#!/usr/bin/env python3
"""kb.py — deterministic guard for the COMPANY knowledge base (shared vault).

company_kb/ is a SHARED, company-level Obsidian-style vault (plain markdown), distinct
from per-user state in users/<id>/. The runtime model READS it directly (tiered: open
_index.md first, then the page it points to). The only WRITE the runtime is allowed is a
lightweight CAPTURE: append a note to the LIVE page company_kb/notes.md — readable
IMMEDIATELY, no ingest engine needed (works on Starter). It NEVER merges/cross-links;
the heavy ingest (notes.md + bulk sources → structured topic pages) is done by a capable
AI at admin/install time via obsidian-wiki (Pro+).

Writes are admin-only (CEO) because the vault is shared across the whole team.

  python3 lib/kb.py status              -> {ok, pages, notes, empty, message}
  python3 lib/kb.py search "<keyword>"  -> {ok, hits:[{path,title}]}  (tiered-retrieval helper)
  python3 lib/kb.py capture "<text>"    -> append note to notes.md (admin only)

(Bộ khung khởi nghiệp Starter Plus có engine RIÊNG: lib/startup_kit.py → company_kb/business/.)

Stdlib only. user_id + org_role resolved the same way identity.py does (registry).
"""
import os
import re
import sys
import json
import glob
import datetime

PLATFORM_PREFIX = {"telegram": "tg", "slack": "sl", "discord": "dc", "whatsapp": "wa"}
ADMIN_ROLES = {"ceo", "admin", "owner"}


def _bundle_root():
    # file at <bundle>/lib/kb.py  → root is parent of lib/
    return os.path.dirname(os.path.dirname(os.path.abspath(__file__)))


ROOT = _bundle_root()
KB = os.path.join(ROOT, "company_kb")
NOTES = os.path.join(KB, "notes.md")          # live page for manual captures (readable immediately)
INDEX = os.path.join(KB, "_index.md")
REGISTRY = os.path.join(ROOT, "users", "_index.json")


def _sess(name, default=""):
    return os.environ.get(name, default) or default


def resolve_user():
    platform = _sess("HERMES_SESSION_PLATFORM", "telegram").lower()
    uid = _sess("HERMES_SESSION_USER_ID") or _sess("HERMES_SESSION_CHAT_ID")
    if not uid:
        return None
    prefix = PLATFORM_PREFIX.get(platform) or (re.sub(r"[^a-z0-9]", "", platform)[:2] or "xx")
    safe = re.sub(r"[^A-Za-z0-9_-]", "", str(uid))
    return ("%s_%s" % (prefix, safe)) if safe else None


def _user_record():
    uid = resolve_user()
    if not uid:
        return uid, {}
    try:
        with open(REGISTRY, encoding="utf-8") as f:
            return uid, json.load(f).get("users", {}).get(uid, {})
    except Exception:
        return uid, {}


def is_admin():
    _, rec = _user_record()
    return str(rec.get("org_role", "")).strip().lower() in ADMIN_ROLES


def _pages():
    if not os.path.isdir(KB):
        return []
    out = []
    for p in glob.glob(os.path.join(KB, "**", "*.md"), recursive=True):
        b = os.path.basename(p)
        if b.startswith("_"):
            continue  # skip _index.md + other underscore meta (notes.md IS included)
        out.append(p)
    return sorted(out)


def _title(path):
    try:
        for line in open(path, encoding="utf-8"):
            line = line.strip()
            if line.startswith("#"):
                return line.lstrip("# ").strip()
    except Exception:
        pass
    return os.path.splitext(os.path.basename(path))[0]


def _note_count():
    if not os.path.isfile(NOTES):
        return 0
    try:
        return sum(1 for ln in open(NOTES, encoding="utf-8") if ln.startswith("## "))
    except Exception:
        return 0


def status():
    structured = [p for p in _pages() if os.path.basename(p) != "notes.md"]
    notes = _note_count()
    empty = not structured and not notes
    if empty:
        msg = ("Kho tri thức công ty đang **trống**. Đây là nơi lưu tài liệu dùng chung cả team "
               "(quy trình, chính sách, hồ sơ sản phẩm…). Anh/chị (admin) có thể nhắn em "
               "\"ghi vào kho tri thức: …\" để em lưu lại — em đọc được ngay.")
    else:
        bits = []
        if structured:
            bits.append("%d trang" % len(structured))
        if notes:
            bits.append("%d ghi chú tay" % notes)
        msg = "Kho tri thức công ty có " + " + ".join(bits) + "."
    return {"ok": True, "pages": len(structured), "notes": notes, "empty": empty, "message": msg}


def search(kw):
    kw = (kw or "").strip().lower()
    hits = []
    if kw:
        for p in _pages():
            hay = (_title(p) + "\n" + _safe_read(p)).lower()
            if kw in hay:
                hits.append({"path": os.path.relpath(p, ROOT), "title": _title(p)})
    return {"ok": True, "hits": hits[:20],
            "message": ("Tìm thấy %d trang." % len(hits)) if hits
            else "Chưa có trang nào khớp trong kho tri thức công ty."}


def _safe_read(path):
    try:
        return open(path, encoding="utf-8").read()
    except Exception:
        return ""


def capture(text):
    text = (text or "").strip()
    if not text:
        return {"ok": False, "reason": "empty",
                "message": "Anh/chị muốn em ghi nội dung gì vào kho tri thức công ty?"}
    if not is_admin():
        return {"ok": False, "reason": "admin_only",
                "message": "Kho tri thức công ty dùng chung cả team nên chỉ **admin/CEO** mới thêm được. "
                           "Anh/chị nhờ admin ghi giúp nhé."}
    os.makedirs(KB, exist_ok=True)
    uid, rec = _user_record()
    now = datetime.datetime.now()
    author = rec.get("display_name") or uid or "?"
    new = not os.path.isfile(NOTES)
    with open(NOTES, "a", encoding="utf-8") as f:
        if new:
            f.write("# Ghi chú tri thức công ty (nạp tay)\n"
                    "> Ghi nhanh qua chat (admin). Trợ lý đọc được ngay. "
                    "Bản Pro+ sẽ gộp các ghi chú này vào trang có cấu trúc.\n")
        f.write("\n## %s — %s\n%s\n" % (now.strftime("%Y-%m-%d %H:%M"), author, text))
    return {"ok": True, "file": os.path.relpath(NOTES, ROOT),
            "message": "Đã lưu vào kho tri thức công ty ✅ — em đọc lại được ngay. "
                       "Anh/chị cứ nhắn thêm bất cứ điều gì muốn cả team dùng chung."}


def main():
    a = sys.argv[1:]
    cmd = a[0] if a else "status"
    if cmd == "status":
        out = status()
    elif cmd == "search":
        out = search(a[1] if len(a) > 1 else "")
    elif cmd == "capture":
        out = capture(a[1] if len(a) > 1 else "")
    else:
        out = {"ok": False, "message": "unknown cmd: %s" % cmd}
    print(json.dumps(out, ensure_ascii=False, indent=2))
    return 0


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