feat(HEL-382): 搭建 datahub 底座和盘后正式数据链路

新增独立 xiaobai-datahub 服务(SQLite WAL、Tushare 盘后发布、/v1 契约和管理后台),不改现站页面与数据链路。

Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: multica-agent <github@multica.ai>
This commit is contained in:
总工
2026-09-02 12:05:26 +08:00
co-authored by Cursor multica-agent
parent c2ebc0ab91
commit 3498dd7a4b
52 changed files with 4259 additions and 0 deletions
+190
View File
@@ -0,0 +1,190 @@
from __future__ import annotations
import base64
import hashlib
import hmac
import os
import secrets
from datetime import timedelta
from typing import Any
from datahub.crypto import SecretVault, mask_secret
from datahub.db import HubDB
from datahub.timeutil import isoformat, now_shanghai
PBKDF2_ROUNDS = 200_000
SESSION_HOURS = 12
LOGIN_FAIL_LIMIT = 5
LOCK_MINUTES = 10
def hash_password(password: str, salt: bytes | None = None) -> tuple[str, str]:
raw_salt = salt or os.urandom(16)
digest = hashlib.pbkdf2_hmac("sha256", password.encode("utf-8"), raw_salt, PBKDF2_ROUNDS, dklen=32)
return (
base64.urlsafe_b64encode(raw_salt).decode("ascii"),
base64.urlsafe_b64encode(digest).decode("ascii"),
)
def verify_password(password: str, salt_text: str, expected_hash: str) -> bool:
try:
salt = base64.urlsafe_b64decode(salt_text.encode("ascii"))
_, actual = hash_password(password, salt)
except (ValueError, TypeError):
return False
return hmac.compare_digest(actual, expected_hash)
def token_hash(token: str) -> str:
return hashlib.sha256(token.encode("utf-8")).hexdigest()
class AuthService:
def __init__(self, db: HubDB, vault: SecretVault, api_token: str, admin_password: str) -> None:
self.db = db
self.vault = vault
self._bootstrap(api_token, admin_password)
def _bootstrap(self, api_token: str, admin_password: str) -> None:
if api_token:
existing = self.db.fetchone("SELECT token_hash FROM api_tokens WHERE name = ?", ("review",))
hashed = token_hash(api_token)
last4 = mask_secret(api_token)
if existing is None:
self.db.execute(
"INSERT INTO api_tokens(token_hash, name, last4, created_at) VALUES (?,?,?,?)",
(hashed, "review", last4, isoformat()),
)
elif existing["token_hash"] != hashed:
self.db.execute(
"UPDATE api_tokens SET token_hash = ?, last4 = ? WHERE name = ?",
(hashed, last4, "review"),
)
admin = self.db.fetchone("SELECT id FROM hub_admin WHERE username = ?", ("hub_admin",))
if admin is None and admin_password:
salt, hashed = hash_password(admin_password)
now = isoformat()
self.db.execute(
"""
INSERT INTO hub_admin(username, password_salt, password_hash, password_must_change, created_at, updated_at)
VALUES (?, ?, ?, 1, ?, ?)
""",
("hub_admin", salt, hashed, now, now),
)
def check_api_token(self, supplied: str) -> bool:
if not supplied:
return False
row = self.db.fetchone(
"SELECT token_hash FROM api_tokens WHERE token_hash = ? AND revoked_at IS NULL",
(token_hash(supplied),),
)
return row is not None
def login(self, username: str, password: str) -> dict[str, Any]:
user = self.db.fetchone("SELECT * FROM hub_admin WHERE username = ?", (username,))
if not user:
raise PermissionError("账号或密码错误")
now = now_shanghai()
locked_until = user.get("locked_until")
if locked_until:
try:
from datetime import datetime
if datetime.fromisoformat(str(locked_until)) > now:
raise PermissionError("账号已锁定,请稍后再试")
except ValueError:
pass
if not verify_password(password, str(user["password_salt"]), str(user["password_hash"])):
fails = int(user["failed_attempts"] or 0) + 1
lock = isoformat(now + timedelta(minutes=LOCK_MINUTES)) if fails >= LOGIN_FAIL_LIMIT else None
self.db.execute(
"UPDATE hub_admin SET failed_attempts = ?, locked_until = ? WHERE id = ?",
(fails, lock, user["id"]),
)
raise PermissionError("账号或密码错误")
self.db.execute(
"UPDATE hub_admin SET failed_attempts = 0, locked_until = NULL WHERE id = ?",
(user["id"],),
)
session = secrets.token_urlsafe(32)
csrf = secrets.token_urlsafe(24)
expires = isoformat(now + timedelta(hours=SESSION_HOURS))
self.db.execute(
"INSERT INTO hub_sessions(token_hash, csrf_token, expires_at, created_at) VALUES (?,?,?,?)",
(token_hash(session), csrf, expires, isoformat(now)),
)
return {
"session": session,
"csrf": csrf,
"must_change": bool(user["password_must_change"]),
"expires_at": expires,
}
def session_user(self, raw_token: str) -> dict[str, Any] | None:
if not raw_token:
return None
row = self.db.fetchone(
"SELECT * FROM hub_sessions WHERE token_hash = ?",
(token_hash(raw_token),),
)
if not row:
return None
if str(row["expires_at"]) < isoformat():
self.db.execute("DELETE FROM hub_sessions WHERE token_hash = ?", (row["token_hash"],))
return None
admin = self.db.fetchone("SELECT username, password_must_change FROM hub_admin WHERE username = ?", ("hub_admin",))
return {
"username": (admin or {}).get("username") or "hub_admin",
"csrf_token": row["csrf_token"],
"must_change": bool((admin or {}).get("password_must_change")),
"token_hash": row["token_hash"],
}
def logout(self, raw_token: str) -> None:
if raw_token:
self.db.execute("DELETE FROM hub_sessions WHERE token_hash = ?", (token_hash(raw_token),))
def change_password(self, current: str, new_password: str) -> None:
if len(new_password) < 8:
raise ValueError("新密码至少 8 位")
user = self.db.fetchone("SELECT * FROM hub_admin WHERE username = ?", ("hub_admin",))
if not user or not verify_password(current, str(user["password_salt"]), str(user["password_hash"])):
raise PermissionError("当前密码错误")
salt, hashed = hash_password(new_password)
self.db.execute(
"UPDATE hub_admin SET password_salt=?, password_hash=?, password_must_change=0, updated_at=? WHERE id=?",
(salt, hashed, isoformat(), user["id"]),
)
def confirm_password(self, password: str) -> bool:
user = self.db.fetchone("SELECT * FROM hub_admin WHERE username = ?", ("hub_admin",))
if not user:
return False
return verify_password(password, str(user["password_salt"]), str(user["password_hash"]))
def credential_status(self, name: str) -> dict[str, Any]:
row = self.db.fetchone("SELECT last4, updated_at FROM credentials WHERE name = ?", (name,))
if not row:
return {"configured": False, "last4": "", "updated_at": ""}
return {"configured": True, "last4": row["last4"], "updated_at": row["updated_at"]}
def store_credential(self, name: str, secret: str) -> None:
payload = self.vault.encrypt_json({name: secret})
self.db.execute(
"""
INSERT INTO credentials(name, encrypted_payload, last4, updated_at)
VALUES (?, ?, ?, ?)
ON CONFLICT(name) DO UPDATE SET
encrypted_payload=excluded.encrypted_payload, last4=excluded.last4, updated_at=excluded.updated_at
""",
(name, payload, mask_secret(secret), isoformat()),
)
def load_credential(self, name: str) -> str:
row = self.db.fetchone("SELECT encrypted_payload FROM credentials WHERE name = ?", (name,))
if not row:
return ""
data = self.vault.decrypt_json(str(row["encrypted_payload"]))
return str(data.get(name) or "")