授权模型切换为个人副本签发；share 脱敏共建；自更新通道

- 公众版不再共享 1000 次授权池：移除 data/.license.json，改置 data/.public_edition.json 标记，首次排盘本机签发个人副本（初始 20 次、不限期）；推送脱敏档案每盘 +10 次（同盘只计一次） - 新增 share/ 共建：结构化脱敏导出（share_export.py）+ 双远端推送（share_push.py）+ 隐私协议（share/README.md）；推送前必须展示脱敏包全文并获用户授权 - 新增自更新：VERSION + manifest.json（SHA256 清单）+ scripts/self_update.py（24h 缓存检查、先验后换、保留授权与档案）；排盘成功后 stderr 提示新版本 - SKILL.md 增「档案共建（share/）」「版本与自更新」两节； LICENSE.md 增第六条（档案权属与撤回） - 测试 100 项全绿
2026-10-09 21:24:25 +08:00
parent 5889d7a150
commit 26a83df3c4
20 changed files with 2760 additions and 22 deletions
@@ -0,0 +1,265 @@
+#!/usr/bin/env python3
+"""脱敏档案导出 —— 从 memory 档案生成结构化脱敏包（JSON）到 share/pending/。
+
+用法：
+    python share_export.py --archive memory/archives/庚午辛巳己卯戊辰-男-a1b2c3.md
+    python share_export.py --birth "1990-05-14 08:30" --place "四川省成都市" --gender 男
+
+脱敏规格（见 share/README.md）：不含姓名；出生信息只保留年月日与时辰（不留分钟
+与钟表时间）；地域只保留到省级；事件一律类目化；正文自由文本（归纳、盘面细节、
+反馈原文）不入包。盘 hash = sha1(四柱+出生地) 前 8 位，用于去重与撤回。
+
+完整度门槛：性别/四柱/大运齐全 + 至少一条反馈事件（任意核实状态）；
+不满足拒绝导出，退出码 2 并说明缺什么。退出码：0 成功；2 不可用/不完整；3 内部错误。
+"""
+from __future__ import annotations
+
+import argparse
+import hashlib
+import json
+import re
+import sys
+import traceback
+from datetime import datetime
+from pathlib import Path
+
+sys.path.insert(0, str(Path(__file__).resolve().parent))
+
+import _memory_store as store
+
+SCHEMA = "share/v1"
+PLACEHOLDERS = {"（无）", "（未记录）", "无", "未记录"}
+
+# 事件/关注方向类目化：首个命中的类目生效，都不命中归「其他」
+CATEGORIES: list[tuple[str, tuple[str, ...]]] = [
+    ("婚姻", ("婚姻", "结婚", "离婚", "配偶", "夫妻", "婚变", "再婚")),
+    ("感情", ("感情", "恋爱", "桃花", "对象", "分手", "复合")),
+    ("事业", ("事业", "工作", "升职", "跳槽", "调动", "失业", "创业", "辞职", "职务", "官非")),
+    ("财运", ("财运", "破财", "投资", "收入", "赚钱", "欠债", "债务", "置业", "买房")),
+    ("健康", ("健康", "生病", "手术", "伤病", "身体", "住院", "疾病", "意外")),
+    ("学业", ("学业", "考试", "升学", "考研", "高考", "留学", "读书")),
+    ("六亲", ("父亲", "母亲", "父母", "子女", "孩子", "兄弟", "姐妹", "六亲", "亲人", "丧")),
+    ("流年运势", ("流年", "运势", "大运", "运程")),
+]
+
+_PROVINCES = (
+    "北京市", "天津市", "上海市", "重庆市",
+    "河北省", "山西省", "辽宁省", "吉林省", "黑龙江省", "江苏省", "浙江省",
+    "安徽省", "福建省", "江西省", "山东省", "河南省", "湖北省", "湖南省",
+    "广东省", "海南省", "四川省", "贵州省", "云南省", "陕西省", "甘肃省",
+    "青海省", "台湾省",
+    "内蒙古自治区", "广西壮族自治区", "西藏自治区", "宁夏回族自治区",
+    "新疆维吾尔自治区", "香港特别行政区", "澳门特别行政区",
+)
+_PROVINCE_SHORT = {
+    "北京": "北京市", "天津": "天津市", "上海": "上海市", "重庆": "重庆市",
+    "内蒙古": "内蒙古自治区", "广西": "广西壮族自治区", "西藏": "西藏自治区",
+    "宁夏": "宁夏回族自治区", "新疆": "新疆维吾尔自治区",
+    "香港": "香港特别行政区", "澳门": "澳门特别行政区",
+}
+
+_STATUS_TAGS = {"【已核实】": "已核实", "【已证伪】": "已证伪", "【未核实】": "未核实"}
+
+
+def _categorize_all(text: str) -> list[str]:
+    return [cat for cat, keywords in CATEGORIES if any(k in text for k in keywords)]
+
+
+def _categorize(text: str) -> str:
+    cats = _categorize_all(text)
+    return cats[0] if cats else "其他"
+
+
+def _province_of(birthplace: str) -> str | None:
+    """地域脱敏：只保留到省级；识别不出省级则舍弃（宁缺毋滥）。"""
+    for p in _PROVINCES:
+        if p in birthplace:
+            return p
+    for short, full in _PROVINCE_SHORT.items():
+        if birthplace.startswith(short):
+            return full
+    m = re.match(r"^(.{2,6}?省)", birthplace)
+    return m.group(1) if m else None
+
+
+def _section_lines(body: str, heading: str) -> list[str]:
+    """取档案正文某固定小节（SECTIONS 内）的行，到下一个固定小节或文末。"""
+    lines = body.splitlines()
+    bounds = {f"## {s}" for s in store.SECTIONS} - {f"## {heading}"}
+    start = next((i for i, ln in enumerate(lines)
+                  if ln.strip() == f"## {heading}"), None)
+    if start is None:
+        return []
+    end = next((i for i in range(start + 1, len(lines))
+                if lines[i].strip() in bounds), len(lines))
+    return lines[start + 1:end]
+
+
+def _chart_section_lines(body: str, heading: str, nxt: str) -> list[str]:
+    """取「盘面」内嵌命盘 Markdown 的小节（如 ## 大运 到 ## 小运）。"""
+    lines = body.splitlines()
+    start = next((i for i, ln in enumerate(lines)
+                  if ln.strip() == f"## {heading}"), None)
+    if start is None:
+        return []
+    end = next((i for i in range(start + 1, len(lines))
+                if lines[i].strip() == f"## {nxt}"), len(lines))
+    return lines[start + 1:end]
+
+
+def _parse_luck(body: str) -> list[dict]:
+    luck = []
+    for ln in _chart_section_lines(body, "大运", "小运"):
+        m = re.match(r"^-\s*第\s*(\d+)\s*步：(\S{2})（[^）]*，(\d{4})\s*年）", ln.strip())
+        if m:
+            luck.append({"seq": int(m.group(1)), "pillar": m.group(2),
+                         "start_year": int(m.group(3))})
+    return luck
+
+
+def _parse_focus(body: str) -> list[str]:
+    cats: list[str] = []
+    for ln in _section_lines(body, "主要关注方向"):
+        t = ln.strip().lstrip("-•· ").strip()
+        if not t or t in PLACEHOLDERS:
+            continue
+        for cat in _categorize_all(t):
+            if cat not in cats:
+                cats.append(cat)
+    return cats
+
+
+def _parse_feedback(body: str) -> list[dict]:
+    """反馈事件：只取年份、类目、核实状态——反馈原文不入包。"""
+    events = []
+    for ln in _section_lines(body, "反馈与澄清"):
+        t = ln.strip().lstrip("-•· ").strip()
+        if not t or t in PLACEHOLDERS:
+            continue
+        status = "未核实"
+        for tag, name in _STATUS_TAGS.items():
+            if tag in t:
+                status = name
+                break
+        m = re.search(r"(?:19|20)\d{2}", t)
+        events.append({"year": int(m.group(0)) if m else None,
+                       "category": _categorize(t),
+                       "status": status})
+    return events
+
+
+def chart_hash(pillars: list[str], birthplace: str) -> str:
+    raw = f"{''.join(pillars)}|{birthplace.strip()}"
+    return hashlib.sha1(raw.encode("utf-8")).hexdigest()[:8]
+
+
+def build_package(meta: dict, body: str) -> tuple[dict, list[str]]:
+    """组装脱敏包；返回 (package, 缺失项列表)。package 保证不含姓名/分钟/省以下地域。"""
+    pillars = meta.get("pillars", [])
+    if isinstance(pillars, str):
+        pillars = pillars.split()
+    pillars = [str(p) for p in pillars]
+    birthplace = str(meta.get("birthplace", ""))
+    gender = str(meta.get("gender", ""))
+
+    pkg: dict = {"schema": SCHEMA,
+                 "exported_at": datetime.now().strftime("%Y-%m-%d %H:%M"),
+                 "chart_hash": chart_hash(pillars, birthplace),
+                 "gender": gender or None,
+                 "pillars": pillars,
+                 "luck_pillars": _parse_luck(body),
+                 "focus_categories": _parse_focus(body),
+                 "feedback": _parse_feedback(body)}
+
+    # 出生信息脱敏：只保留年月日与时辰（时支），不留分钟与钟表时间
+    m = re.match(r"^(\d{4})-(\d{2})-(\d{2})", str(meta.get("birth_clock", "")).strip())
+    if m:
+        pkg["birth_date"] = f"{m.group(1)}-{m.group(2)}-{m.group(3)}"
+    if len(pillars) == 4 and len(pillars[3]) == 2:
+        pkg["birth_shichen"] = pillars[3][1]
+    province = _province_of(birthplace)
+    if province:
+        pkg["region_province"] = province
+
+    missing = []
+    if not gender:
+        missing.append("性别")
+    if len(pillars) != 4:
+        missing.append("四柱")
+    if not pkg["luck_pillars"]:
+        missing.append("大运序列")
+    if not pkg["feedback"]:
+        missing.append("反馈事件（至少一条，任意核实状态）")
+    return pkg, missing
+
+
+def _find_archive(args) -> Path | None:
+    if args.archive:
+        p = Path(args.archive).expanduser().resolve()
+        return p if p.is_file() else None
+    root = store.resolve_root(args.memory_root, __file__)
+    for path in store.iter_archives(root):
+        meta, _ = store.parse_frontmatter(path.read_text(encoding="utf-8"))
+        if meta and store.matches(meta, name=args.name, birth=args.birth,
+                                  place=args.place, gender=args.gender,
+                                  pillars=args.pillars):
+            return path
+    return None
+
+
+def main() -> int:
+    ap = argparse.ArgumentParser(
+        prog="share_export.py",
+        description="从 memory 档案导出结构化脱敏包（JSON）到 share/pending/；不含姓名、分钟、省以下地域")
+    ap.add_argument("--archive", default=None, help="档案文件路径（直接指定）")
+    ap.add_argument("--name", default=None, help="姓名（检索条件）")
+    ap.add_argument("--gender", default=None, choices=["男", "女"], help="性别（检索条件）")
+    ap.add_argument("--birth", default=None,
+                    help='出生时刻，格式 "YYYY-MM-DD HH:MM"（检索条件）')
+    ap.add_argument("--place", default=None, help="出生地（检索条件）")
+    ap.add_argument("--pillars", default=None,
+                    help='四柱干支，如 "丁亥 乙巳 庚子 辛巳"（检索条件）')
+    ap.add_argument("--memory-root", default=None,
+                    help=f"档案根目录（默认技能根 memory/；亦可用环境变量 {store.ENV_VAR}）")
+    ap.add_argument("--share-root", default=None,
+                    help="share/ 目录（默认技能根 share/，测试覆盖用）")
+    a = ap.parse_args()
+
+    if not a.archive and all(v is None for v in (a.name, a.gender, a.birth,
+                                                 a.place, a.pillars)):
+        print("ERROR: 给 --archive <档案路径>，或至少一个检索条件"
+              "（--name/--birth/--place/--gender/--pillars）。", file=sys.stderr)
+        return 2
+
+    try:
+        arch = _find_archive(a)
+        if arch is None:
+            print("ERROR: 未找到匹配档案（先用 scripts/memory_save.py 建档，"
+                  "并确认检索条件与档案一致）。", file=sys.stderr)
+            return 2
+        meta, body = store.parse_frontmatter(arch.read_text(encoding="utf-8"))
+        pkg, missing = build_package(meta, body)
+        if missing:
+            print(f"ERROR: 档案完整度不足，拒绝导出。缺：{'、'.join(missing)}。"
+                  f"（性别/四柱/大运齐全 + 至少一条反馈事件才可导出）", file=sys.stderr)
+            return 2
+
+        share_root = (Path(a.share_root).expanduser().resolve() if a.share_root
+                      else Path(__file__).resolve().parents[1] / "share")
+        pending = share_root / "pending"
+        pending.mkdir(parents=True, exist_ok=True)
+        out = pending / f"{pkg['chart_hash']}.json"
+        out.write_text(json.dumps(pkg, ensure_ascii=False, indent=2) + "\n",
+                       encoding="utf-8")
+        print(out)
+        return 0
+    except SystemExit:
+        raise
+    except Exception:
+        print("ERROR: 导出内部错误。", file=sys.stderr)
+        traceback.print_exc(file=sys.stderr)
+        return 3
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())