# -*- coding: utf-8 -*-
"""契约 ↔ 代码 接口实现度比对（可复现的盘点工具）。

用途：回答"121 个契约接口里，哪些已经实现、哪些还没做"。
      —— 这个数在《08-完成度盘点与问题清单.md》里被引用，必须可复现、不能靠记忆。

契约来源：docs/03-接口设计.md 的 §4 接口总览（合计 121 个，全文权威清单）。
代码来源：backend 所有 controller 的类级 @RequestMapping + 方法级 @*Mapping。

★ 两个必须容忍的写法（踩过）：
  1. 全限定注解 `@org.springframework.web.bind.annotation.PostMapping("/x")`
     —— FillController 就是这么写的；用 `@PostMapping` 短名匹配会漏掉 3 个接口；
  2. `@RequestMapping(value="/x", method=RequestMethod.POST)` 这种显式写法。

用法：
    python scripts/check-api-parity.py            # 打印逐分类实现度 + 缺口清单
    python scripts/check-api-parity.py --quiet    # 只打印汇总（CI 里用）
"""
from __future__ import annotations

import glob
import os
import re
import sys

if hasattr(sys.stdout, "reconfigure"):
    sys.stdout.reconfigure(encoding="utf-8")

ROOT = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
DOC = os.path.join(ROOT, "docs", "03-接口设计.md")
JAVA = os.path.join(ROOT, "backend", "xt-server", "src", "main", "java", "cn", "tianbaotong", "fill")

# §4 里这两节是索引（交叉引用别处已计数的接口），不重复计入
INDEX_SECTIONS = ("4.9", "4.11")


def norm(path: str) -> str:
    """归一化：去掉 /api/v1 前缀、统一占位符（`{id}` 与 `{fileId}` 视为同一个）。"""
    p = path.replace("/api/v1", "")
    p = re.sub(r"\{[^}]+\}", "{}", p)
    return p.rstrip("/") or "/"


def doc_overview() -> dict[str, list[tuple[str, str]]]:
    """解析 §4.1~§4.10 的清单表，返回 分类 -> [(METHOD, path)]。"""
    text = open(DOC, encoding="utf-8").read()
    sec = text[text.index("## §4"):text.index("## §5")]
    out: dict[str, list[tuple[str, str]]] = {}
    cur = None
    for line in sec.splitlines():
        h = re.match(r"^### (4\.\d+) (.+)$", line)
        if h:
            cur = f"{h.group(1)} {h.group(2)}"
            out[cur] = []
            continue
        if cur is None or not line.startswith("|") or cur.startswith(INDEX_SECTIONS):
            continue
        cells = [c.strip() for c in line.strip("|").split("|")]
        if len(cells) < 2 or set(cells[0]) <= set("-: "):
            continue
        method, path = cells[0].replace("⭐", "").strip(), cells[1].strip("` ")
        if not path.startswith("/"):
            continue
        for m in re.findall(r"GET|POST|PUT|DELETE|PATCH", method.upper()) or ["?"]:
            out[cur].append((m, path))
    return out


def code_endpoints() -> list[tuple[str, str, str]]:
    """解析所有 controller，返回 [(文件, METHOD, 完整路径)]。"""
    out = []
    for f in glob.glob(os.path.join(JAVA, "module", "**", "controller", "*.java"), recursive=True):
        src = open(f, encoding="utf-8").read()
        mb = re.search(r'@RequestMapping\(\s*(?:value\s*=\s*)?"([^"]+)"', src)
        base = mb.group(1) if mb else ""
        for mm in re.finditer(
                r'@(?:[\w.]+\.)?(Get|Post|Put|Delete|Patch)Mapping(?:\(\s*(?:value\s*=\s*)?"([^"]*)")?',
                src):
            out.append((os.path.basename(f), mm.group(1).upper(),
                        (base + (mm.group(2) or "")) or "/"))
        for mm in re.finditer(
                r'@RequestMapping\(\s*value\s*=\s*"([^"]+)"\s*,\s*method\s*=\s*'
                r'(?:RequestMethod\.)?(GET|POST|PUT|DELETE|PATCH)', src):
            out.append((os.path.basename(f), mm.group(2), base + mm.group(1)))
    return sorted(set(out), key=lambda x: (x[0], x[2]))


def main() -> int:
    quiet = "--quiet" in sys.argv
    only_missing = "--missing" in sys.argv
    doc = doc_overview()
    code = code_endpoints()
    code_keys = {(m, norm(p)) for _, m, p in code}

    claimed = implemented = 0
    missing: list[tuple[str, str, str]] = []
    rows = []
    for cat, eps in doc.items():
        uniq = sorted(set(eps))
        done = [e for e in uniq if (e[0], norm(e[1])) in code_keys]
        miss = [e for e in uniq if (e[0], norm(e[1])) not in code_keys]
        claimed += len(uniq)
        implemented += len(done)
        missing += [(cat, m, p) for m, p in miss]
        rows.append((cat, len(uniq), len(done), len(miss)))

    if not quiet and not only_missing:
        print("=" * 78)
        print("代码注册接口：%d 个" % len(code))
        print("=" * 78)
        for f, m, p in code:
            print("  %-30s %-6s %s" % (f, m, p))
        print()

    print("=" * 78)
    print("契约 §4 总览逐分类")
    print("=" * 78)
    for cat, n, d, miss_n in rows:
        print("  %-22s 声称 %2d  已实现 %2d  缺口 %2d" % (cat, n, d, miss_n))

    print()
    print("合计：契约 %d ／ 已实现 %d ／ 缺口 %d ／ 实现度 %.0f%%"
          % (claimed, implemented, len(missing), 100.0 * implemented / claimed))

    if not quiet:
        print()
        cur = None
        for cat, m, p in missing:
            if cat != cur:
                print("\n【%s】" % cat)
                cur = cat
            print("    %-6s %s" % (m, p))
        # `--missing` 就是"只看还差什么"：上面那两段（代码清单 + 逐分类数字）
        # 在"接着往下做"时反而是噪声，而这是最常被问的问题
        if only_missing:
            print()
            print("说明：--missing 只列缺口；不带参数时会先打全部已注册接口。")

    # 契约必须恰好 121（§2 的口径），不等就说明解析或文档出了问题
    if claimed != 121:
        print("  [FAIL] 契约接口数应为 121，实际解析出 %d —— 检查 §4 表格格式" % claimed)
        return 1
    return 0


if __name__ == "__main__":
    raise SystemExit(main())