User: Password:
Subscribe / Log in / New account

A proposal for "rebooted" Python asynchronous I/O support

From:  Guido van Rossum <>
To:  Python-Dev <>, Python-Ideas <>
Subject:  PEP 3156 - Asynchronous IO Support Rebooted
Date:  Fri, 21 Dec 2012 10:57:12 -0800
Message-ID:  <>
Archive-link:  Article

Dear python-dev *and* python-ideas,

I am posting PEP 3156 here for early review and discussion. As you can
see from the liberally sprinkled TBD entries it is not done, but I am
about to disappear on vacation for a few weeks and I am reasonably
happy with the state of things so far. (Of course feedback may change
this. :-) Also, there has already been some discussion on python-ideas
(and even on Twitter) so I don't want python-dev to feel out of the
loop -- this *is* a proposal for a new standard library module. (But
no, I haven't picked the module name yet. :-)

There's an -- also incomplete -- reference implementation at -- unlike the first version of tulip,
this version actually has (some) unittests.

Let the bikeshedding begin!

(Oh, happy holidays too. :-)

--Guido van Rossum (
PEP: 3156
Title: Asynchronous IO Support Rebooted
Version: $Revision$
Last-Modified: $Date$
Author: Guido van Rossum <>
Status: Draft
Type: Standards Track
Content-Type: text/x-rst
Created: 12-Dec-2012
Post-History: TBD


This is a proposal for asynchronous I/O in Python 3, starting with
Python 3.3.  Consider this the concrete proposal that is missing from
PEP 3153.  The proposal includes a pluggable event loop API, transport
and protocol abstractions similar to those in Twisted, and a
higher-level scheduler based on ``yield from`` (PEP 380).  A reference
implementation is in the works under the code name tulip.


The event loop is the place where most interoperability occurs.  It
should be easy for (Python 3.3 ports of) frameworks like Twisted,
Tornado, or ZeroMQ to either adapt the default event loop
implementation to their needs using a lightweight wrapper or proxy, or
to replace the default event loop implementation with an adaptation of
their own event loop implementation.  (Some frameworks, like Twisted,
have multiple event loop implementations.  This should not be a
problem since these all have the same interface.)

It should even be possible for two different third-party frameworks to
interoperate, either by sharing the default event loop implementation
(each using its own adapter), or by sharing the event loop
implementation of either framework.  In the latter case two levels of
adaptation would occur (from framework A's event loop to the standard
event loop interface, and from there to framework B's event loop).
Which event loop implementation is used should be under control of the
main program (though a default policy for event loop selection is

Thus, two separate APIs are defined:

- getting and setting the current event loop object
- the interface of a conforming event loop and its minimum guarantees

An event loop implementation may provide additional methods and

The event loop interface does not depend on ``yield from``.  Rather, it
uses a combination of callbacks, additional interfaces (transports and
protocols), and Futures.  The latter are similar to those defined in
PEP 3148, but have a different implementation and are not tied to
threads.  In particular, they have no wait() method; the user is
expected to use callbacks.

For users (like myself) who don't like using callbacks, a scheduler is
provided for writing asynchronous I/O code as coroutines using the PEP
380 ``yield from`` expressions.  The scheduler is not pluggable;
pluggability occurs at the event loop level, and the scheduler should
work with any conforming event loop implementation.

For interoperability between code written using coroutines and other
async frameworks, the scheduler has a Task class that behaves like a
Future.  A framework that interoperates at the event loop level can
wait for a Future to complete by adding a callback to the Future.
Likewise, the scheduler offers an operation to suspend a coroutine
until a callback is called.

Limited interoperability with threads is provided by the event loop
interface; there is an API to submit a function to an executor (see
PEP 3148) which returns a Future that is compatible with the event


Interoperability with systems like Stackless Python or
greenlets/gevent is not a goal of this PEP.



Python 3.3 is required.  No new language or standard library features
beyond Python 3.3 are required.  No third-party modules or packages
are required.

Module Namespace

The specification here will live in a new toplevel package.  Different
components will live in separate submodules of that package.  The
package will import common APIs from their respective submodules and
make them available as package attributes (similar to the way the
email package works).

The name of the toplevel package is currently unspecified.  The
reference implementation uses the name 'tulip', but the name will
change to something more boring if and when the implementation is
moved into the standard library (hopefully for Python 3.4).

Until the boring name is chosen, this PEP will use 'tulip' as the
toplevel package name.  Classes and functions given without a module
name are assumed to be accessed via the toplevel package.

Event Loop Policy: Getting and Setting the Event Loop

To get the current event loop, use ``get_event_loop()``.  This returns
an instance of the ``EventLoop`` class defined below or an equivalent
object.  It is possible that ``get_event_loop()`` returns a different
object depending on the current thread, or depending on some other
notion of context.

To set the current event loop, use ``set_event_loop(event_loop)``,
where ``event_loop`` is an instance of the ``EventLoop`` class or
equivalent.  This uses the same notion of context as

For the benefit of unit tests and other special cases there's a third
policy function: ``init_event_loop()``, which creates a new EventLoop
instance and calls ``set_event_loop()`` with it.  TBD: Maybe we should
have a ``create_default_event_loop_instance()`` function instead?

To change the way the above three functions work
(including their notion of context), call
``set_event_loop_policy(policy)``, where ``policy`` is an event loop
policy object.  The policy object can be any object that has methods
``get_event_loop()``, ``set_event_loop(event_loop)``
and ``init_event_loop()`` behaving like
the functions described above.  The default event loop policy is an
instance of the class ``DefaultEventLoopPolicy``.  The current event loop
policy object can be retrieved by calling ``get_event_loop_policy()``.

An event loop policy may but does not have to enforce that there is
only one event loop in existence.  The default event loop policy does
not enforce this, but it does enforce that there is only one event
loop per thread.

Event Loop Interface

(A note about times: as usual in Python, all timeouts, intervals and
delays are measured in seconds, and may be ints or floats.  The
accuracy and precision of the clock are up to the implementation; the
default implementation uses ``time.monotonic()``.)

A conforming event loop object has the following methods:

- ``run()``.  Runs the event loop until there is nothing left to do.
  This means, in particular:

  - No more calls scheduled with ``call_later()``,
  ``call_repeatedly()``, ``call_soon()``, or
  ``call_soon_threadsafe()``, except for cancelled calls.

  - No more registered file descriptors.  It is up to the registering
    party to unregister a file descriptor when it is closed.

  Note: ``run()`` blocks until the termination condition is met,
  or until ``stop()`` is called.

  Note: if you schedule a call with ``call_repeatedly()``, ``run()``
  will not exit until you cancel it.

  TBD: How many variants of this do we really need?

- ``stop()``.  Stops the event loop as soon as it is convenient.  It
  is fine to restart the loop with ``run()`` (or one of its variants)

  Note: How soon exactly is up to the implementation.  All immediate
  callbacks that were already scheduled to run before ``stop()`` is
  called must still be run, but callbacks scheduled after it is called
  (or scheduled to be run later) will not be run.

- ``run_forever()``.  Runs the event loop until ``stop()`` is called.

- ``run_until_complete(future, timeout=None)``.  Runs the event loop
  until the Future is done.  If a timeout is given, it waits at most
  that long.  If the Future is done, its result is returned, or its
  exception is raised; if the timeout expires before the Future is
  done, or if ``stop()`` is called, ``TimeoutError`` is raised (but
  the Future is not cancelled).  This cannot be called when the event
  loop is already running.
  Note: This API is most useful for tests and the like.  It should not
  be used as a substitute for ``yield from future`` or other ways to
  wait for a Future (e.g. registering a done callback).

- ``run_once(timeout=None)``.  Run the event loop for a little while.
  If a timeout is given, an I/O poll made will block at most that
  long; otherwise, an I/O poll is not constrained in time.

  Note: Exactlly how much work this does is up to the implementation.
  One constraint: if a callback immediately schedules itself using
  ``call_soon()``, causing an infinite loop, ``run_once()`` should
  still return.

- ``call_later(delay, callback, *args)``.  Arrange for
  ``callback(*args)`` to be called approximately ``delay`` seconds in
  the future, once, unless cancelled.  Returns
  a ``Handler`` object representing the callback, whose
  ``cancel()`` method can be used to cancel the callback.

- ``call_repeatedly(interval, callback, **args)``.  Like ``call_later()``
  but calls the callback repeatedly, every ``interval`` seconds,
  until the ``Handler`` returned is cancelled.  The first call is in
  ``interval`` seconds.

- ``call_soon(callback, *args)``.  Equivalent to ``call_later(0,
  callback, *args)``.

- ``call_soon_threadsafe(callback, *args)``.  Like
  ``call_soon(callback, *args)``, but when called from another thread
  while the event loop is blocked waiting for I/O, unblocks the event
  loop.  This is the *only* method that is safe to call from another
  thread or from a signal handler.  (To schedule a callback for a
  later time in a threadsafe manner, you can use
  ``ev.call_soon_threadsafe(ev.call_later, when, callback, *args)``.)

- TBD: A way to register a callback that is already wrapped in a
  ``Handler``.  Maybe ``call_soon()`` could just check
  ``isinstance(callback, Handler)``?  It should silently skip
  a cancelled callback.

Some methods in the standard conforming interface return Futures:

- ``wrap_future(future)``.  This takes a PEP 3148 Future (i.e., an
  instance of ``concurrent.futures.Future``) and returns a Future
  compatible with the event loop (i.e., a ``tulip.Future`` instance).

- ``run_in_executor(executor, function, *args)``.  Arrange to call
  ``function(*args)`` in an executor (see PEP 3148).  Returns a Future
  whose result on success is the return value that call.  This is
  equivalent to ``wrap_future(executor.submit(function, *args))``.  If
  ``executor`` is ``None``, a default ``ThreadPoolExecutor`` with 5
  threads is used.  (TBD: Should the default executor be shared
  between different event loops?  Should we even have a default
  executor?  Should be be able to set its thread count?  Shoul we even
  have this method?)

- ``set_default_executor(executor)``.  Set the default executor used
  by ``run_in_executor()``.

- ``getaddrinfo(host, port, family=0, type=0, proto=0, flags=0)``.
  Similar to the ``socket.getaddrinfo()`` function but returns a
  Future.  The Future's result on success will be a list of the same
  format as returned by ``socket.getaddrinfo()``.  The default
  implementation calls ``socket.getaddrinfo()`` using
  ``run_in_executor()``, but other implementations may choose to
  implement their own DNS lookup.

- ``getnameinfo(sockaddr, flags=0)``.  Similar to
  ``socket.getnameinfo()`` but returns a Future.  The Future's result
  on success will be a tuple ``(host, port)``.  Same implementation
  remarks as for ``getaddrinfo()``.

- ``create_transport(protocol_factory, host, port, **kwargs)``.
  Creates a transport and a protocol and ties them together.  Returns
  a Future whose result on success is a (transport, protocol) pair.
  Note that when the Future completes, the protocol's
  ``connection_made()`` method has not yet been called; that will
  happen when the connection handshake is complete.  When it is
  impossible to connect to the given host and port, the Future will
  raise an exception instead.

  Optional keyword arguments:

  - ``family``, ``type``, ``proto``, ``flags``: Address familty,
    socket type, protcol, and miscellaneous flags to be passed through
    to ``getaddrinfo()``.  These all default to ``0`` except ``type``
    which defaults to ``socket.SOCK_STREAM``.

  - ``ssl``: Pass ``True`` to create an SSL transport (by default a
    plain TCP is created).  Or pass an ``ssl.SSLContext`` object to
    override the default SSL context object to be used.

  TBD: Should this be called create_connection()?

- ``start_serving(...)``.  Enters a loop that accepts connections.
  TBD: Signature.  There are two possibilities:

  1. You pass it a non-blocking socket that you have already prepared
     with ``bind()`` and ``listen()`` (these system calls do not block
     AFAIK), a protocol factory (I hesitate to use this word :-), and
     optional flags that control the transport creation (e.g. ssl).

  2. Instead of a socket, you pass it a host and port, and some more
     optional flags (e.g. to control IPv4 vs IPv6, or to set the
     backlog value to be passed to ``listen()``).

  In either case, once it has a socket, it will wrap it in a
  transport, and then enter a loop accepting connections (the best way
  to implement such a loop depends on the platform).  Each time a
  connection is accepted, a transport and protocol are created for it.

  This should return an object that can be used to control the serving
  loop, e.g. to stop serving, abort all active connections, and (if
  supported) adjust the backlog or other parameters.  It may also have
  an API to inquire about active connections.  If version (2) is
  selected, it should probably return a Future whose result on success
  will be that control object, and which becomes done once the accept
  loop is started.

  TBD: It may be best to use version (2), since on some platforms the
  best way to start a server may not involve sockets (but will still
  involve transports and protocols).

  TBD: Be more specific.

TBD: Some platforms may not be interested in implementing all of
these, e.g. start_serving() may be of no interest to mobile apps.
(Although, there's a Minecraft server on my iPad...)

The following methods for registering callbacks for file descriptors
are optional.  If they are not implemented, accessing the method
(without calling it) returns AttributeError.  The default
implementation provides them but the user normally doesn't use these
directly -- they are used by the transport implementations
exclusively.  Also, on Windows these may be present or not depending
on whether a select-based or IOCP-based event loop is used.  These
take integer file descriptors only, not objects with a fileno()
method.  The file descriptor should represent something pollable --
i.e. no disk files.

- ``add_reader(fd, callback, *args)``.  Arrange for
  ``callback(*args)`` to be called whenever file descriptor ``fd`` is
  ready for reading.  Returns a ``Handler`` object which can be
  used to cancel the callback.  Note that, unlike ``call_later()``,
  the callback may be called many times.  Calling ``add_reader()``
  again for the same file descriptor implicitly cancels the previous
  callback for that file descriptor.  (TBD: Returning a
  ``Handler`` that can be cancelled seems awkward.  Let's forget
  about that.)  (TBD: Change this to raise an exception if a handler
  is already set.)

- ``add_writer(fd, callback, *args)``.  Like ``add_reader()``,
  but registers the callback for writing instead of for reading.

- ``remove_reader(fd)``.  Cancels the current read callback for file
  descriptor ``fd``, if one is set.  A no-op if no callback is
  currently set for the file descriptor.  (The reason for providing
  this alternate interface is that it is often more convenient to
  remember the file descriptor than to remember the ``Handler``
  object.)  (TBD: Return ``True`` if a handler was removed, ``False``
  if not.)

- ``remove_writer(fd)``.  This is to ``add_writer()`` as
  ``remove_reader()`` is to ``add_reader()``.

- ``add_connector(fd, callback, *args)``.  Like ``add_writer()`` but
  meant to wait for ``connect()`` operations, which on some platforms
  require different handling (e.g. ``WSAPoll()`` on Windows).

- ``remove_connector(fd)``.  This is to ``remove_writer()`` as
  ``add_connector()`` is to ``add_writer()``.

TBD: What about multiple callbacks per fd?  The current semantics is
that ``add_reader()/add_writer()`` replace a previously registered
callback.  Change this to raise an exception if a callback is already

The following methods for doing async I/O on sockets are optional.
They are alternative to the previous set of optional methods, intended
for transport implementations on Windows using IOCP (if the event loop
supports it).  The socket argument has to be a non-blocking socket.

- ``sock_recv(sock, n)``.  Receive up to ``n`` bytes from socket
  ``sock``.  Returns a Future whose result on success will be a
  bytes object on success.

- ``sock_sendall(sock, data)``.  Send bytes ``data`` to the socket
  ``sock``.  Returns a Future whose result on success will be
  ``None``.  (TBD: Is it better to emulate ``sendall()`` or ``send()``
  semantics?  I think ``sendall()`` -- but perhaps it should still
  be *named* ``send()``?)

- ``sock_connect(sock, address)``.  Connect to the given address.
  Returns a Future whose result on success will be ``None``.

- ``sock_accept(sock)``.  Accept a connection from a socket.  The
  socket must be in listening mode and bound to an address.  Returns a
  Future whose result on success will be a tuple ``(conn, peer)``
  where ``conn`` is a connected non-blocking socket and ``peer`` is
  the peer address.  (TBD: People tell me that this style of API is
  too slow for high-volume servers.  So there's also
  ``start_serving()`` above.  Then do we still need this?)

TBD: Optional methods are not so good.  Perhaps these should be
required?  It may still depend on the platform which set is more

Callback Sequencing

When two callbacks are scheduled for the same time, they are run
in the order in which they are registered.  For example::


guarantees that ``foo()`` is called before ``bar()``.

If ``call_soon()`` is used, this guarantee is true even if the system
clock were to run backwards.  This is also the case for
``call_later(0, callback, *args)``.  However, if ``call_later()`` is
used with a nonzero delay, all bets are off if the system
clock were to runs backwards.  (A good event loop implementation
should use ``time.monotonic()`` to avoid problems when the clock runs
backward.  See PEP 418.)


All event loops have a notion of context.  For the default event loop
implementation, the context is a thread.  An event loop implementation
should run all callbacks in the same context.  An event loop
implementation should run only one callback at a time, so callbacks
can assume automatic mutual exclusion with other callbacks scheduled
in the same event loop.


There are two categories of exceptions in Python: those that derive
from the ``Exception`` class and those that derive from
``BaseException``.  Exceptions deriving from ``Exception`` will
generally be caught and handled appropriately; for example, they will
be passed through by Futures, and they will be logged and ignored when
they occur in a callback.

However, exceptions deriving only from ``BaseException`` are never
caught, and will usually cause the program to terminate with a
traceback.  (Examples of this category include ``KeyboardInterrupt``
and ``SystemExit``; it is usually unwise to treat these the same as
most other exceptions.)

The Handler Class

The various methods for registering callbacks (e.g. ``call_later()``)
all return an object representing the registration that can be used to
cancel the callback.  For want of a better name this object is called
a ``Handler``, although the user never needs to instantiate
instances of this class.  There is one public method:

- ``cancel()``.  Attempt to cancel the callback.
  TBD: Exact specification.

Read-only public attributes:

- ``callback``.  The callback function to be called.

- ``args``.  The argument tuple with which to call the callback function.

- ``cancelled``.  True if ``cancel()`` has been called.

Note that some callbacks (e.g. those registered with ``call_later()``)
are meant to be called only once.  Others (e.g. those registered with
``add_reader()``) are meant to be called multiple times.

TBD: An API to call the callback (encapsulating the exception handling
necessary)?  Should it record how many times it has been called?
Maybe this API should just be ``__call__()``?  (But it should suppress

TBD: Public attribute recording the realtime value when the callback
is scheduled?  (Since this is needed anyway for storing it in a heap.)


The ``tulip.Future`` class here is intentionally similar to the
``concurrent.futures.Future`` class specified by PEP 3148, but there
are slight differences.  The supported public API is as follows,
indicating the differences with PEP 3148:

- ``cancel()``.
  TBD: Exact specification.

- ``cancelled()``.

- ``running()``.  Note that the meaning of this method is essentially
  "cannot be cancelled and isn't done yet".  (TBD: Would be nice if
  this could be set *and* cleared in some cases, e.g. sock_recv().)

- ``done()``.

- ``result()``.  Difference with PEP 3148: This has no timeout
  argument and does *not* wait; if the future is not yet done, it
  raises an exception.

- ``exception()``.  Difference with PEP 3148: This has no timeout
  argument and does *not* wait; if the future is not yet done, it
  raises an exception.

- ``add_done_callback(fn)``.  Difference with PEP 3148: The callback
  is never called immediately, and always in the context of the
  caller.  (Typically, a context is a thread.)  You can think of this
  as calling the callback through ``call_soon_threadsafe()``.  Note
  that the callback (unlike all other callbacks defined in this PEP,
  and ignoring the convention from the section "Callback Style" below)
  is always called with a single argument, the Future object.

The internal methods defined in PEP 3148 are not supported.  (TBD:
Maybe we do need to support these, in order to make it easy to write
user code that returns a Future?)

A ``tulip.Future`` object is not acceptable to the ``wait()`` and
``as_completed()`` functions in the ``concurrent.futures`` package.

A ``tulip.Future`` object is acceptable to a ``yield from`` expression
when used in a coroutine.  This is implemented through the
``__iter__()`` interface on the Future.  See the section "Coroutines
and the Scheduler" below.

When a Future is garbage-collected, if it has an associated exception
but neither ``result()`` nor ``exception()`` nor ``__iter__()`` has
ever been called (or the latter hasn't raised the exception yet --
details TBD), the exception should be logged.  TBD: At what level?

In the future (pun intended) we may unify ``tulip.Future`` and
``concurrent.futures.Future``, e.g. by adding an ``__iter__()`` method
to the latter that works with ``yield from``.  To prevent accidentally
blocking the event loop by calling e.g. ``result()`` on a Future
that's not don yet, the blocking operation may detect that an event
loop is active in the current thread and raise an exception instead.
However the current PEP strives to have no dependencies beyond Python
3.3, so changes to ``concurrent.futures.Future`` are off the table for


A transport is an abstraction on top of a socket or something similar
(for example, a UNIX pipe or an SSL connection).  Transports are
strongly influenced by Twisted and PEP 3153.  Users rarely implement
or instantiate transports -- rather, event loops offer utility methods
to set up transports.

Transports work in conjunction with protocols.  Protocols are
typically written without knowing or caring about the exact type of
transport used, and transports can be used with a wide variety of
protocols.  For example, an HTTP client protocol implementation may be
used with either a plain socket transport or an SSL transport.  The
plain socket transport can be used with many different protocols
besides HTTP (e.g. SMTP, IMAP, POP, FTP, IRC, SPDY).

Most connections have an asymmetric nature: the client and server
usually have very different roles and behaviors.  Hence, the interface
between transport and protocol is also asymmetric.  From the
protocol's point of view, *writing* data is done by calling the
``write()`` method on the transport object; this buffers the data and
returns immediately.  However, the transport takes a more active role
in *reading* data: whenever some data is read from the socket (or
other data source), the transport calls the protocol's
``data_received()`` method.

Transports have the following public methods:

- ``write(data)``.  Write some bytes.  The argument must be a bytes
  object.  Returns ``None``.  The transport is free to buffer the
  bytes, but it must eventually cause the bytes to be transferred to
  the entity at the other end, and it must maintain stream behavior.
  That is, ``t.write(b'abc'); t.write(b'def')`` is equivalent to
  ``t.write(b'abcdef')``, as well as to::


- ``writelines(iterable)``.  Equivalent to::

    for data in iterable:

- ``write_eof()``.  Close the writing end of the connection.
  Subsequent calls to ``write()`` are not allowed.  Once all buffered
  data is transferred, the transport signals to the other end that no
  more data will be received.  Some protocols don't support this
  operation; in that case, calling ``write_eof()`` will raise an
  exception.  (Note: This used to be called ``half_close()``, but
  unless you already know what it is for, that name doesn't indicate
  *which* end is closed.)

- ``can_write_eof()``.  Return ``True`` if the protocol supports
  ``write_eof()``, ``False`` if it does not.  (This method is needed
  because some protocols need to change their behavior when
  ``write_eof()`` is unavailable.  For example, in HTTP, to send data
  whose size is not known ahead of time, the end of the data is
  typically indicated using ``write_eof()``; however, SSL does not
  support this, and an HTTP protocol implementation would have to use
  the "chunked" transfer encoding in this case.  But if the data size
  is known ahead of time, the best approach in both cases is to use
  the Content-Length header.)

- ``pause()``.  Suspend delivery of data to the protocol until a
  subsequent ``resume()`` call.  Between ``pause()`` and ``resume()``,
  the protocol's ``data_received()`` method will not be called.  This
  has no effect on ``write()``.

- ``resume()``.  Restart delivery of data to the protocol via

- ``close()``.  Sever the connection with the entity at the other end.
  Any data buffered by ``write()`` will (eventually) be transferred
  before the connection is actually closed.  The protocol's
  ``data_received()`` method will not be called again.  Once all
  buffered data has been flushed, the protocol's ``connection_lost()``
  method will be called with ``None`` as the argument.  Note that
  this method does not wait for all that to happen.

- ``abort()``.  Immediately sever the connection.  Any data still
  buffered by the transport is thrown away.  Soon, the protocol's
  ``connection_lost()`` method will be called with ``None`` as
  argument.  (TBD: Distinguish in the ``connection_lost()`` argument
  between ``close()``, ``abort()`` or a close initated by the other
  end?  Or add a transport method to inquire about this?  Glyph's
  proposal was to pass different exceptions for this purpose.)

TBD: Provide flow control the other way -- the transport may need to
suspend the protocol if the amount of data buffered becomes a burden.
Proposal: let the transport call ``protocol.pause()`` and
``protocol.resume()`` if they exist; if they don't exist, the
protocol doesn't support flow control.  (Perhaps different names
to avoid confusion between protocols and transports?)


Protocols are always used in conjunction with transports.  While a few
common protocols are provided (e.g. decent though not necessarily
excellent HTTP client and server implementations), most protocols will
be implemented by user code or third-party libraries.

A protocol must implement the following methods, which will be called
by the transport.  Consider these callbacks that are always called by
the event loop in the right context.  (See the "Context" section

- ``connection_made(transport)``.  Indicates that the transport is
  ready and connected to the entity at the other end.  The protocol
  should probably save the transport reference as an instance variable
  (so it can call its ``write()`` and other methods later), and may
  write an initial greeting or request at this point.

- ``data_received(data)``.  The transport has read some bytes from the
  connection.  The argument is always a non-empty bytes object.  There
  are no guarantees about the minimum or maximum size of the data
  passed along this way.  ``p.data_received(b'abcdef')`` should be
  treated exactly equivalent to::


- ``eof_received()``.  This is called when the other end called
  ``write_eof()`` (or something equivalent).  The default
  implementation calls ``close()`` on the transport, which causes
  ``connection_lost()`` to be called (eventually) on the protocol.

- ``connection_lost(exc)``.  The transport has been closed or aborted,
  has detected that the other end has closed the connection cleanly,
  or has encountered an unexpected error.  In the first three cases
  the argument is ``None``; for an unexpected error, the argument is
  the exception that caused the transport to give up.  (TBD: Do we
  need to distinguish between the first three cases?)

Here is a chart indicating the order and multiplicity of calls:

  1. ``connection_made()`` -- exactly once
  2. ``data_received()`` -- zero or more times
  3. ``eof_received()`` -- at most once
  4. ``connection_lost()`` -- exactly once

TBD: Discuss whether user code needs to do anything to make sure that
protocol and transport aren't garbage-collected prematurely.

Callback Style

Most interfaces taking a callback also take positional arguments.  For
instance, to arrange for ``foo("abc", 42)`` to be called soon, you
call ``ev.call_soon(foo, "abc", 42)``.  To schedule the call
``foo()``, use ``ev.call_soon(foo)``.  This convention greatly reduces
the number of small lambdas required in typical callback programming.

This convention specifically does *not* support keyword arguments.
Keyword arguments are used to pass optional extra information about
the callback.  This allows graceful evolution of the API without
having to worry about whether a keyword might be significant to a
callee somewhere.  If you have a callback that *must* be called with a
keyword argument, you can use a lambda or ``functools.partial``.  For

  ev.call_soon(functools.partial(foo, "abc", repeat=42))

Choosing an Event Loop Implementation

TBD.  (This is about the choice to use e.g. select vs. poll vs. epoll,
and how to override the choice.  Probably belongs in the event loop

Coroutines and the Scheduler

This is a separate toplevel section because its status is different
from the event loop interface.  Usage of coroutines is optional, and
it is perfectly fine to write code using callbacks only.  On the other
hand, there is only one implementation of the scheduler/coroutine API,
and if you're using coroutines, that's the one you're using.


A coroutine is a generator that follows certain conventions.  For
documentation purposes, all coroutines should be decorated with
``@tulip.coroutine``, but this cannot be strictly enforced.

Coroutines use the ``yield from`` syntax introduced in PEP 380,
instead of the original ``yield`` syntax.

The word "coroutine", like the word "generator", is used for two
different (though related) concepts:

- The function that defines a coroutine (a function definition
  decorated with ``tulip.coroutine``).  If disambiguation is needed,
  we call this a *coroutine function*.

- The object obtained by calling a coroutine function.  This object
  represents a computation or an I/O operation (usually a combination)
  that will complete eventually.  For disambiguation we call it a
  *coroutine object*.

Things a coroutine can do:

- ``result = yield from future`` -- suspends the coroutine until the
  future is done, then returns the future's result, or raises its
  exception, which will be propagated.

- ``result = yield from coroutine`` -- wait for another coroutine to
  produce a result (or raise an exception, which will be propagated).
  The ``coroutine`` expression must be a *call* to another coroutine.

- ``results = yield from tulip.par(futures_and_coroutines)`` -- Wait
  for a list of futures and/or coroutines to complete and return a
  list of their results.  If one of the futures or coroutines raises
  an exception, that exception is propagated, after attempting to
  cancel all other futures and coroutines in the list.

- ``return result`` -- produce a result to the coroutine that is
  waiting for this one using ``yield from``.

- ``raise exception`` -- raise an exception in the coroutine that is
  waiting for this one using ``yield from``.

Calling a coroutine does not start its code running -- it is just a
generator, and the coroutine object returned by the call is really a
generator object, which doesn't do anything until you iterate over it.
In the case of a coroutine object, there are two basic ways to start
it running: call ``yield from coroutine`` from another coroutine
(assuming the other coroutine is already running!), or convert it to a

Coroutines can only run when the event loop is running.


A Task is an object that manages an independently running coroutine.
The Task interface is the same as the Future interface.  The task
becomes done when its coroutine returns or raises an exception; if it
returns a result, that becomes the task's result, if it raises an
exception, that becomes the task's exception.

Cancelling a task that's not done yet prevents its coroutine from
completing; in this case an exception is thrown into the coroutine
that it may catch to further handle cancellation, but it doesn't have
to (this is done using the standard ``close()`` method on generators,
described in PEP 342).

The ``par()`` function described above runs coroutines in parallel by
converting them to Tasks.  (Arguments that are already Tasks or
Futures are not converted.)

Tasks are also useful for interoperating between coroutines and
callback-based frameworks like Twisted.  After converting a coroutine
into a Task, callbacks can be added to the Task.

You may ask, why not convert all coroutines to Tasks?  The
``@tulip.coroutine`` decorator could do this.  This would slow things
down considerably in the case where one coroutine calls another (and
so on), as switching to a "bare" coroutine has much less overhead than
switching to a Task.

The Scheduler

The scheduler has no public interface.  You interact with it by using
``yield from future`` and ``yield from task``.  In fact, there is no
single object representing the scheduler -- its behavior is
implemented by the ``Task`` and ``Future`` classes using only the
public interface of the event loop, so it will work with third-party
event loop implementations, too.


TBD: ``yield sleep(seconds)``.  Can use ``sleep(0)`` to suspend to
poll for I/O.

Wait for First

TBD: Need an interface to wait for the first of a collection of Futures.

Coroutines and Protocols

The best way to use coroutines to implement protocols is probably to
use a streaming buffer that gets filled by ``data_received()`` and can
be read asynchronously using methods like ``read(n)`` and
``readline()`` that return a Future.  When the connection is closed,
``read()`` should return a Future whose result is ``b''``, or raise an
exception if ``connection_closed()`` is called with an exception.

To write, the ``write()`` method (and friends) on the transport can be
used -- these do not return Futures.  A standard protocol
implementation should be provided that sets this up and kicks off the
coroutine when ``connection_made()`` is called.

TBD: Be more specific.


TBD.  When a Task is cancelled its coroutine may see an exception at
any point where it is yielding to the scheduler (i.e., potentially at
any ``yield from`` operation).  We need to spell out which exception
is raised.

Also TBD: timeouts.

Open Issues

- A debugging API?  E.g. something that logs a lot of stuff, or logs
  unusual conditions (like queues filling up faster than they drain)
  or even callbacks taking too much time...

- Do we need introspection APIs?  E.g. asking for the read callback
  given a file descriptor.  Or when the next scheduled call is.  Or
  the list of file descriptors registered with callbacks.

- Should we have ``future.add_callback(callback, *args)``, using the
  convention from the section "Callback Style" above, or should we
  stick with the PEP 3148 specification of
  ``future.add_done_callback(callback)`` which calls
  ``callback(future)``?  (Glyph suggested using a different method
  name since add_done_callback() does not guarantee that the callback
  will be called in the right context.)

- Returning a Future is relatively expensive, and it is quite possible
  that some types of calls *usually* complete immediately
  (e.g. writing small amounts of data to a socket).  A trick used by
  Richard Oudkerk in the tulip project's proactor branch makes calls
  like recv() either return a regular result or *raise* a Future.  The
  caller (likely a transport) must then write code like this::

        res = ev.sock_recv(sock, 8192)
    except Future as f:
        yield from sch.block_future(f)
        res = f.result()

- Do we need a larger vocabulary of operations for combining
  coroutines and/or futures?  E.g. in addition to par() we could have
  a way to run several coroutines sequentially (returning all results
  or passing the result of one to the next and returning the final
  result?).  We might also introduce explicit locks (though these will
  be a bit of a pain to use, as we can't use the ``with lock: block``
  syntax).  Anyway, I think all of these are easy enough to write
  using ``Task``.

  Proposal: ``f = yield from wait_one(fs)`` takes a set of Futures and
  sets f to the first of those that is done.  (Yes, this requires an
  intermediate Future to wait for.)  You can then write::

    while fs:
        f = tulip.wait_one(fs)
        <inspect f>        

- Support for datagram protocols, "connected" or otherwise?  Probably
  need more socket I/O methods, e.g. ``sock_sendto()`` and
  ``sock_recvfrom()``.  Or users can write their own (it's not rocket
  science).  Is it reasonable to map ``write()``, ``writelines()``,
  ``data_received()`` to single datagrams?

- Task or callback priorities?  (I hope not.)

- An EventEmitter in the style of NodeJS?  Or make this a separate
  PEP?  It's easy enough to do in user space, though it may benefit
  from standardization.  (See and for examples.)


Apart from PEP 3153, influences include PEP 380 and Greg Ewing's
tutorial for ``yield from``, Twisted, Tornado, ZeroMQ, pyftpdlib, tulip
(the author's attempts at synthesis of all these), wattle (Steve
Dower's counter-proposal), numerous discussions on python-ideas from
September through December 2012, a Skype session with Steve Dower and
Dino Viehland, email exchanges with Ben Darnell, an audience with
Niels Provos (original author of libevent), and two in-person meetings
with several Twisted developers, including Glyph, Brian Warner, David
Reid, and Duncan McGreggor.  Also, the author's previous work on async
support in the NDB library for Google App Engine was an important


This document has been placed in the public domain.

   Local Variables:
   mode: indented-text
   indent-tabs-mode: nil
   sentence-end-double-space: t
   fill-column: 70
   coding: utf-8

(Log in to post comments)

A proposal for "rebooted" Python asynchronous I/O support

Posted Dec 29, 2012 11:38 UTC (Sat) by cglass (subscriber, #52152) [Link]

I'll go ahead and ask the obvious question here, apologies if this seems obvious to people more familiar with the topic at hand:
Why not twisted? Why reimplement the same functionality with a slightly different interface?
It certainly isn't perfect, but I would be interested in understanding the reasons.

A proposal for "rebooted" Python asynchronous I/O support

Posted Dec 29, 2012 13:30 UTC (Sat) by marduk (subscriber, #3831) [Link]

I didn't read the entire PEP but it seems that Twisted, ZeroMQ and Tornado were thought of and can easily implement the interface for the PEP. It seems that one of the goals is to have a a common interface (think wsgi or the db-api) while a second goal is to have a reference implementation included in the standard library.

A proposal for "rebooted" Python asynchronous I/O support

Posted Dec 29, 2012 14:32 UTC (Sat) by njs (guest, #40338) [Link]

Twisted is seriously awesome. But I would be very surprised to see it end up in the standard library. Just to start with, twisted as a whole is a huge sprawling mess of variable quality code, so you'd have to extract some small pieces (just the reactor interface?), and even those interfaces probably have stuff in them that you'd want to clean up given how many years its been since they were first designed. And twisted relies heavily on zope-style interfaces, which have been rejected from the stdlib before. And twisted is still a long way from running on Python 3.

A proposal for "rebooted" Python asynchronous I/O support

Posted Dec 29, 2012 16:26 UTC (Sat) by tack (subscriber, #12542) [Link]

With a model like Twisted's, you have a twisty little maze of callbacks.

So one of the primary drivers of this PEP is coroutines, which let us solve the callback hell problem seen in languages like Javascript (especially with node.js). To standardize on coroutines used for async IO, you also need to standardize a bunch of other things like a Future-like object and a minimal event loop.

Once these things are defined and codified, libraries like Twisted and Kaa can conform to them, and this just might, if we play our cards right, pave the way for async library interoperability in Python land. Today each library takes a different and incompatible approach.

Also Twisted's API is not PEP8 compliant. :)

A proposal for "rebooted" Python asynchronous I/O support

Posted Dec 30, 2012 3:35 UTC (Sun) by wahern (subscriber, #37304) [Link]

From the PEP Guido says they're basically just generators underneath, which means these aren't like more modern coroutine implementations which can yield from sub-subroutine invocations. Being restricted to the first-level invocation will hardly remove the necessity for callbacks in high-level application code, unfortunately.

This is why I much prefer Lua. It might not have the huge library, but at least the VM is done right.

A proposal for "rebooted" Python asynchronous I/O support

Posted Dec 30, 2012 4:36 UTC (Sun) by tack (subscriber, #12542) [Link]

You've misunderstood the PEP. Coroutines are implemented using generators, yes, but they can be arbitrarily nested (within runtime stack depth limits of course). In other words, coroutine A can "yield from" coroutine B which can "yield from" coroutine C and so on. If C yields a Future object, the stack unravels -- through B and A -- and the coroutine scheduler will resume C's execution when the Future it yielded is done.

A proposal for "rebooted" Python asynchronous I/O support

Posted Dec 31, 2012 23:47 UTC (Mon) by wahern (subscriber, #37304) [Link]

I think you misunderstood what I was saying. To understand the fundamental problem in the CPython (and Jython) VM, just think about the difference between CPython and Stackless. You can't `yield from' inside a new function invocation because of the way call frames are implemented in CPython. Only the callee of a `yield from' statement can yield. But you can explicitly stack yield froms in a call chain. Kinda messy, but it's a fundamental limitation of CPython.

Coroutines in languages like Lua are implemented more like threads, so coroutine semantics aren't limited to a single callee invocation. In Lua, invocation C in A->B->C can yield, even though coroutine resumption began in A. In CPython, when resuming A only A can yield. A can, however, _explicitly_ invoke B as a coroutine. But B (perhaps a library call) can't hide the details. A needs to adorn it's call to B in order to get coroutine semantics. Again, not so in Lua or some other languages.

A proposal for "rebooted" Python asynchronous I/O support

Posted Jan 1, 2013 0:10 UTC (Tue) by wahern (subscriber, #37304) [Link]

FWIW, I believe Python coroutines will work much like Javascript generator trampolines:

Self-fulfilling prophecy

Posted Dec 29, 2012 12:19 UTC (Sat) by man_ls (guest, #15091) [Link]

Let the bikeshedding begin!
Not a very good start for a healthy discussion. As could be expected there are to date only a handful of messages in the thread.

Self-fulfilling prophecy

Posted Dec 29, 2012 13:28 UTC (Sat) by robert_s (subscriber, #42402) [Link]

"Let the bikeshedding begin!"

You could say the same thing about any PEP. But miraculously a lot of them get accepted.

Self-fulfilling prophecy

Posted Dec 29, 2012 17:48 UTC (Sat) by augustz (guest, #37348) [Link]

The comments look constructive and on-point to me.

Self-fulfilling prophecy

Posted Dec 29, 2012 18:38 UTC (Sat) by man_ls (guest, #15091) [Link]

Yes, you are right on that respect. Not too much discussion though, especially for such an interesting topic. Python lost momentum in the async server area when it could have led the pack.

Self-fulfilling prophecy

Posted Dec 29, 2012 18:52 UTC (Sat) by corbet (editor, #1) [Link]

Bear in mind that this PEP comes after a few months of quite intense conversation. Have a look at the python-ideas list starting in October. It's not like the conversation is starting now, in the middle of the holiday lull.

Self-fulfilling prophecy

Posted Dec 29, 2012 19:02 UTC (Sat) by man_ls (guest, #15091) [Link]

Probably van Rossum said the quoted sentence as a joke, and my humor detector is on holidays too. It just sounded a bit harsh coming from a "benevolent dictator" to me.

Self-fulfilling prophecy

Posted Jan 3, 2013 22:18 UTC (Thu) by JanC_ (subscriber, #34940) [Link]

<p>You forget that this particular BDFL named his programming language after <a href="">a comedy group</a></p>

Self-fulfilling prophecy

Posted Jan 3, 2013 22:23 UTC (Thu) by JanC_ (subscriber, #34940) [Link]

And of course this was meant to read like this:

You forget that this particular BDFL named his programming language after a comedy group

(Can we get a post editor, please? ;) )

Self-fulfilling prophecy

Posted Jan 4, 2013 11:53 UTC (Fri) by man_ls (guest, #15091) [Link]

I would have very much appreciated an explicit Python mention, e.g.:
Let the discussion about the color of Arthur "Two Sheds" Jackson's second shed begin!
But bikeshedding? Heresy!

A proposal for "rebooted" Python asynchronous I/O support

Posted Dec 30, 2012 0:33 UTC (Sun) by ppisa (subscriber, #67307) [Link]

I like proposal and please, invest time to to ensure good compatibility with libevent-2 main loop and consider even its buffer events. I think that that library is good candidate for generic mainloop for most of applications.

But there is need to live with GLIB based mainloop as well. Its interface is so rigid and unfortunate that only option for other GUI systems (like Qt) was to switch to GLIB mainloop when GLIB based loop libraries are considered for them. I hope that GLIB community would invest time to that problem some day but it cannot be done without API level compatibility break.

You can look at my ul_evpoll effort as a source for some ideas too. It started in libevent-1 days as an solution to write libraries requiring system level events registration and compatibility with GLIB and yet to be able to switch to libevent. Then the project little extended to provide own implementation of select, poll, epoll and winevent mechanism and ability to plug-in into GLIB, libevent-1 and libevent-2 based applications.

The ul_evpoll library is part of uLAN and OrtCAN repositories
I have prepared snapshot for simple standalone testing
Documentation - qute short, written in a hurry and with many sins against English language

A proposal for "rebooted" Python asynchronous I/O support

Posted Jan 5, 2013 19:53 UTC (Sat) by cmccabe (guest, #60281) [Link]

From the libabc readme:

> Avoid thinking about main loops/event dispatchers.
>  - Get your stuff right in the kernel: fds are awesome, expose them
>    in userspace and in the library, because people can easily integrate
>    them with their own poll() loops of whatever kind they have.
>  - Don't hide file descriptors away in your headers.
>  - Never add blocking kernel syscalls, and never add blocking library
>    calls either (with very few exceptions). Userspace code is primarily
>    asynchronous around event loops, and blocking calls are generally
>    incompatible with that.
>  - Corollary of that: always O_NONBLOCK!
I can't think of any library I have used in the past few years that has used a glib-based main loop. So basically, if your library depends on some event loop framework... you're doing it wrong.

A proposal for "rebooted" Python asynchronous I/O support

Posted Dec 31, 2012 22:53 UTC (Mon) by Arach (guest, #58847) [Link]

Rigid and obsolete in advance.

A proposal for "rebooted" Python asynchronous I/O support

Posted Jan 3, 2013 18:12 UTC (Thu) by intgr (subscriber, #39733) [Link]

In "advance" of what? This isn't a specification, this is a proposal, and it's not ready yet.

You are kindly invited to collaborate and improve it.

A proposal for "rebooted" Python asynchronous I/O support

Posted Jan 4, 2013 11:23 UTC (Fri) by smurf (subscriber, #17840) [Link]

Given that a model like gevent's is (a) a whole damn lot more natural to code to than all that "yield from" stuff, and (b) explicitly out of scope, I'd have to agree to the "rigid and obsolete" comment.

Debugging a maze of "yield from"s and/or explicitly registered callbacks is a *major* PITA. Been there, done that, given up, and re-implemented a couple of ten thousand lines with gevent instead of Twisted.

Fortunately, implementing a Twisted reactor on top of gevent has already been done, so the migration can be done in steps and is rather painless.

I am fairly sure that the same thing will be done for this PEP, once it's moderately final.

A proposal for "rebooted" Python asynchronous I/O support

Posted Jan 6, 2013 2:35 UTC (Sun) by tack (subscriber, #12542) [Link]

In which framework were you tasked with debugging a maze of yield froms?

This is very different from the callback world of Twisted. The code reads fairly linearly like it would with gevent, except that your exit/entry points are explicit rather than implicit.

Having coded with both approaches (that is, explicit yields vs. greenlet-style implicitness) I prefer to be explicit. I rather like the visual cue that signals where my function will suspend and resume and I find this improves debuggability rather than hinders it. IMO, yielding gives you a nice bit of async magic while not violating the principle of least astonishment.

Copyright © 2012, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds