#!/usr/bin/env python3
"""
导出每个角色每节课的完课记录（含耗时、是否首通）
输出：用户ID、角色ID、课程名称、完课耗时(分钟)、是否首通
"""
import os
import sys
import pandas as pd
import psycopg2
from openpyxl import Workbook
from openpyxl.utils.dataframe import dataframe_to_rows

# 数据库连接
conn = psycopg2.connect(
    host=os.environ.get('PG_ONLINE_HOST', 'bj-postgres-16pob4sg.sql.tencentcdb.com'),
    port=int(os.environ.get('PG_ONLINE_PORT', 28591)),
    dbname='vala_bi',
    user='ai_member',
    password=os.environ['PG_ONLINE_PASSWORD']
)

print("开始查询...")

# 构建 8 个分表的 UNION ALL 查询
chapter_unions = " UNION ALL ".join(
    f"SELECT * FROM bi_user_chapter_play_record_{i}" for i in range(8)
)

component_unions = " UNION ALL ".join(
    f"SELECT * FROM bi_user_component_play_record_{i}" for i in range(8)
)

query = f"""
WITH 
-- 1. 聚合组件耗时（按 chapter_unique_id）
component_agg AS (
    SELECT 
        chapter_unique_id,
        SUM(interval_time) AS total_interval_ms
    FROM ({component_unions}) AS comp
    GROUP BY chapter_unique_id
),
-- 2. 有效课时完成记录（关联有效账号 + 角色 + 课程）
chapter_data AS (
    SELECT 
        cpr.id AS record_id,
        cpr.user_id AS character_id,
        chr.account_id,
        cpr.chapter_id,
        cpr.chapter_unique_id,
        cpr.created_at,
        lul.course_level || ' ' || lul.course_season || ' ' || lul.course_unit || ' ' || lul.course_lesson AS course_name
    FROM ({chapter_unions}) AS cpr
    INNER JOIN bi_vala_app_character chr ON cpr.user_id = chr.id
    INNER JOIN bi_vala_app_account acc ON chr.account_id = acc.id 
        AND acc.status = 1 
        AND acc.deleted_at IS NULL
    INNER JOIN bi_level_unit_lesson lul ON cpr.chapter_id = lul.id
    WHERE cpr.play_status = 1
),
-- 3. 关联耗时并标记首通
final AS (
    SELECT 
        cd.account_id,
        cd.character_id,
        cd.course_name,
        ROUND(COALESCE(ca.total_interval_ms, 0) / 60000.0, 1) AS duration_min,
        CASE 
            WHEN cd.created_at = MIN(cd.created_at) OVER (
                PARTITION BY cd.character_id, cd.chapter_id
            ) THEN '是'
            ELSE '否'
        END AS is_first_complete
    FROM chapter_data cd
    LEFT JOIN component_agg ca ON cd.chapter_unique_id = ca.chapter_unique_id
)
SELECT 
    account_id,
    character_id,
    course_name,
    duration_min,
    is_first_complete
FROM final
ORDER BY account_id, character_id, course_name, is_first_complete DESC, duration_min
"""

print("执行 SQL 查询...")
df = pd.read_sql_query(query, conn)
conn.close()

print(f"查询完成，共 {len(df)} 条记录")
print(f"列：{list(df.columns)}")
print(f"前5行预览：\n{df.head()}")

# 导出 Excel
output_path = '/root/.openclaw/workspace/output/chapter_completion_all.xlsx'
print(f"正在导出到 {output_path}...")

# 重命名列
df.columns = ['用户ID', '角色ID', '课程名称', '完课耗时(分钟)', '是否首通']

# 使用 openpyxl 写入（更好的格式）
wb = Workbook()
ws = wb.active
ws.title = '完课记录'

# 写入表头
for col_idx, col_name in enumerate(df.columns, 1):
    ws.cell(row=1, column=col_idx, value=col_name)

# 批量写入数据
for row_idx, row in enumerate(df.itertuples(index=False), 2):
    ws.cell(row=row_idx, column=1, value=row[0])  # 用户ID
    ws.cell(row=row_idx, column=2, value=row[1])  # 角色ID
    ws.cell(row=row_idx, column=3, value=row[2])  # 课程名称
    ws.cell(row=row_idx, column=4, value=row[3])  # 完课耗时
    ws.cell(row=row_idx, column=5, value=row[4])  # 是否首通

# 调整列宽
ws.column_dimensions['A'].width = 12
ws.column_dimensions['B'].width = 12
ws.column_dimensions['C'].width = 22
ws.column_dimensions['D'].width = 18
ws.column_dimensions['E'].width = 12

wb.save(output_path)
print(f"✅ 导出完成！文件: {output_path}")
print(f"   总记录数: {len(df)}")
print(f"   首通记录: {len(df[df['是否首通'] == '是'])}")
print(f"   非首通记录: {len(df[df['是否首通'] == '否'])}")

# 文件大小
size_mb = os.path.getsize(output_path) / (1024 * 1024)
print(f"   文件大小: {size_mb:.1f} MB")