# -*- coding: utf-8 -*-
"""行业深度研报查询脚本 —— 供 Claude Skill 使用。

连接 MySQL 数据库，查询行业报告的结构化数据和原始 Markdown 文本。
依赖: pymysql（项目已安装）。

用法:
    python query_industry.py <command> [args...]

子命令:
    lookup <keyword>              按关键词或代码查找行业
    summary <code>                获取行业核心指标摘要
    module <code> <num|all>       获取结构化 Module JSON
    part <code> <nums|all>        获取 Part 原始 Markdown
    rank <field> [--top N]        跨行业排名
    list                          列出所有行业
"""
from __future__ import annotations

import json
import sys
import os

# Windows 终端 GBK 编码问题
sys.stdout.reconfigure(encoding="utf-8")
sys.stderr.reconfigure(encoding="utf-8")

# 数据库配置
sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
from db_config import DB_CONFIG

try:
    import pymysql
except ImportError:
    import subprocess
    print("pymysql 未安装，正在自动安装...", file=sys.stderr)
    subprocess.check_call([sys.executable, "-m", "pip", "install", "pymysql", "-q"])
    import pymysql


def _get_conn():
    """获取数据库连接。"""
    return pymysql.connect(**DB_CONFIG, cursorclass=pymysql.cursors.DictCursor)


# ---------------------------------------------------------------------------
# 子命令实现
# ---------------------------------------------------------------------------

def cmd_lookup(keyword: str):
    """按关键词查找行业，模糊匹配 name 和 code。"""
    conn = _get_conn()
    try:
        with conn.cursor() as cur:
            like = f"%{keyword}%"
            cur.execute(
                """SELECT code, name, stars, stars_label, headline
                   FROM industry_reports
                   WHERE name LIKE %s OR code LIKE %s
                   ORDER BY stars DESC""",
                (like, like),
            )
            rows = cur.fetchall()
    finally:
        conn.close()

    if not rows:
        print(f"未找到匹配「{keyword}」的行业")
        return

    print(f"找到 {len(rows)} 个行业：\n")
    for r in rows:
        stars_display = "★" * r["stars"] + "☆" * (5 - r["stars"])
        print(f"  {r['code']}  {r['name']}  {stars_display} ({r['stars_label']})")
        if r["headline"]:
            print(f"         {r['headline']}")
        print()


def cmd_summary(code: str):
    """获取行业核心指标摘要。"""
    conn = _get_conn()
    try:
        with conn.cursor() as cur:
            cur.execute(
                """SELECT code, name, stars, stars_label, cagr, gross_margin,
                          net_margin, cr5, cr10, hhi, concentration_trend,
                          headline, total_characters, report_version, generated_at
                   FROM industry_reports WHERE code = %s""",
                (code,),
            )
            row = cur.fetchone()
    finally:
        conn.close()

    if not row:
        print(f"未找到行业: {code}")
        return

    print(json.dumps(row, ensure_ascii=False, indent=2, default=str))


def cmd_module(code: str, module_num: str):
    """获取结构化 Module JSON 数据。"""
    conn = _get_conn()
    try:
        with conn.cursor() as cur:
            if module_num == "all":
                cur.execute(
                    """SELECT rm.module_num, rm.module_json
                       FROM report_modules rm
                       JOIN industry_reports ir ON ir.id = rm.report_id
                       WHERE ir.code = %s
                       ORDER BY rm.module_num""",
                    (code,),
                )
            else:
                cur.execute(
                    """SELECT rm.module_num, rm.module_json
                       FROM report_modules rm
                       JOIN industry_reports ir ON ir.id = rm.report_id
                       WHERE ir.code = %s AND rm.module_num = %s""",
                    (code, int(module_num)),
                )
            rows = cur.fetchall()
    finally:
        conn.close()

    if not rows:
        print(f"未找到行业 {code} 的 Module 数据")
        return

    for r in rows:
        mj = r["module_json"]
        data = json.loads(mj) if isinstance(mj, str) else mj
        print(f"=== Module {r['module_num']} ===")
        print(json.dumps(data, ensure_ascii=False, indent=2))
        print()


def cmd_part(code: str, part_nums: str):
    """获取 Part 原始 Markdown 文本。"""
    conn = _get_conn()
    try:
        with conn.cursor() as cur:
            if part_nums == "all":
                cur.execute(
                    """SELECT rrp.part_num, rrp.content, rrp.char_count
                       FROM report_raw_parts rrp
                       JOIN industry_reports ir ON ir.id = rrp.report_id
                       WHERE ir.code = %s
                       ORDER BY rrp.part_num""",
                    (code,),
                )
            else:
                nums = [int(x.strip()) for x in part_nums.split(",")]
                placeholders = ",".join(["%s"] * len(nums))
                cur.execute(
                    f"""SELECT rrp.part_num, rrp.content, rrp.char_count
                        FROM report_raw_parts rrp
                        JOIN industry_reports ir ON ir.id = rrp.report_id
                        WHERE ir.code = %s AND rrp.part_num IN ({placeholders})
                        ORDER BY rrp.part_num""",
                    [code] + nums,
                )
            rows = cur.fetchall()
    finally:
        conn.close()

    if not rows:
        print(f"未找到行业 {code} 的 Part 数据")
        return

    for r in rows:
        label = "meta (目录+执行摘要)" if r["part_num"] == 0 else f"Part {r['part_num']}"
        print(f"{'=' * 60}")
        print(f"  {label}  ({r['char_count']} 字)")
        print(f"{'=' * 60}")
        print(r["content"])
        print()


def cmd_rank(field: str, top: int = 5):
    """跨行业排名，按指定字段降序。"""
    field_map = {
        "stars": "stars",
        "cagr": "cagr_value",
        "cagr_value": "cagr_value",
        "gross_margin": "gross_margin_value",
        "gross_margin_value": "gross_margin_value",
    }
    column = field_map.get(field)
    if not column:
        print(f"不支持的排序字段: {field}")
        print(f"可用字段: {', '.join(field_map.keys())}")
        return

    conn = _get_conn()
    try:
        with conn.cursor() as cur:
            cur.execute(
                f"""SELECT code, name, stars, stars_label, cagr, gross_margin, headline
                    FROM industry_reports
                    WHERE {column} IS NOT NULL
                    ORDER BY {column} DESC
                    LIMIT %s""",
                (top,),
            )
            rows = cur.fetchall()
    finally:
        conn.close()

    if not rows:
        print("暂无数据")
        return

    print(f"Top {top} — 按 {field} 排序:\n")
    for i, r in enumerate(rows, 1):
        print(f"  {i}. {r['code']}  {r['name']}")
        print(f"     景气度={r['stars']}★  CAGR={r['cagr']}  毛利率={r['gross_margin']}")
        if r["headline"]:
            print(f"     {r['headline']}")
        print()


def cmd_list():
    """列出所有行业。"""
    conn = _get_conn()
    try:
        with conn.cursor() as cur:
            cur.execute(
                "SELECT code, name, stars FROM industry_reports ORDER BY code"
            )
            rows = cur.fetchall()
    finally:
        conn.close()

    if not rows:
        print("暂无行业数据")
        return

    print(f"共 {len(rows)} 个行业:\n")
    for r in rows:
        print(f"  {r['code']}  {r['name']}  {'★' * r['stars']}")


# ---------------------------------------------------------------------------
# 主入口
# ---------------------------------------------------------------------------

def main():
    if len(sys.argv) < 2:
        print(__doc__)
        sys.exit(0)

    cmd = sys.argv[1]

    if cmd == "lookup":
        if len(sys.argv) < 3:
            print("用法: query_industry.py lookup <keyword>")
            sys.exit(1)
        cmd_lookup(sys.argv[2])

    elif cmd == "summary":
        if len(sys.argv) < 3:
            print("用法: query_industry.py summary <code>")
            sys.exit(1)
        cmd_summary(sys.argv[2])

    elif cmd == "module":
        if len(sys.argv) < 4:
            print("用法: query_industry.py module <code> <1-5|all>")
            sys.exit(1)
        cmd_module(sys.argv[2], sys.argv[3])

    elif cmd == "part":
        if len(sys.argv) < 4:
            print("用法: query_industry.py part <code> <0-8|all>")
            sys.exit(1)
        cmd_part(sys.argv[2], sys.argv[3])

    elif cmd == "rank":
        if len(sys.argv) < 3:
            print("用法: query_industry.py rank <field> [--top N]")
            sys.exit(1)
        field = sys.argv[2]
        top = 5
        if "--top" in sys.argv:
            idx = sys.argv.index("--top")
            if idx + 1 < len(sys.argv):
                top = int(sys.argv[idx + 1])
        cmd_rank(field, top)

    elif cmd == "list":
        cmd_list()

    else:
        print(f"未知命令: {cmd}")
        print(__doc__)
        sys.exit(1)


if __name__ == "__main__":
    main()