# 文件路径：scripts/orchestrator.py
# 多 Skill 协作的编排入口脚本

import subprocess
import sys
import json
import os
import shutil

SKILLS_BASE  = "/mnt/skills/public"
WORK_DIR     = "/home/claude/report_work"
OUTPUT_DIR   = "/mnt/user-data/outputs"

def run(script_path: str, *args) -> dict:
    """运行指定脚本，返回 JSON 结果"""
    cmd    = [sys.executable, script_path] + list(args)
    result = subprocess.run(cmd, capture_output=True, text=True, timeout=120)
    try:
        return json.loads(result.stdout)
    except json.JSONDecodeError:
        return {
            "status":  "error",
            "message": result.stderr or result.stdout or "脚本无输出"
        }

def orchestrate(input_file: str) -> dict:
    """执行完整的报告生成工作流"""
    os.makedirs(WORK_DIR, exist_ok=True)

    # ── 阶段一：数据清洗 ─────────────────────────────────
    print("【1/3】正在清洗数据...", flush=True)
    clean_result = run(
        f"{SKILLS_BASE}/data-cleaner/scripts/clean_data.py",
        input_file,
        os.path.join(WORK_DIR, "cleaned.csv")
    )
    if clean_result.get("status") != "success":
        return {"status": "error", "stage": "数据清洗",
                "message": clean_result.get("message")}

    print(f"  已清洗：删除重复行 {clean_result.get('dup_removed', 0)} 条，"
          f"填充空值 {clean_result.get('nulls_filled', 0)} 处", flush=True)

    # ── 阶段二：统计分析 ─────────────────────────────────
    print("【2/3】正在生成统计报告...", flush=True)
    stats_result = run(
        f"{SKILLS_BASE}/data-cleaner/scripts/calc_stats.py",
        clean_result["output"],
        os.path.join(WORK_DIR, "stats.json")
    )
    if stats_result.get("status") != "success":
        return {"status": "error", "stage": "统计分析",
                "message": stats_result.get("message")}

    # ── 阶段三：生成 Excel 报告 ──────────────────────────
    print("【3/3】正在生成 Excel 报告...", flush=True)
    output_name  = os.path.splitext(os.path.basename(input_file))[0] + "_report.xlsx"
    output_path  = os.path.join(OUTPUT_DIR, output_name)

    report_result = run(
        f"{SKILLS_BASE}/data-cleaner/scripts/gen_report.py",
        clean_result["output"],
        stats_result.get("output", ""),
        output_path
    )
    if report_result.get("status") != "success":
        return {"status": "error", "stage": "报告生成",
                "message": report_result.get("message")}

    # 清理工作目录
    shutil.rmtree(WORK_DIR, ignore_errors=True)

    return {
        "status":     "success",
        "output":     output_path,
        "clean_stats": {
            "dup_removed":  clean_result.get("dup_removed", 0),
            "nulls_filled": clean_result.get("nulls_filled", 0)
        }
    }

if __name__ == "__main__":
    if len(sys.argv) < 2:
        print(json.dumps({"status": "error", "message": "用法：python orchestrator.py <文件路径>"}))
        sys.exit(1)

    result = orchestrate(sys.argv[1])
    print(json.dumps(result, ensure_ascii=False, indent=2))