#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""docs-index-stats.py — 从 INDEX.md §二 清单表**算**出状态分布，并可就地刷新「状态摘要」行。

为什么需要它
────────────
「状态摘要」原本是**人工手写**的硬数字，档案一多必然漂移（2026-09-13 实测：摘要写「档案 72 份」，
而表格实际 76 行）。本脚本把摘要变成**机器生成**：读表 → 统计 → 打印；`--write` 时回写该行，
并顺带做**图例自检**；**完整保持文件原有行尾**（CRLF 文件不会被转成 LF）（表格用到的图标必须在「图例」行里有定义，缺则补）。

用法
────
  python3 07-scripts/docs-index-stats.py            # 只打印 + 一致性判定
  python3 07-scripts/docs-index-stats.py --write    # 就地刷新「状态摘要」行（必要时补图例）
退出码：0 = 一致（或已刷新）；1 = 漂移且未加 --write；2 = 结构异常。
"""

import io
import os
import re
import sys
import collections

ROOT = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
INDEX = os.path.join(ROOT, "INDEX.md")

ORDER = ["✅", "🔄", "🧪", "📝", "🔍", "📋", "🟡", "🗄", "🔧", "⚠️"]
LEGEND_OF = {
    "✅": "已落地", "🔄": "维护中", "🧪": "PoC", "📝": "待开发", "🔍": "核查完成",
    "📋": "评估", "🟡": "保留兜底", "🗄": "归档", "🔧": "修复", "⚠️": "警示",
}
HEADER_RE = re.compile(r"^\|\s*号\s*\|\s*状态\s*\|")
LEAF_RE = re.compile(r"^0\d$")


def parse(index_path):
    """返回 (lines, arch_rows, leaf_rows, counter, other_count)。"""
    # ⚠️ 必须 newline=""（通用换行会把 CRLF 静默转成 LF —— 2026-09-13 踩过两次）
    raw = io.open(index_path, encoding="utf-8", newline="").read()
    lines = raw.split("\n")
    try:
        start = next(i for i, l in enumerate(lines) if HEADER_RE.match(l))
    except StopIteration:
        raise SystemExit("ERROR: INDEX.md 里找不到「| 号 | 状态 | 一句话 |」表头")
    arch, leaf, other = [], [], 0
    for l in lines[start + 2:]:
        if not l.startswith("|"):
            break
        cells = [x.strip() for x in l.split("|")]
        if len(cells) < 4 or not cells[1]:
            continue
        no, st = cells[1], cells[2]
        if no.startswith("04-"):
            arch.append((no, st))
        elif LEAF_RE.match(no):
            leaf.append((no, st))
        else:
            other += 1
    cnt = collections.Counter(st for _, st in arch)
    cnt.update(st for _, st in leaf)
    eol = "\r\n" if raw.count("\r\n") > 0 else "\n"
    return lines, arch, leaf, cnt, other, eol


def summary_text(cnt, n_arch, leaf_names, other):
    parts = ["%s %d" % (k, cnt[k]) for k in ORDER if cnt.get(k)]
    unmarked = sum(v for k, v in cnt.items() if k not in ORDER)
    if unmarked:
        parts.append("未标记 %d" % unmarked)
    leaf = ("，另含根级编号 %d 条（%s）" % (len(leaf_names), "/".join(leaf_names))) if leaf_names else ""
    return (
        "> **状态摘要**（**机器生成，勿手改**）：档案 **%d** 份（`04-*`）%s，"
        "另有非编号行 %d 条（README / INDEX / 技能 / poc 等）—— %s。"
        "复跑 `python3 07-scripts/docs-index-stats.py` 取数，`--write` 就地刷新本行。"
        % (n_arch, leaf, other, " ｜ ".join(parts))
    )


def main():
    write = "--write" in sys.argv
    lines, arch, leaf, cnt, other, eol = parse(INDEX)
    leaf_names = [n for n, _ in leaf]
    new = summary_text(cnt, len(arch), leaf_names, other)

    print("档案 04-* %d 份 ｜ 根级编号 %s ｜ 非编号行 %d 条" % (len(arch), "/".join(leaf_names) or "—", other))
    for k in ORDER:
        if cnt.get(k):
            print("  %s %-6s %d" % (k, LEGEND_OF.get(k, ""), cnt[k]))
    unmarked = sum(v for k, v in cnt.items() if k not in ORDER)
    if unmarked:
        print("  未标记 %d" % unmarked)

    changed = False

    # ── 图例自检 ─────────────────────────────────────────────────────────
    li = [i for i, l in enumerate(lines) if l.startswith("> 图例：")]
    if li:
        legend = lines[li[0]]
        missing = [k for k in cnt if k not in legend]
        if missing:
            print("⚠️ 图例缺图标：%s" % " ".join(missing))
            if write:
                lines[li[0]] = legend.rstrip().rstrip("｜") + "".join(
                    "｜%s%s" % (k, LEGEND_OF.get(k, "")) for k in missing
                )
                changed = True
                print("✓ 已补进图例行")

    # ── 摘要行 ───────────────────────────────────────────────────────────
    idx = [i for i, l in enumerate(lines) if l.startswith("> **状态摘要**")]
    if not idx:
        print("ERROR: 未找到「状态摘要」行", file=sys.stderr)
        return 2
    old = lines[idx[0]]
    if old.strip() == new.strip() and not changed:
        print("✓ 摘要与表格一致")
        return 0
    if not write:
        print("✗ 摘要与表格不一致（加 --write 刷新）")
        print("  旧: %s" % old.strip()[:90])
        print("  新: %s" % new.strip()[:90])
        return 1
    lines[idx[0]] = new
    io.open(INDEX, "w", encoding="utf-8", newline="").write(eol.join(lines))
    print("✓ 已刷新 INDEX.md（摘要行%s）" % ("+图例" if changed else ""))
    return 0


if __name__ == "__main__":
    sys.exit(main())