Question

我正在尝试建立一个优雅地将数据库操作推迟到单独线程的系统，以避免在Twisted回调期间阻塞。

到目前为止，这是我的方法：

from contextlib import contextmanager

from sqlalchemy import create_engine
from sqlalchemy.orm import scoped_session, sessionmaker

from twisted.internet.threads import deferToThread

_engine = create_engine(initialization_string)
Session = scoped_session(sessionmaker(bind=_engine))


@contextmanager
def transaction_context():
    session = Session()
    try:
        yield session
        session.commit()
    except:
        # No need to do session.rollback().  session.remove will do it.
        raise
    finally:
        session.remove()


def threaded(fn):
    @wraps(fn)
    def wrapper(*args, **kwargs):
        return deferToThread(fn, *args, **kwargs)
    return wrapper

这应该允许我用threaded装饰器包装一个函数，然后在所述函数体中使用transaction_context上下文管理器。以下是一个例子：

from __future__ import print_function
from my_lib.orm import User, transaction_context, threaded
from twisted.internet import reactor


@threaded
def get_n_users(n):
    with transaction_context() as session:
        return session.query(User).limit(n).all()

if __name__ == '__main__':
    get_n_users(n).addBoth(len)
    reactor.run()

但是，当我运行上面的脚本时，我得到一个包含以下回溯的失败：

Unhandled error in Deferred:
Unhandled Error
Traceback (most recent call last):
  File "/usr/lib/python2.7/threading.py", line 781, in __bootstrap
    self.__bootstrap_inner()
  File "/usr/lib/python2.7/threading.py", line 808, in __bootstrap_inner
    self.run()
  File "/usr/lib/python2.7/threading.py", line 761, in run
    self.__target(*self.__args, **self.__kwargs)
--- <exception caught here> ---
  File "/usr/local/lib/python2.7/dist-packages/twisted/python/threadpool.py", line 191, in _worker
    result = context.call(ctx, function, *args, **kwargs)
  File "/usr/local/lib/python2.7/dist-packages/twisted/python/context.py", line 118, in callWithContext
    return self.currentContext().callWithContext(ctx, func, *args, **kw)
  File "/usr/local/lib/python2.7/dist-packages/twisted/python/context.py", line 81, in callWithContext
    return func(*args,**kw)
  File "testaccess.py", line 9, in get_n_users
    return session.query(User).limit(n).all()
  File "/usr/lib/python2.7/contextlib.py", line 24, in __exit__
    self.gen.next()
  File "/home/louis/Documents/Python/knacki/knacki/db.py", line 36, in transaction_context
    session.remove()
exceptions.AttributeError: 'Session' object has no attribute 'remove'

我根本没想到这一点。我错过了什么？我没有正确地实例化scoped_session吗？

修改： Here是关于将此设置与Twisted集成的相关问题。它可能有助于澄清我正在努力实现的目标。

Answer 1

简短回答

在.remove()上致电Session，而不是session。

答案很长：

scoped_session并没有真正返回Session课程。相反，它创建了一个对象，它关注它所调用的线程。调用它将返回与该线程关联的现有Session实例或关联新线程并返回该实例。 thread local是线程与会话关联的原因。

remove对象上的scoped_session方法删除当前与其调用的线程关联的会话对象。这意味着它与scoped_session.__call__相反，这是一种令人困惑的API。

这是一个简短的Python脚本来说明行为。

import threading
from sqlalchemy import create_engine
from sqlalchemy.orm import scoped_session, sessionmaker

_engine = create_engine('sqlite:///:memory:')
Session = scoped_session(sessionmaker(_engine))


def scoped_session_demo(remove=False):
    ids = []

    def push_ids():
        thread_name = threading.currentThread().getName()
        data = [thread_name]

        data.append(Session())
        if remove:
            Session.remove()
        data.append(Session())

        ids.append(data)

    t = threading.Thread(target=push_ids)
    t.start()
    t.join()

    push_ids()

    sub_thread, main_thread = ids

    sub_name, sub_session_a, sub_session_b = sub_thread
    main_name, main_session_a, main_session_b = main_thread

    print sub_name, sub_session_a == sub_session_b
    print main_name, main_session_a == main_session_b
    print sub_name, '==', main_name, sub_session_a == main_session_b


print 'Without remove:'
scoped_session_demo()
print 'With remove:'
scoped_session_demo(True)

其输出：

Without remove:
Thread-1 True
MainThread True
Thread-1 == MainThread False
With remove:

Thread-2 False
MainThread False
Thread-2 == MainThread False

为什么我的scoped_session引发了一个AttributeError：'Session'对象没有属性'remove'

1 个答案: