#!/usr/bin/env python3
"""B2: 动态跟踪表 → 中东项目启动跟踪看板 (2026-06-08)"""
import pandas as pd, warnings, json
from pathlib import Path
warnings.filterwarnings('ignore')

import sys
DATA_DATE = sys.argv[1] if len(sys.argv) > 1 else '2026-06-08'
RAW_DIR = Path(f'/mnt/y/Openclaw_Hub/03.资源/实施项目 wiki/raw/月报数据/{DATA_DATE}')
OUT_DIR = Path(f'/mnt/y/Openclaw_Hub/03.资源/实施项目 wiki/dashboard/data/{DATA_DATE}/cleaned')
OUT_DIR.mkdir(parents=True, exist_ok=True)

INPUT_FILE = RAW_DIR / '动态跟踪表_REPORT3.xlsx'
df = pd.read_excel(INPUT_FILE, sheet_name='REPORT3', header=None)

# Row 0 = title, Row 1 = 驻外机构, Row 2 = project names, Row 3+ = tasks
# Find Middle East columns
me_cols = []  # (start_col, region_name, project_name)
current_region = None
current_project = None

for c in range(0, df.shape[1], 3):
    region = str(df.iloc[1, c]) if c < df.shape[1] and pd.notna(df.iloc[1, c]) else ''
    project = str(df.iloc[2, c]) if c < df.shape[1] and pd.notna(df.iloc[2, c]) else ''
    
    if region and 'nan' not in region:
        current_region = region
    if project and project != 'nan' and '项目名称' not in project:
        current_project = project
    
    if current_region and '中东' in str(current_region):
        if current_project and current_project != 'nan':
            me_cols.append((c, current_region, current_project))

print(f'Found {len(me_cols)} Middle East project columns')

# Build task rows for each project
all_tasks = []
TASK_ROWS_START = 3  # Row index where task data starts (0-based)

for start_col, region, project in me_cols:
    for row_idx in range(TASK_ROWS_START, df.shape[0]):
        task = str(df.iloc[row_idx, start_col + 0]) if pd.notna(df.iloc[row_idx, start_col + 0]) else ''
        plan_date = df.iloc[row_idx, start_col + 1] if start_col + 1 < df.shape[1] else None
        gap = df.iloc[row_idx, start_col + 2] if start_col + 2 < df.shape[1] else None
        
        if task and task != 'nan' and task.strip():
            all_tasks.append({
                '项目名称': project,
                '区域': region,
                '工作任务': task.strip(),
                '计划完成日期': plan_date if pd.notna(plan_date) else '',
                '差距天数': gap if pd.notna(gap) else '',
            })

df_tasks = pd.DataFrame(all_tasks)
print(f'Extracted {len(df_tasks)} task rows across {df_tasks["项目名称"].nunique()} projects')

# Analyze completion status
df_tasks['差距数值'] = pd.to_numeric(df_tasks['差距天数'], errors='coerce')
df_tasks['计划日期_parsed'] = pd.to_datetime(df_tasks['计划完成日期'], errors='coerce')

# Project-level summary
proj_summary = df_tasks.groupby('项目名称').agg(
    总任务数=('工作任务', 'count'),
    已完成=('差距天数', lambda x: (x.astype(str).str.strip() == '0').sum()),
    逾期=('差距数值', lambda x: (x > 0).sum()),
    未开始=('差距天数', lambda x: (x.astype(str).str.strip().isin(['', 'nan', '不涉及', '/', 'None'])).sum()),
    不涉及=('差距天数', lambda x: (x.astype(str).str.contains('不涉及')).sum()),
).reset_index()

proj_summary['完成率'] = (proj_summary['已完成'] / proj_summary['总任务数'] * 100).round(0).astype(int).astype(str) + '%'
proj_summary = proj_summary.sort_values('逾期', ascending=False)

print(f'\n📊 项目启动跟踪汇总:')
print(f'  项目数: {len(proj_summary)}')
print(f'  总任务数: {proj_summary["总任务数"].sum()}')
print(f'  已完成: {proj_summary["已完成"].sum()}')
print(f'  逾期: {proj_summary["逾期"].sum()}')
print(f'\n逾期项目:')
print(proj_summary[proj_summary['逾期'] > 0].head(15).to_string())

# Output
df_tasks.to_parquet(OUT_DIR / 'tracking_tasks.parquet', index=False)
df_tasks.to_csv(OUT_DIR / 'tracking_tasks.csv', index=False, encoding='utf-8-sig')
proj_summary.to_csv(OUT_DIR / 'tracking_project_summary.csv', index=False, encoding='utf-8-sig')

report = {
    'source': str(INPUT_FILE),
    'me_projects': len(proj_summary),
    'total_tasks': int(proj_summary['总任务数'].sum()),
    'completed': int(proj_summary['已完成'].sum()),
    'overdue': int(proj_summary['逾期'].sum()),
}
with open(OUT_DIR / 'tracking_validation.json', 'w', encoding='utf-8') as f:
    json.dump(report, f, ensure_ascii=False, indent=2)

print(f'\n✅ tracking_tasks.parquet + project_summary')
print(f'\n完整项目列表:')
print(proj_summary[['项目名称','总任务数','已完成','逾期','完成率']].to_string())