#!/usr/bin/env python3
"""
课程学生表现分析脚本
用法:
  python scripts/analyze_course_student.py \\
      --course-id 6975983771e15346c9e8fdc0 \\
      --uid 248 --name 王梓骏 --lessons 8
"""

import io
import json
import os
import re
import sys
from collections import defaultdict
from datetime import datetime
from html import unescape
from pathlib import Path

import httpx

sys.stdout = io.TextIOWrapper(sys.stdout.buffer, encoding="utf-8", errors="replace")
sys.stderr = io.TextIOWrapper(sys.stderr.buffer, encoding="utf-8", errors="replace")

PROJECT_ROOT = Path(__file__).parent.parent


def load_env(env_path: Path) -> dict:
    d = {}
    if env_path.exists():
        for line in env_path.read_text(encoding="utf-8").split("\n"):
            line = line.strip()
            if line and not line.startswith("#") and "=" in line:
                k, v = line.split("=", 1)
                d[k.strip()] = v.strip()
    return d


env = load_env(PROJECT_ROOT / ".env")
OJ_BASE_URL = os.environ.get("OJ_BASE_URL") or env.get(
    "OJ_BASE_URL", "https://oj.qonnwolf.com"
)
USERNAME = os.environ.get("OJ_USERNAME") or env.get("OJ_USERNAME", "")
PASSWORD = os.environ.get("OJ_PASSWORD") or env.get("OJ_PASSWORD", "")

STATUS_MAP = {
    "Accepted": "AC",
    "Wrong Answer": "WA",
    "Compile Error": "CE",
    "Time Limit Exceeded": "TLE",
    "Memory Limit Exceeded": "MLE",
    "Runtime Error": "RE",
    "Presentation Error": "PE",
    "Output Limit Exceeded": "OLE",
    "Runtime Error on Test": "RE",
    "Partial Accepted": "PA",
}

# 作业名称 → 类型
HW_TYPE = {
    "课堂练习": "A",
    "课后作业": "B",
    "拓展练习": "C",
}


# ──────────────────────────────────────────
# 基础工具
# ──────────────────────────────────────────


def strip_tags(html: str) -> str:
    return unescape(re.sub(r"<[^>]+>", "", html)).replace("\xa0", " ").strip()


def detect_status(row_html: str) -> str:
    for kw, st in STATUS_MAP.items():
        if kw in row_html:
            return st
    return "UNKNOWN"


def normalize_time(raw: str) -> str:
    m = re.search(r"(\d{4})-(\d{1,2})-(\d{1,2})\s+(\d{1,2}):(\d{2}):(\d{2})", raw)
    if not m:
        return ""
    y, mo, d, h, mi, s = m.groups()
    return f"{y}-{int(mo):02d}-{int(d):02d}T{int(h):02d}:{mi}:{s}"


# ──────────────────────────────────────────
# 登录
# ──────────────────────────────────────────


def login(client: httpx.Client) -> bool:
    try:
        r = client.post("/login", json={"uname": USERNAME, "password": PASSWORD})
        r.raise_for_status()
        ok = any(c.name == "sid" for c in client.cookies.jar)
        print(f"  {'[OK] 登录成功' if ok else '[X] 登录失败'}: {USERNAME}")
        return ok
    except Exception as e:
        print(f"  [X] 登录异常: {e}")
        return False


# ──────────────────────────────────────────
# 课程结构抓取（解析 sectionHWMap）
# ──────────────────────────────────────────


def fetch_course_structure(client: httpx.Client, course_id: str) -> list[dict]:
    """
    从课程页面解析所有节次和作业 ID。
    返回: [{"title": "知识回顾", "homeworks": [{"hw_id": "...", "type": "A/B/C", "name": "..."}]}, ...]
    按页面顺序排列，跳过 homeworks 为空的节次。
    """
    r = client.get(f"/course/{course_id}")
    html = r.text

    # 1) 提取节次标题（按 HTML 顺序）
    #    openGlobalPanel('<section_id>', '<title>')
    panel_calls = re.findall(
        r"openGlobalPanel\('([A-Fa-f0-9]{24})',\s*'([^']+)'\)", html
    )
    section_titles: dict[str, str] = {}  # id → title
    section_order: list[str] = []  # ordered section ids
    for sid, title in panel_calls:
        if sid not in section_titles:
            section_titles[sid] = title
            section_order.append(sid)

    # 2) 提取 sectionHWMap
    m = re.search(r"sectionHWMap:\s*(\{.*?\}),\s*\n\s*open", html, re.DOTALL)
    if not m:
        print("  [X] 未找到 sectionHWMap")
        return []

    raw_map = m.group(1)
    section_hws: dict[str, list[dict]] = {}  # section_id → [{hw_id, name}]
    for sid_m in re.finditer(r'"([A-Fa-f0-9]{24})":\s*\[(.*?)\]', raw_map, re.DOTALL):
        sid = sid_m.group(1)
        entries = re.findall(
            r'\{\s*id:\s*"([^"]+)",\s*name:\s*"([^"]+)"', sid_m.group(2)
        )
        section_hws[sid] = [{"hw_id": hw_id, "name": name} for hw_id, name in entries]

    # 3) 按顺序组合
    lessons = []
    for sid in section_order:
        hws_raw = section_hws.get(sid, [])
        if not hws_raw:
            continue  # 该节次无作业，跳过
        homeworks = []
        for hw in hws_raw:
            # 去掉 emoji，识别类型
            name_clean = re.sub(r"[^\w\s]", "", hw["name"]).strip()
            hw_type = "other"
            for key, t in HW_TYPE.items():
                if key in hw["name"]:
                    hw_type = t
                    break
            homeworks.append(
                {"hw_id": hw["hw_id"], "type": hw_type, "name": hw["name"]}
            )
        lessons.append(
            {
                "section_id": sid,
                "title": section_titles.get(sid, sid),
                "homeworks": homeworks,
            }
        )

    return lessons


# ──────────────────────────────────────────
# 提交记录抓取
# ──────────────────────────────────────────


def parse_record_rows(html: str) -> list[dict]:
    rows = re.findall(r"<tr[^>]*>(.*?)</tr>", html, re.IGNORECASE | re.DOTALL)
    records = []
    for i, row in enumerate(rows, 1):
        if "/user/" not in row or "/p/" not in row:
            continue
        user_m = re.search(r'href="/user/(\d+)"[^>]*>(.*?)</a>', row, re.DOTALL)
        if not user_m:
            continue
        prob_m = re.search(
            r'href="/p/([^"?/]+)(?:\?[^"]*)?"[^>]*>(.*?)</a>', row, re.DOTALL
        )
        rec_id_m = re.search(r"/record/([A-Za-z0-9]+)", row)
        time_m = re.search(r"(\d{4}-\d{1,2}-\d{1,2}\s+\d{1,2}:\d{2}:\d{2})", row)
        records.append(
            {
                "id": rec_id_m.group(1) if rec_id_m else str(i),
                "uid": int(user_m.group(1)),
                "uname": strip_tags(user_m.group(2)),
                "pid": strip_tags(prob_m.group(1)) if prob_m else "",
                "status": detect_status(row),
                "time": normalize_time(time_m.group(1)) if time_m else "",
            }
        )
    return records


def fetch_hw_official_pids(client: httpx.Client, hw_id: str) -> list[str]:
    """
    从作业页面读取官方题目列表，作为分母。
    避免用提交记录反推题目导致计数偏差。
    """
    r = client.get(f"/homework/{hw_id}")
    pids = re.findall(r'href="/p/([A-Za-z0-9]+)"', r.text)
    return list(dict.fromkeys(pids))  # 去重保序


def fetch_hw_records(
    client: httpx.Client, hw_id: str, max_pages: int = 30
) -> list[dict]:
    records, seen = [], set()
    for page in range(1, max_pages + 1):
        r = client.get(f"/record?tid={hw_id}&page={page}")
        if r.status_code != 200:
            break
        rows = parse_record_rows(r.text)
        if not rows:
            break
        new = 0
        for rec in rows:
            if rec["id"] not in seen:
                seen.add(rec["id"])
                records.append(rec)
                new += 1
        if new == 0:
            break
        if f"page={page + 1}" not in r.text:
            break
    records.sort(key=lambda x: x.get("time", ""))
    return records


# ──────────────────────────────────────────
# 分析逻辑
# ──────────────────────────────────────────


def pattern_label(attempts: int, solved: bool) -> str:
    if attempts == 0:
        return "未提交"
    if not solved:
        return "尝试未通过"
    if attempts == 1:
        return "一气呵成 ⚡"
    if attempts <= 3:
        return "调试改进"
    if attempts <= 6:
        return "多次尝试"
    return "耐心调试"


def analyze_student(uid: int, uname: str, lessons: list[dict]) -> dict:
    """
    以题目为粒度统计完成情况：
    - total[type]["total"]  = 该类型所有作业的总题目数
    - total[type]["done"]   = 学生 AC 的题目数
    - total[type]["attempts"] = 学生该类型总提交次数
    """
    result = {
        "uid": uid,
        "uname": uname,
        "lessons": [],
        "total": {
            "A": {"done": 0, "total": 0, "attempts": 0},
            "B": {"done": 0, "total": 0, "attempts": 0},
            "C": {"done": 0, "total": 0, "attempts": 0},
        },
        "all_errors": defaultdict(int),
    }

    for lesson in lessons:
        ls = {"title": lesson["title"], "problems": [], "present": False}
        for hw in lesson["homeworks"]:
            recs = hw["records"]

            # ── 官方题目列表（从作业页面读取，不从提交记录推断）──
            official_pids = hw.get("official_pids", [])
            all_pids = (
                official_pids
                if official_pids
                else sorted(set(r["pid"] for r in recs if r["pid"]))
            )
            n_total_pids = len(all_pids)

            # ── 该学生的提交 ──
            stu_recs = [r for r in recs if r["uid"] == uid]
            attempts = len(stu_recs)

            # 该学生 AC 的题目（只统计在官方题目列表内的）
            official_pids_set = set(all_pids)
            stu_ac_pids = set(
                r["pid"]
                for r in stu_recs
                if r["status"] == "AC" and r["pid"] and r["pid"] in official_pids_set
            )
            n_solved = len(stu_ac_pids)

            # 错误类型（非 AC 提交）
            errors = [r["status"] for r in stu_recs if r["status"] != "AC"]
            err_counts = defaultdict(int)
            for e in errors:
                err_counts[e] += 1
                result["all_errors"][e] += 1

            t = hw["type"]
            if t in ("A", "B", "C"):
                result["total"][t]["total"] += n_total_pids
                result["total"][t]["attempts"] += attempts
                result["total"][t]["done"] += n_solved

            # 完成情况文字
            if attempts == 0:
                pat = "未提交"
            elif n_solved == n_total_pids:
                pat = f"全部完成 ({n_solved}/{n_total_pids})"
            else:
                pat = f"部分完成 ({n_solved}/{n_total_pids})"

            ls["problems"].append(
                {
                    "name": hw["name"],
                    "type": t,
                    "attempts": attempts,
                    "n_solved": n_solved,
                    "n_total": n_total_pids,
                    "solved": n_solved == n_total_pids and n_total_pids > 0,
                    "pattern": pat,
                    "errors": dict(err_counts),
                }
            )
            if attempts > 0:
                ls["present"] = True
        result["lessons"].append(ls)

    result["all_errors"] = dict(result["all_errors"])
    return result


# ──────────────────────────────────────────
# 报告生成
# ──────────────────────────────────────────

MEDAL = {0: "🥇 第1名", 1: "🥈 第2名", 2: "🥉 第3名"}


def pct(done, total):
    return 0.0 if total == 0 else done / total * 100


def rank_scores(all_analyses: list[dict], key: str):
    scores = []
    for a in all_analyses:
        t = a["total"][key]
        scores.append((a["uid"], a["uname"], pct(t["done"], t["total"])))
    scores.sort(key=lambda x: -x[2])
    return scores


def build_report(
    target_uid: int,
    target_name: str,
    all_analyses: list[dict],
    lessons: list[dict],
) -> str:
    target = next((a for a in all_analyses if a["uid"] == target_uid), None)
    if not target:
        return f"❌ 未找到 uid={target_uid} 的提交记录（该学生可能还未提交过任何题目）"

    lines = []
    W = 62
    lines.append("=" * W)
    lines.append(f"  {target_name} · 2026春季班 CSP04 学习报告")
    lines.append(
        f"  统计课次：前 {len(lessons)} 节  |  生成: {datetime.now().strftime('%Y-%m-%d %H:%M')}"
    )
    lines.append("=" * W)

    t = target["total"]
    a_rate = pct(t["A"]["done"], t["A"]["total"])
    b_rate = pct(t["B"]["done"], t["B"]["total"])
    c_rate = pct(t["C"]["done"], t["C"]["total"])
    attend = sum(1 for l in target["lessons"] if l["present"])

    # ── 一、总体概览 ──
    lines.append("")
    lines.append("【一、总体学习概览】")
    lines.append(f"  出勤情况：{attend}/{len(lessons)} 节有提交记录")
    lines.append(
        f"  课堂练习（A）：{t['A']['done']}/{t['A']['total']} 题  完成率 {a_rate:.0f}%  累计 {t['A']['attempts']} 次提交"
    )
    lines.append(
        f"  课后作业（B）：{t['B']['done']}/{t['B']['total']} 题  完成率 {b_rate:.0f}%  累计 {t['B']['attempts']} 次提交"
    )
    lines.append(
        f"  拓展练习（C）：{t['C']['done']}/{t['C']['total']} 题  完成率 {c_rate:.0f}%  累计 {t['C']['attempts']} 次提交"
    )

    # ── 二、逐节课 ──
    lines.append("")
    lines.append("【二、逐节课做题明细】")
    for i, ls in enumerate(target["lessons"], 1):
        tag = "✅" if ls["present"] else "⬜"
        lines.append(f"\n  {tag} 第{i}节  {ls['title']}")
        for p in ls["problems"]:
            if p["attempts"] == 0:
                icon = "—"
            elif p["solved"]:
                icon = "✓"
            else:
                icon = "○"
            type_label = {"A": "课堂练习", "B": "课后作业", "C": "拓展练习"}.get(
                p["type"], "其他"
            )
            err_str = ""
            if p["errors"]:
                err_str = (
                    "  [" + ", ".join(f"{k}×{v}" for k, v in p["errors"].items()) + "]"
                )
            lines.append(
                f"    {icon} {type_label}  {p['attempts']}次提交  {p['pattern']}{err_str}"
            )

    # ── 三、班级对比 ──
    lines.append("")
    lines.append("【三、在班级中的位置】")
    n = len(all_analyses)
    for key, label in [("A", "课堂练习"), ("B", "课后作业"), ("C", "拓展练习")]:
        scores = rank_scores(all_analyses, key)
        avg = sum(s for _, _, s in scores) / len(scores) if scores else 0
        rank = next(i for i, (uid, _, _) in enumerate(scores) if uid == target_uid)
        my_sc = next(s for uid, _, s in scores if uid == target_uid)
        medal_str = MEDAL.get(rank, f"第{rank + 1}名")
        lines.append(
            f"\n  {label}  {medal_str}  {target_name}: {my_sc:.0f}%  （均值: {avg:.0f}%，共{n}人）"
        )
        lines.append("  ┌" + "─" * 42)
        for j, (uid, uname, sc) in enumerate(scores):
            bar = "█" * int(sc / 5)
            me = " ← 本人" if uid == target_uid else ""
            lines.append(f"  │ {j + 1:2d}. {uname[:5]:<5s}  {sc:5.1f}%  {bar}{me}")
        lines.append("  └" + "─" * 42)

    # ── 四、亮点 ──
    lines.append("")
    lines.append("【四、做得好的地方 ✨】")
    strengths = []

    if attend == len(lessons):
        strengths.append(f"全勤出勤：{len(lessons)} 节课均有提交记录，学习连贯性强")
    elif attend >= len(lessons) * 0.75:
        strengths.append(f"出勤率高达 {attend}/{len(lessons)} 节，学习节奏稳定")

    scores_a = rank_scores(all_analyses, "A")
    rank_a = next(i for i, (u, _, _) in enumerate(scores_a) if u == target_uid)
    scores_b = rank_scores(all_analyses, "B")
    rank_b = next(i for i, (u, _, _) in enumerate(scores_b) if u == target_uid)
    scores_c = rank_scores(all_analyses, "C")
    rank_c = next(i for i, (u, _, _) in enumerate(scores_c) if u == target_uid)

    if rank_a == 0:
        strengths.append(
            f"课堂练习完成率全班第一（{a_rate:.0f}%），课堂专注度和吸收能力突出"
        )
    elif rank_a <= 2:
        strengths.append(f"课堂练习完成率班级前三（{a_rate:.0f}%），课堂表现优秀")
    elif a_rate >= 80:
        strengths.append(f"课堂练习完成率 {a_rate:.0f}%，基础掌握较扎实")

    if b_rate >= 80:
        strengths.append(
            f"课后作业完成率 {b_rate:.0f}%，课后巩固习惯{'很好' if b_rate >= 90 else '较好'}"
        )
    if rank_b == 0 and b_rate > 0:
        strengths.append("课后作业完成率全班最高，自律性强、学习投入度高")

    if c_rate > 0:
        strengths.append(
            f"有完成拓展练习（{c_rate:.0f}%），主动挑战高难度题，很有进取心"
        )
    if rank_c == 0 and c_rate > 0:
        strengths.append("拓展练习全班第一，学有余力且善于钻研")

    total_solved = t["A"]["done"] + t["B"]["done"] + t["C"]["done"]
    total_attempts = t["A"]["attempts"] + t["B"]["attempts"] + t["C"]["attempts"]
    if total_solved > 0:
        avg_tries = total_attempts / total_solved
        if avg_tries <= 2:
            strengths.append(
                f"平均仅需 {avg_tries:.1f} 次提交即可通过，思路清晰、代码质量高"
            )

    if not strengths:
        strengths.append("坚持参与学习，持续积累是进步的基础")

    for s in strengths:
        lines.append(f"  ✅ {s}")

    # ── 五、待改进 ──
    lines.append("")
    lines.append("【五、需要加强的地方 📌】")
    gaps = []

    if a_rate < 60:
        gaps.append(
            f"课堂练习完成率偏低（{a_rate:.0f}%），课堂时间利用率有待提升，"
            "遇到卡点建议及时向老师请教，不要在一个问题上停留太久"
        )
    elif a_rate < 80:
        gaps.append(
            f"课堂练习完成率 {a_rate:.0f}%，还有提升空间，建议下笔前先把题意和思路整理清楚"
        )

    if b_rate < 30:
        gaps.append(
            f"课后作业完成率仅 {b_rate:.0f}%，课后练习严重不足——知识点在课堂上理解了，"
            "但不做题就很容易遗忘，建议每次课后至少完成必做作业"
        )
    elif b_rate < 60:
        gaps.append(f"课后作业完成率 {b_rate:.0f}%，课后练习不够充分，建议提高完成频率")

    if c_rate == 0 and t["C"]["total"] > 0:
        gaps.append("拓展练习暂无提交——不要求全部做完，但尝试一下对思维提升很有帮助")

    if total_solved > 0:
        avg_tries = total_attempts / total_solved
        if avg_tries > 8:
            gaps.append(
                f"平均每题需要 {avg_tries:.1f} 次才通过，建议做题前先在草稿纸上梳理逻辑，"
                "把样例手动跑一遍，再提交，大幅减少无效试错"
            )
        elif avg_tries > 5:
            gaps.append(
                f"平均每题约 {avg_tries:.1f} 次提交通过，养成'提交前自检'习惯可以进一步减少次数"
            )

    errs = target["all_errors"]
    if errs.get("WA", 0) >= 10:
        gaps.append(
            f"答案错误（WA）累计 {errs['WA']} 次，建议重点练习边界条件判断和分类讨论，"
            "每次WA后认真分析是哪种情况遗漏了"
        )
    if errs.get("CE", 0) >= 5:
        gaps.append(f"编译错误（CE）累计 {errs['CE']} 次，提交前先确认代码没有语法错误")
    if errs.get("RE", 0) >= 5:
        gaps.append(f"运行错误（RE）{errs['RE']} 次，注意数组下标越界和递归终止条件")
    if errs.get("TLE", 0) >= 3:
        gaps.append(
            f"超时（TLE）{errs['TLE']} 次，需要关注算法时间复杂度，避免暴力解法"
        )

    if not gaps:
        gaps.append("整体表现均衡，建议进一步挑战拓展练习，向更高水平迈进")

    for g in gaps:
        lines.append(f"  📌 {g}")

    # ── 六、家长建议 ──
    lines.append("")
    lines.append("【六、给家长的话】")
    if b_rate < 50:
        lines.append("  • 课后作业完成率不高，建议每次上完课后家长提醒孩子完成练习，")
        lines.append('    最好当天做，养成"当天课当天练"的好习惯，效果事半功倍。')
    else:
        lines.append("  • 课后练习完成情况不错！鼓励孩子继续保持，")
        lines.append("    也可以问问孩子今天学了什么，帮助他用语言组织、加深理解。")

    if c_rate == 0 and t["C"]["total"] > 0:
        lines.append("  • 拓展练习还没开始尝试，如果孩子平时有余力，")
        lines.append("    可以鼓励他试试看，不要求做完，重要的是训练思维的过程。")

    lines.append("  • 信奥学习贵在坚持，遇到难题时多鼓励孩子，")
    lines.append("    告诉他“卡住了很正常，调试出来才是成长”。")
    lines.append("    有任何疑问随时联系老师～")

    lines.append("")
    lines.append("=" * W)
    return "\n".join(lines)


# ──────────────────────────────────────────
# 主流程
# ──────────────────────────────────────────


def main():
    import argparse

    parser = argparse.ArgumentParser()
    parser.add_argument("--course-id", required=True)
    parser.add_argument("--uid", required=True, type=int)
    parser.add_argument("--name", default="")
    parser.add_argument("--lessons", default=8, type=int, help="分析前 N 节课")
    args = parser.parse_args()

    client = httpx.Client(base_url=OJ_BASE_URL, timeout=60.0, follow_redirects=True)

    # 1. 登录
    print("\n[1] 登录 OJ...")
    if not login(client):
        sys.exit(1)

    # 2. 解析课程结构
    print(f"\n[2] 解析课程结构: {args.course_id}")
    all_lessons = fetch_course_structure(client, args.course_id)
    print(f"  共找到 {len(all_lessons)} 个有作业的节次")
    for i, l in enumerate(all_lessons, 1):
        hw_summary = " | ".join(f"{h['name']}" for h in l["homeworks"])
        print(f"  第{i}节: {l['title']}  [{hw_summary}]")

    lessons = all_lessons[: args.lessons]
    print(f"\n  分析前 {len(lessons)} 节")

    # 3. 抓取每节课的题目列表 + 提交记录
    print(f"\n[3] 抓取作业题目列表与提交记录...")
    for i, lesson in enumerate(lessons, 1):
        print(f"  第{i}节: {lesson['title']}")
        for hw in lesson["homeworks"]:
            # 先拉官方题目列表
            official_pids = fetch_hw_official_pids(client, hw["hw_id"])
            hw["official_pids"] = official_pids
            # 再拉提交记录
            recs = fetch_hw_records(client, hw["hw_id"])
            hw["records"] = recs
            n_students = len({r["uid"] for r in recs})
            print(
                f"    {hw['name']} (tid={hw['hw_id'][:8]}...): "
                f"{len(official_pids)} 道题 / {len(recs)} 条记录 / {n_students} 名学生"
            )

    client.close()

    # 4. 汇总所有学生
    print(f"\n[4] 汇总学生列表...")
    uid_uname: dict[int, str] = {}
    for lesson in lessons:
        for hw in lesson["homeworks"]:
            for rec in hw["records"]:
                uid_uname[rec["uid"]] = rec["uname"]

    if args.uid not in uid_uname:
        uid_uname[args.uid] = args.name or f"uid{args.uid}"
    elif args.name:
        uid_uname[args.uid] = args.name

    print(f"  共 {len(uid_uname)} 名学生有记录")
    for uid, uname in sorted(uid_uname.items(), key=lambda x: x[1]):
        tag = " ← 目标" if uid == args.uid else ""
        print(f"    uid={uid}  {uname}{tag}")

    # 5. 分析所有学生
    print(f"\n[5] 分析各学生数据...")
    all_analyses = []
    for uid, uname in sorted(uid_uname.items()):
        a = analyze_student(uid, uname, lessons)
        all_analyses.append(a)
        t = a["total"]
        print(
            f"  {uname}(uid={uid}):  "
            f"A={t['A']['done']}/{t['A']['total']}  "
            f"B={t['B']['done']}/{t['B']['total']}  "
            f"C={t['C']['done']}/{t['C']['total']}"
        )

    # 6. 生成报告
    target_name = uid_uname.get(args.uid, args.name or f"uid{args.uid}")
    print(f"\n[6] 生成报告: {target_name} (uid={args.uid})")
    report = build_report(args.uid, target_name, all_analyses, lessons)

    print("\n" + report)

    # 7. 保存
    out_dir = PROJECT_ROOT / ".claude" / "memory" / "oj" / "analysis"
    out_dir.mkdir(parents=True, exist_ok=True)

    out_txt = out_dir / f"course_{args.course_id[:8]}_{args.uid}_report.txt"
    out_txt.write_text(report, encoding="utf-8")
    print(f"\n📄 报告已保存: {out_txt}")

    out_json = out_dir / f"course_{args.course_id[:8]}_{args.uid}_raw.json"
    payload = {
        "course_id": args.course_id,
        "target_uid": args.uid,
        "target_name": target_name,
        "generated_at": datetime.now().isoformat(),
        "lessons_analyzed": len(lessons),
        "lesson_titles": [l["title"] for l in lessons],
        "all_analyses": [
            {k: v for k, v in a.items() if k != "lessons"} for a in all_analyses
        ],
    }
    out_json.write_text(
        json.dumps(payload, ensure_ascii=False, indent=2), encoding="utf-8"
    )
    print(f"📊 数据已保存: {out_json}")


if __name__ == "__main__":
    main()