+    ap.add_argument(
+        "--out-json",
+        default=str(Path(__file__).with_name("recent_vs_former_report.json")),
+        help="输出 json 路径（含错误信息与统计）",
+    )
+    args = ap.parse_args()
+
+    cfg = AppConfig.load().config
+    tz = ZoneInfo(cfg["app"]["timezone"])
+    store_id = int(cfg["app"]["store_id"])
+
+    recent_start, recent_end = _window_from_dates(args.recent_start, args.recent_end, tz)
+    former_start, former_end = _window_from_dates(args.former_start, args.former_end, tz)
+
+    if not cfg["api"].get("token"):
+        raise SystemExit("缺少 api.token（请在 .env 配置 API_TOKEN 或 FICOO_TOKEN）")
+
+    client = APIClient(
+        base_url=cfg["api"]["base_url"],
+        token=cfg["api"]["token"],
+        timeout=int(cfg["api"].get("timeout_sec") or 20),
+        retry_max=int(cfg["api"].get("retries", {}).get("max_attempts") or 3),
+        headers_extra=cfg["api"].get("headers_extra") or {},
+    )
+
+    common_params = cfg["api"].get("params", {}) or {}
+    if not isinstance(common_params, dict):
+        common_params = {}
+
+    results: list[EndpointCheckResult] = []
+    specs = _load_specs_for_range_only()
+
+    for spec in specs:
+        name_zh = CHINESE_NAMES.get(spec.endpoint) or Path(spec.endpoint).name
+        former_endpoint = derive_former_endpoint(spec.endpoint)
+
+        # recent
+        recent_params = dict(common_params)
+        recent_params.update(_build_window_params(spec.window_style, store_id, recent_start, recent_end, tz))
+
+        recent_ok = False
+        recent_records: list = []
+        recent_payload: dict = {}
+        recent_err: str | None = None
+
+        try:
+            recent_payload, recent_records = _post_first_page(
+                client=client,
+                endpoint=spec.endpoint,
+                params=recent_params,
+                page_size=args.page_size,
+                spec=spec,
+            )
+            recent_ok = True
+        except Exception as e:
+            recent_err = f"{type(e).__name__}: {e}"
+
+        # former
+        former_params = dict(common_params)
+        former_params.update(_build_window_params(spec.window_style, store_id, former_start, former_end, tz))
+
+        former_ok = False
+        former_records: list = []
+        former_payload: dict = {}
+        former_err: str | None = None
+
+        if not former_endpoint:
+            former_err = "未提供历史记录接口 path"
+        else:
+            try:
+                former_payload, former_records = _post_first_page(
+                    client=client,
+                    endpoint=former_endpoint,
+                    params=former_params,
+                    page_size=args.page_size,
+                    spec=spec,
+                )
+                former_ok = True
+            except Exception as e:
+                former_err = f"{type(e).__name__}: {e}"
+
+        extracted_key: str | None = spec.list_key
+        if recent_ok and former_ok:
+            # 用“更能提取出 records 的方式”来做字段对比
+            recent_extracted, recent_key_used = _extract_records(recent_payload, spec)
+            former_extracted, former_key_used = _extract_records(former_payload, spec)
+            extracted_key = recent_key_used or former_key_used or spec.list_key
+
+            if recent_extracted and former_extracted:
+                recent_schema = _schema_from_records(recent_extracted, args.max_records, args.max_depth)
+                former_schema = _schema_from_records(former_extracted, args.max_records, args.max_depth)
+                has_diff, detail = _format_diff(recent_schema, former_schema)
+            elif (not recent_extracted) and (not former_extracted):
+                has_diff, detail = "未知", "两侧 records 均为空，无法判断字段差异"
+            else:
+                has_diff, detail = (
+                    "未知",
+                    f"一侧 records 为空（近期={len(recent_extracted)} 历史={len(former_extracted)}），无法判断字段差异",
+                )
+        else:
+            if former_endpoint is None:
+                has_diff, detail = "未知", "无历史记录接口，跳过字段对比"
+            else:
+                has_diff, detail = "未知", "请求失败，无法对比字段"
+
+        results.append(
+            EndpointCheckResult(
+                name_zh=name_zh,
+                recent_endpoint=spec.endpoint,
+                recent_ok=recent_ok,
+                former_endpoint=former_endpoint or "无",
+                former_ok=former_ok,
+                has_schema_diff=has_diff,
+                diff_detail=detail,
+                recent_records=len(recent_records) if recent_ok else None,
+                former_records=len(former_records) if former_ok else None,
+                recent_error=recent_err,
+                former_error=former_err,
+                extracted_list_key=extracted_key,
+            )
+        )
+
+    # markdown report
+    out_md = Path(args.out)
+    out_json = Path(args.out_json)
+    out_md.parent.mkdir(parents=True, exist_ok=True)
+
+    header = [
+        "# 近期记录 vs 历史记录(Former) 接口对比报告",
+        "",
+        f"- 近期窗口: `{recent_start.isoformat()}` ~ `{recent_end.isoformat()}`（end 为次日 00:00:00）",
+        f"- 历史窗口: `{former_start.isoformat()}` ~ `{former_end.isoformat()}`（end 为次日 00:00:00）",
+        f"- store_id: `{store_id}`",
+        f"- base_url: `{cfg['api']['base_url']}`",
+        "",
+        "表头：接口名称（中文）；近期记录接口 path；近期记录是否返回；历史记录接口 path；历史记录是否返回；是否存在返回字段差异；差异字段详情。",
+        "",
+        "| 接口名称（中文） | 近期记录接口 path | 近期记录是否返回 | 历史记录接口 path | 历史记录是否返回 | 是否存在返回字段差异 | 差异字段详情 |",
+        "| --- | --- | --- | --- | --- | --- | --- |",
+    ]
+
+    rows: list[str] = []
+    for r in results:
+        rows.append(
+            "| "
+            + " | ".join(
+                [
+                    _cell(r.name_zh),
+                    _cell(r.recent_endpoint),
+                    "是" if r.recent_ok else "否",
+                    _cell(r.former_endpoint),
+                    "是" if r.former_ok else "否",
+                    _cell(r.has_schema_diff),
+                    _cell(r.diff_detail or ""),
+                ]
+            )
+            + " |"
+        )
+
+    out_md.write_text("\n".join(header + rows) + "\n", encoding="utf-8")
+
+    out_json.write_text(
+        json.dumps(
+            {
+                "recent_window": {"start": recent_start.isoformat(), "end": recent_end.isoformat()},
+                "former_window": {"start": former_start.isoformat(), "end": former_end.isoformat()},
+                "store_id": store_id,
+                "base_url": cfg["api"]["base_url"],
+                "results": [asdict(r) for r in results],
+            },
+            ensure_ascii=False,
+            indent=2,
+        )
+        + "\n",
+        encoding="utf-8",
+    )
+
+    print(f"OK: wrote {out_md}")
+    print(f"OK: wrote {out_json}")
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())

etl_billiards/fetch-test/recent_vs_former_report.json

@@ -0,0 +1,196 @@
+{
+  "recent_window": {
+    "start": "2025-12-01T00:00:00+08:00",
+    "end": "2025-12-16T00:00:00+08:00"
+  },
+  "former_window": {
+    "start": "2025-08-01T00:00:00+08:00",
+    "end": "2025-08-16T00:00:00+08:00"
+  },
+  "store_id": 2790685415443269,
+  "base_url": "https://pc.ficoo.vip/apiprod/admin/v1/",
+  "results": [
+    {
+      "name_zh": "会员余额变动",
+      "recent_endpoint": "/MemberProfile/GetMemberCardBalanceChange",
+      "recent_ok": true,
+      "former_endpoint": "/MemberProfile/GetFormerMemberCardBalanceChange",
+      "former_ok": true,
+      "has_schema_diff": "否",
+      "diff_detail": "",
+      "recent_records": 50,
+      "former_records": 50,
+      "recent_error": null,
+      "former_error": null,
+      "extracted_list_key": null
+    },
+    {
+      "name_zh": "助教服务记录",
+      "recent_endpoint": "/AssistantPerformance/GetOrderAssistantDetails",
+      "recent_ok": true,
+      "former_endpoint": "/AssistantPerformance/GetFormerOrderAssistantDetails",
+      "former_ok": true,
+      "has_schema_diff": "否",
+      "diff_detail": "",
+      "recent_records": 50,
+      "former_records": 50,
+      "recent_error": null,
+      "former_error": null,
+      "extracted_list_key": "orderAssistantDetails"
+    },
+    {
+      "name_zh": "助教撤销/作废记录",
+      "recent_endpoint": "/AssistantPerformance/GetAbolitionAssistant",
+      "recent_ok": true,
+      "former_endpoint": "无",
+      "former_ok": false,
+      "has_schema_diff": "未知",
+      "diff_detail": "无历史记录接口，跳过字段对比",
+      "recent_records": 18,
+      "former_records": null,
+      "recent_error": null,
+      "former_error": "未提供历史记录接口 path",
+      "extracted_list_key": "abolitionAssistants"
+    },
+    {
+      "name_zh": "商品销售记录",
+      "recent_endpoint": "/TenantGoods/GetGoodsSalesList",
+      "recent_ok": true,
+      "former_endpoint": "/TenantGoods/GetFormerGoodsSalesList",
+      "former_ok": true,
+      "has_schema_diff": "否",
+      "diff_detail": "",
+      "recent_records": 50,
+      "former_records": 50,
+      "recent_error": null,
+      "former_error": null,
+      "extracted_list_key": "orderGoodsLedgers"
+    },
+    {
+      "name_zh": "团购核销记录",
+      "recent_endpoint": "/Site/GetSiteTableUseDetails",
+      "recent_ok": true,
+      "former_endpoint": "/Site/GetSiteTableUseDetails",
+      "former_ok": true,
+      "has_schema_diff": "未知",
+      "diff_detail": "一侧 records 为空（近期=50 历史=0），无法判断字段差异",
+      "recent_records": 50,
+      "former_records": 0,
+      "recent_error": null,
+      "former_error": null,
+      "extracted_list_key": "siteTableUseDetailsList"
+    },
+    {
+      "name_zh": "台费订单明细",
+      "recent_endpoint": "/Site/GetSiteTableOrderDetails",
+      "recent_ok": true,
+      "former_endpoint": "/Site/GetFormerSiteTableOrderDetails",
+      "former_ok": true,
+      "has_schema_diff": "否",
+      "diff_detail": "",
+      "recent_records": 50,
+      "former_records": 50,
+      "recent_error": null,
+      "former_error": null,
+      "extracted_list_key": "siteTableUseDetailsList"
+    },
+    {
+      "name_zh": "台费调整/优惠记录",
+      "recent_endpoint": "/Site/GetTaiFeeAdjustList",
+      "recent_ok": true,
+      "former_endpoint": "/Site/GetFormerTaiFeeAdjustList",
+      "former_ok": true,
+      "has_schema_diff": "否",
+      "diff_detail": "",
+      "recent_records": 50,
+      "former_records": 50,
+      "recent_error": null,
+      "former_error": null,
+      "extracted_list_key": "taiFeeAdjustInfos"
+    },
+    {
+      "name_zh": "出库单/出库记录",
+      "recent_endpoint": "/GoodsStockManage/QueryGoodsOutboundReceipt",
+      "recent_ok": true,
+      "former_endpoint": "/GoodsStockManage/QueryFormerGoodsOutboundReceipt",
+      "former_ok": true,
+      "has_schema_diff": "否",
+      "diff_detail": "",
+      "recent_records": 50,
+      "former_records": 50,
+      "recent_error": null,
+      "former_error": null,
+      "extracted_list_key": "queryDeliveryRecordsList"
+    },
+    {
+      "name_zh": "平台券核销记录",
+      "recent_endpoint": "/Promotion/GetOfflineCouponConsumePageList",
+      "recent_ok": true,
+      "former_endpoint": "/Promotion/GetOfflineCouponConsumePageList",
+      "former_ok": true,
+      "has_schema_diff": "否",
+      "diff_detail": "",
+      "recent_records": 50,
+      "former_records": 50,
+      "recent_error": null,
+      "former_error": null,
+      "extracted_list_key": null
+    },
+    {
+      "name_zh": "退款记录",
+      "recent_endpoint": "/Order/GetRefundPayLogList",
+      "recent_ok": true,
+      "former_endpoint": "无",
+      "former_ok": false,
+      "has_schema_diff": "未知",
+      "diff_detail": "无历史记录接口，跳过字段对比",
+      "recent_records": 5,
+      "former_records": null,
+      "recent_error": null,
+      "former_error": "未提供历史记录接口 path",
+      "extracted_list_key": null
+    },
+    {
+      "name_zh": "结账记录",
+      "recent_endpoint": "/Site/GetAllOrderSettleList",
+      "recent_ok": true,
+      "former_endpoint": "/Site/GetFormerOrderSettleList",
+      "former_ok": true,
+      "has_schema_diff": "否",
+      "diff_detail": "",
+      "recent_records": 50,
+      "former_records": 50,
+      "recent_error": null,
+      "former_error": null,
+      "extracted_list_key": "settleList"
+    },
+    {
+      "name_zh": "充值结算记录",
+      "recent_endpoint": "/Site/GetRechargeSettleList",
+      "recent_ok": true,
+      "former_endpoint": "/Site/GetFormerRechargeSettleList",
+      "former_ok": true,
+      "has_schema_diff": "否",
+      "diff_detail": "",
+      "recent_records": 39,
+      "former_records": 47,
+      "recent_error": null,
+      "former_error": null,
+      "extracted_list_key": "settleList"
+    },
+    {
+      "name_zh": "支付记录",
+      "recent_endpoint": "/PayLog/GetPayLogListPage",
+      "recent_ok": true,
+      "former_endpoint": "/PayLog/GetFormerPayLogListPage",
+      "former_ok": true,
+      "has_schema_diff": "否",
+      "diff_detail": "",
+      "recent_records": 50,
+      "former_records": 50,
+      "recent_error": null,
+      "former_error": null,
+      "extracted_list_key": null
+    }
+  ]
+}

etl_billiards/fetch-test/recent_vs_former_report.md

@@ -0,0 +1,24 @@
+# 近期记录 vs 历史记录(Former) 接口对比报告
+
+- 近期窗口: `2025-12-01T00:00:00+08:00` ~ `2025-12-16T00:00:00+08:00`（end 为次日 00:00:00）
+- 历史窗口: `2025-08-01T00:00:00+08:00` ~ `2025-08-16T00:00:00+08:00`（end 为次日 00:00:00）
+- store_id: `2790685415443269`
+- base_url: `https://pc.ficoo.vip/apiprod/admin/v1/`
+
+表头：接口名称（中文）；近期记录接口 path；近期记录是否返回；历史记录接口 path；历史记录是否返回；是否存在返回字段差异；差异字段详情。
+
+| 接口名称（中文） | 近期记录接口 path | 近期记录是否返回 | 历史记录接口 path | 历史记录是否返回 | 是否存在返回字段差异 | 差异字段详情 |
+| --- | --- | --- | --- | --- | --- | --- |
+| 会员余额变动 | /MemberProfile/GetMemberCardBalanceChange | 是 | /MemberProfile/GetFormerMemberCardBalanceChange | 是 | 否 |  |
+| 助教服务记录 | /AssistantPerformance/GetOrderAssistantDetails | 是 | /AssistantPerformance/GetFormerOrderAssistantDetails | 是 | 否 |  |
+| 助教撤销/作废记录 | /AssistantPerformance/GetAbolitionAssistant | 是 | 无 | 否 | 未知 | 无历史记录接口，跳过字段对比 |
+| 商品销售记录 | /TenantGoods/GetGoodsSalesList | 是 | /TenantGoods/GetFormerGoodsSalesList | 是 | 否 |  |
+| 团购核销记录 | /Site/GetSiteTableUseDetails | 是 | /Site/GetSiteTableUseDetails | 是 | 未知 | 一侧 records 为空（近期=50 历史=0），无法判断字段差异 |
+| 台费订单明细 | /Site/GetSiteTableOrderDetails | 是 | /Site/GetFormerSiteTableOrderDetails | 是 | 否 |  |
+| 台费调整/优惠记录 | /Site/GetTaiFeeAdjustList | 是 | /Site/GetFormerTaiFeeAdjustList | 是 | 否 |  |
+| 出库单/出库记录 | /GoodsStockManage/QueryGoodsOutboundReceipt | 是 | /GoodsStockManage/QueryFormerGoodsOutboundReceipt | 是 | 否 |  |
+| 平台券核销记录 | /Promotion/GetOfflineCouponConsumePageList | 是 | /Promotion/GetOfflineCouponConsumePageList | 是 | 否 |  |
+| 退款记录 | /Order/GetRefundPayLogList | 是 | 无 | 否 | 未知 | 无历史记录接口，跳过字段对比 |
+| 结账记录 | /Site/GetAllOrderSettleList | 是 | /Site/GetFormerOrderSettleList | 是 | 否 |  |
+| 充值结算记录 | /Site/GetRechargeSettleList | 是 | /Site/GetFormerRechargeSettleList | 是 | 否 |  |
+| 支付记录 | /PayLog/GetPayLogListPage | 是 | /PayLog/GetFormerPayLogListPage | 是 | 否 |  |