python_doc_md/Beginner/whatsnew/3.5.md

# What\'s New In Python 3.5

Editors
:   Elvis Pranskevichus \<elvis@magic.io\>, Yury Selivanov \<yury@magic.io\>

> \* Anyone can add text to this document. Do not spend very much time on the wording of your changes, because your text will probably get rewritten to some degree.
>
> \* The maintainer will go through Misc/NEWS periodically and add changes; it\'s therefore more important to add your changes to Misc/NEWS than to this file.
>
> \* This is not a complete list of every single change; completeness is the purpose of Misc/NEWS. Some changes I consider too small or esoteric to include. If such a change is added to the text, I\'ll just remove it. (This is another reason you shouldn\'t spend too much time on writing your addition.)
>
> \* If you want to draw your new text to the attention of the maintainer, add \'XXX\' to the beginning of the paragraph or section.
>
> \* It\'s OK to just add a fragmentary note about a change. For example: \"XXX Describe the transmogrify() function added to the socket module.\" The maintainer will research the change and write the necessary text.
>
> \* You can comment out your additions if you like, but it\'s not necessary (especially when a final release is some months away).
>
> \* Credit the author of a patch or bugfix. Just the name is sufficient; the e-mail address isn\'t necessary.
>
> - It\'s helpful to add the bug/patch number as a comment:
>
> XXX Describe the transmogrify() function added to the socket module. (Contributed by P.Y. Developer in `12345`{.interpreted-text role="issue"}.)
>
> This saves the maintainer the effort of going through the Mercurial log when researching a change.

This article explains the new features in Python 3.5, compared to 3.4. Python 3.5 was released on September 13, 2015. See the [changelog](https://docs.python.org/3.5/whatsnew/changelog.html) for a full list of changes.

::: seealso
`478`{.interpreted-text role="pep"} - Python 3.5 Release Schedule
:::

## Summary \-- Release highlights

New syntax features:

- `PEP 492 <whatsnew-pep-492>`{.interpreted-text role="ref"}, coroutines with async and await syntax.
- `PEP 465 <whatsnew-pep-465>`{.interpreted-text role="ref"}, a new matrix multiplication operator: `a @ b`.
- `PEP 448 <whatsnew-pep-448>`{.interpreted-text role="ref"}, additional unpacking generalizations.

New library modules:

- `typing`{.interpreted-text role="mod"}: `PEP 484 -- Type Hints <whatsnew-pep-484>`{.interpreted-text role="ref"}.
- `zipapp`{.interpreted-text role="mod"}: `PEP 441 Improving Python ZIP Application Support
  <whatsnew-zipapp>`{.interpreted-text role="ref"}.

New built-in features:

- `bytes % args`, `bytearray % args`: `PEP 461 <whatsnew-pep-461>`{.interpreted-text role="ref"} \-- Adding `%` formatting to bytes and bytearray.
- New `bytes.hex`{.interpreted-text role="meth"}, `bytearray.hex`{.interpreted-text role="meth"} and `memoryview.hex`{.interpreted-text role="meth"} methods. (Contributed by Arnon Yaari in `9951`{.interpreted-text role="issue"}.)
- `memoryview`{.interpreted-text role="class"} now supports tuple indexing (including multi-dimensional). (Contributed by Antoine Pitrou in `23632`{.interpreted-text role="issue"}.)
- Generators have a new `gi_yieldfrom` attribute, which returns the object being iterated by `yield from` expressions. (Contributed by Benno Leslie and Yury Selivanov in `24450`{.interpreted-text role="issue"}.)
- A new `RecursionError`{.interpreted-text role="exc"} exception is now raised when maximum recursion depth is reached. (Contributed by Georg Brandl in `19235`{.interpreted-text role="issue"}.)

CPython implementation improvements:

- When the `LC_TYPE` locale is the POSIX locale (`C` locale), `sys.stdin`{.interpreted-text role="py:data"} and `sys.stdout`{.interpreted-text role="py:data"} now use the `surrogateescape` error handler, instead of the `strict` error handler. (Contributed by Victor Stinner in `19977`{.interpreted-text role="issue"}.)
- `.pyo` files are no longer used and have been replaced by a more flexible scheme that includes the optimization level explicitly in `.pyc` name. (See `PEP 488 overview <whatsnew-pep-488>`{.interpreted-text role="ref"}.)
- Builtin and extension modules are now initialized in a multi-phase process, which is similar to how Python modules are loaded. (See `PEP 489 overview <whatsnew-pep-489>`{.interpreted-text role="ref"}.)

Significant improvements in the standard library:

- `collections.OrderedDict`{.interpreted-text role="class"} is now `implemented in C <whatsnew-ordereddict>`{.interpreted-text role="ref"}, which makes it 4 to 100 times faster.
- The `ssl`{.interpreted-text role="mod"} module gained `support for Memory BIO <whatsnew-sslmemorybio>`{.interpreted-text role="ref"}, which decouples SSL protocol handling from network IO.
- The new `os.scandir`{.interpreted-text role="func"} function provides a `better and significantly faster way <whatsnew-pep-471>`{.interpreted-text role="ref"} of directory traversal.
- `functools.lru_cache`{.interpreted-text role="func"} has been mostly `reimplemented in C <whatsnew-lrucache>`{.interpreted-text role="ref"}, yielding much better performance.
- The new `subprocess.run`{.interpreted-text role="func"} function provides a `streamlined way to run subprocesses <whatsnew-subprocess>`{.interpreted-text role="ref"}.
- The `traceback`{.interpreted-text role="mod"} module has been significantly `enhanced <whatsnew-traceback>`{.interpreted-text role="ref"} for improved performance and developer convenience.

Security improvements:

- SSLv3 is now disabled throughout the standard library. It can still be enabled by instantiating a `ssl.SSLContext`{.interpreted-text role="class"} manually. (See `22638`{.interpreted-text role="issue"} for more details; this change was backported to CPython 3.4 and 2.7.)
- HTTP cookie parsing is now stricter, in order to protect against potential injection attacks. (Contributed by Antoine Pitrou in `22796`{.interpreted-text role="issue"}.)

Windows improvements:

- A new installer for Windows has replaced the old MSI. See `using-on-windows`{.interpreted-text role="ref"} for more information.
- Windows builds now use Microsoft Visual C++ 14.0, and extension modules should use the same.

Please read on for a comprehensive list of user-facing changes, including many other smaller improvements, CPython optimizations, deprecations, and potential porting issues.

## New Features

### PEP 492 - Coroutines with async and await syntax {#whatsnew-pep-492}

`492`{.interpreted-text role="pep"} greatly improves support for asynchronous programming in Python by adding `awaitable objects <awaitable>`{.interpreted-text role="term"}, `coroutine functions <coroutine function>`{.interpreted-text role="term"}, `asynchronous iteration <asynchronous iterable>`{.interpreted-text role="term"}, and `asynchronous context managers <asynchronous context manager>`{.interpreted-text role="term"}.

Coroutine functions are declared using the new `async def`{.interpreted-text role="keyword"} syntax:

    >>> async def coro():
    ...     return 'spam'

Inside a coroutine function, the new `await`{.interpreted-text role="keyword"} expression can be used to suspend coroutine execution until the result is available. Any object can be *awaited*, as long as it implements the `awaitable`{.interpreted-text role="term"} protocol by defining the `~object.__await__`{.interpreted-text role="meth"} method.

PEP 492 also adds `async for`{.interpreted-text role="keyword"} statement for convenient iteration over asynchronous iterables.

An example of a rudimentary HTTP client written using the new syntax:

    import asyncio

    async def http_get(domain):
        reader, writer = await asyncio.open_connection(domain, 80)

        writer.write(b'\r\n'.join([
            b'GET / HTTP/1.1',
            b'Host: %b' % domain.encode('latin-1'),
            b'Connection: close',
            b'', b''
        ]))

        async for line in reader:
            print('>>>', line)

        writer.close()

    loop = asyncio.get_event_loop()
    try:
        loop.run_until_complete(http_get('example.com'))
    finally:
        loop.close()

Similarly to asynchronous iteration, there is a new syntax for asynchronous context managers. The following script:

    import asyncio

    async def coro(name, lock):
        print('coro {}: waiting for lock'.format(name))
        async with lock:
            print('coro {}: holding the lock'.format(name))
            await asyncio.sleep(1)
            print('coro {}: releasing the lock'.format(name))

    loop = asyncio.get_event_loop()
    lock = asyncio.Lock()
    coros = asyncio.gather(coro(1, lock), coro(2, lock))
    try:
        loop.run_until_complete(coros)
    finally:
        loop.close()

will output:

    coro 2: waiting for lock
    coro 2: holding the lock
    coro 1: waiting for lock
    coro 2: releasing the lock
    coro 1: holding the lock
    coro 1: releasing the lock

Note that both `async for`{.interpreted-text role="keyword"} and `async with`{.interpreted-text role="keyword"} can only be used inside a coroutine function declared with `async def`{.interpreted-text role="keyword"}.

Coroutine functions are intended to be run inside a compatible event loop, such as the `asyncio loop <asyncio-event-loop>`{.interpreted-text role="ref"}.

::::: note
::: title
Note
:::

::: versionchanged
3.5.2 Starting with CPython 3.5.2, `__aiter__` can directly return `asynchronous iterators <asynchronous iterator>`{.interpreted-text role="term"}. Returning an `awaitable`{.interpreted-text role="term"} object will result in a `PendingDeprecationWarning`{.interpreted-text role="exc"}.

See more details in the `async-iterators`{.interpreted-text role="ref"} documentation section.
:::
:::::

::: seealso

`492`{.interpreted-text role="pep"} \-- Coroutines with async and await syntax

:   PEP written and implemented by Yury Selivanov.
:::

### PEP 465 - A dedicated infix operator for matrix multiplication {#whatsnew-pep-465}

`465`{.interpreted-text role="pep"} adds the `@` infix operator for matrix multiplication. Currently, no builtin Python types implement the new operator, however, it can be implemented by defining `~object.__matmul__`{.interpreted-text role="meth"}, `~object.__rmatmul__`{.interpreted-text role="meth"}, and `~object.__imatmul__`{.interpreted-text role="meth"} for regular, reflected, and in-place matrix multiplication. The semantics of these methods is similar to that of methods defining other infix arithmetic operators.

Matrix multiplication is a notably common operation in many fields of mathematics, science, engineering, and the addition of `@` allows writing cleaner code:

    S = (H @ beta - r).T @ inv(H @ V @ H.T) @ (H @ beta - r)

instead of:

    S = dot((dot(H, beta) - r).T,
            dot(inv(dot(dot(H, V), H.T)), dot(H, beta) - r))

NumPy 1.10 has support for the new operator:

    >>> import numpy

    >>> x = numpy.ones(3)
    >>> x
    array([ 1., 1., 1.])

    >>> m = numpy.eye(3)
    >>> m
    array([[ 1., 0., 0.],
           [ 0., 1., 0.],
           [ 0., 0., 1.]])

    >>> x @ m
    array([ 1., 1., 1.])

::: seealso

`465`{.interpreted-text role="pep"} \-- A dedicated infix operator for matrix multiplication

:   PEP written by Nathaniel J. Smith; implemented by Benjamin Peterson.
:::

### PEP 448 - Additional Unpacking Generalizations {#whatsnew-pep-448}

`448`{.interpreted-text role="pep"} extends the allowed uses of the `*` iterable unpacking operator and `**` dictionary unpacking operator. It is now possible to use an arbitrary number of unpackings in `function calls <calls>`{.interpreted-text role="ref"}:

    >>> print(*[1], *[2], 3, *[4, 5])
    1 2 3 4 5

    >>> def fn(a, b, c, d):
    ...     print(a, b, c, d)
    ...

    >>> fn(**{'a': 1, 'c': 3}, **{'b': 2, 'd': 4})
    1 2 3 4

Similarly, tuple, list, set, and dictionary displays allow multiple unpackings (see `exprlists`{.interpreted-text role="ref"} and `dict`{.interpreted-text role="ref"}):

    >>> *range(4), 4
    (0, 1, 2, 3, 4)

    >>> [*range(4), 4]
    [0, 1, 2, 3, 4]

    >>> {*range(4), 4, *(5, 6, 7)}
    {0, 1, 2, 3, 4, 5, 6, 7}

    >>> {'x': 1, **{'y': 2}}
    {'x': 1, 'y': 2}

::: seealso

`448`{.interpreted-text role="pep"} \-- Additional Unpacking Generalizations

:   PEP written by Joshua Landau; implemented by Neil Girdhar, Thomas Wouters, and Joshua Landau.
:::

### PEP 461 - percent formatting support for bytes and bytearray {#whatsnew-pep-461}

`461`{.interpreted-text role="pep"} adds support for the `%` `interpolation operator <bytes-formatting>`{.interpreted-text role="ref"} to `bytes`{.interpreted-text role="class"} and `bytearray`{.interpreted-text role="class"}.

While interpolation is usually thought of as a string operation, there are cases where interpolation on `bytes` or `bytearrays` makes sense, and the work needed to make up for this missing functionality detracts from the overall readability of the code. This issue is particularly important when dealing with wire format protocols, which are often a mixture of binary and ASCII compatible text.

Examples:

    >>> b'Hello %b!' % b'World'
    b'Hello World!'

    >>> b'x=%i y=%f' % (1, 2.5)
    b'x=1 y=2.500000'

Unicode is not allowed for `%b`, but it is accepted by `%a` (equivalent of `repr(obj).encode('ascii', 'backslashreplace')`):

    >>> b'Hello %b!' % 'World'
    Traceback (most recent call last):
      File "<stdin>", line 1, in <module>
    TypeError: %b requires bytes, or an object that implements __bytes__, not 'str'

    >>> b'price: %a' % '10€'
    b"price: '10\\u20ac'"

Note that `%s` and `%r` conversion types, although supported, should only be used in codebases that need compatibility with Python 2.

::: seealso

`461`{.interpreted-text role="pep"} \-- Adding % formatting to bytes and bytearray

:   PEP written by Ethan Furman; implemented by Neil Schemenauer and Ethan Furman.
:::

### PEP 484 - Type Hints {#whatsnew-pep-484}

Function annotation syntax has been a Python feature since version 3.0 (`3107`{.interpreted-text role="pep"}), however the semantics of annotations has been left undefined.

Experience has shown that the majority of function annotation uses were to provide type hints to function parameters and return values. It became evident that it would be beneficial for Python users, if the standard library included the base definitions and tools for type annotations.

`484`{.interpreted-text role="pep"} introduces a `provisional module <provisional API>`{.interpreted-text role="term"} to provide these standard definitions and tools, along with some conventions for situations where annotations are not available.

For example, here is a simple function whose argument and return type are declared in the annotations:

    def greeting(name: str) -> str:
        return 'Hello ' + name

While these annotations are available at runtime through the usual `~object.__annotations__`{.interpreted-text role="attr"} attribute, *no automatic type checking happens at runtime*. Instead, it is assumed that a separate off-line type checker (e.g. [mypy](https://mypy-lang.org)) will be used for on-demand source code analysis.

The type system supports unions, generic types, and a special type named `~typing.Any`{.interpreted-text role="class"} which is consistent with (i.e. assignable to and from) all types.

::: seealso
- `typing`{.interpreted-text role="mod"} module documentation

- 

  `484`{.interpreted-text role="pep"} \-- Type Hints

  :   PEP written by Guido van Rossum, Jukka Lehtosalo, and Łukasz Langa; implemented by Guido van Rossum.

- 

  `483`{.interpreted-text role="pep"} \-- The Theory of Type Hints

  :   PEP written by Guido van Rossum
:::

### PEP 471 - os.scandir() function \-- a better and faster directory iterator {#whatsnew-pep-471}

`471`{.interpreted-text role="pep"} adds a new directory iteration function, `os.scandir`{.interpreted-text role="func"}, to the standard library. Additionally, `os.walk`{.interpreted-text role="func"} is now implemented using `scandir`, which makes it 3 to 5 times faster on POSIX systems and 7 to 20 times faster on Windows systems. This is largely achieved by greatly reducing the number of calls to `os.stat`{.interpreted-text role="func"} required to walk a directory tree.

Additionally, `scandir` returns an iterator, as opposed to returning a list of file names, which improves memory efficiency when iterating over very large directories.

The following example shows a simple use of `os.scandir`{.interpreted-text role="func"} to display all the files (excluding directories) in the given *path* that don\'t start with `'.'`. The `entry.is_file() <os.DirEntry.is_file>`{.interpreted-text role="meth"} call will generally not make an additional system call:

    for entry in os.scandir(path):
        if not entry.name.startswith('.') and entry.is_file():
            print(entry.name)

::: seealso

`471`{.interpreted-text role="pep"} \-- os.scandir() function \-- a better and faster directory iterator

:   PEP written and implemented by Ben Hoyt with the help of Victor Stinner.
:::

### PEP 475: Retry system calls failing with EINTR {#whatsnew-pep-475}

An `errno.EINTR`{.interpreted-text role="py:const"} error code is returned whenever a system call, that is waiting for I/O, is interrupted by a signal. Previously, Python would raise `InterruptedError`{.interpreted-text role="exc"} in such cases. This meant that, when writing a Python application, the developer had two choices:

1.  Ignore the `InterruptedError`.
2.  Handle the `InterruptedError` and attempt to restart the interrupted system call at every call site.

The first option makes an application fail intermittently. The second option adds a large amount of boilerplate that makes the code nearly unreadable. Compare:

    print("Hello World")

and:

    while True:
        try:
            print("Hello World")
            break
        except InterruptedError:
            continue

`475`{.interpreted-text role="pep"} implements automatic retry of system calls on `EINTR`. This removes the burden of dealing with `EINTR` or `InterruptedError`{.interpreted-text role="exc"} in user code in most situations and makes Python programs, including the standard library, more robust. Note that the system call is only retried if the signal handler does not raise an exception.

Below is a list of functions which are now retried when interrupted by a signal:

- `open`{.interpreted-text role="func"} and `io.open`{.interpreted-text role="func"};
- functions of the `faulthandler`{.interpreted-text role="mod"} module;
- `os`{.interpreted-text role="mod"} functions: `~os.fchdir`{.interpreted-text role="func"}, `~os.fchmod`{.interpreted-text role="func"}, `~os.fchown`{.interpreted-text role="func"}, `~os.fdatasync`{.interpreted-text role="func"}, `~os.fstat`{.interpreted-text role="func"}, `~os.fstatvfs`{.interpreted-text role="func"}, `~os.fsync`{.interpreted-text role="func"}, `~os.ftruncate`{.interpreted-text role="func"}, `~os.mkfifo`{.interpreted-text role="func"}, `~os.mknod`{.interpreted-text role="func"}, `~os.open`{.interpreted-text role="func"}, `~os.posix_fadvise`{.interpreted-text role="func"}, `~os.posix_fallocate`{.interpreted-text role="func"}, `~os.pread`{.interpreted-text role="func"}, `~os.pwrite`{.interpreted-text role="func"}, `~os.read`{.interpreted-text role="func"}, `~os.readv`{.interpreted-text role="func"}, `~os.sendfile`{.interpreted-text role="func"}, `~os.wait3`{.interpreted-text role="func"}, `~os.wait4`{.interpreted-text role="func"}, `~os.wait`{.interpreted-text role="func"}, `~os.waitid`{.interpreted-text role="func"}, `~os.waitpid`{.interpreted-text role="func"}, `~os.write`{.interpreted-text role="func"}, `~os.writev`{.interpreted-text role="func"};
- special cases: `os.close`{.interpreted-text role="func"} and `os.dup2`{.interpreted-text role="func"} now ignore `~errno.EINTR`{.interpreted-text role="py:const"} errors; the syscall is not retried (see the PEP for the rationale);
- `select`{.interpreted-text role="mod"} functions: `devpoll.poll() <select.devpoll.poll>`{.interpreted-text role="func"}, `epoll.poll() <select.epoll.poll>`{.interpreted-text role="func"}, `kqueue.control() <select.kqueue.control>`{.interpreted-text role="func"}, `poll.poll() <select.poll.poll>`{.interpreted-text role="func"}, `~select.select`{.interpreted-text role="func"};
- methods of the `~socket.socket`{.interpreted-text role="class"} class: `~socket.socket.accept`{.interpreted-text role="meth"}, `~socket.socket.connect`{.interpreted-text role="meth"} (except for non-blocking sockets), `~socket.socket.recv`{.interpreted-text role="meth"}, `~socket.socket.recvfrom`{.interpreted-text role="meth"}, `~socket.socket.recvmsg`{.interpreted-text role="meth"}, `~socket.socket.send`{.interpreted-text role="meth"}, `~socket.socket.sendall`{.interpreted-text role="meth"}, `~socket.socket.sendmsg`{.interpreted-text role="meth"}, `~socket.socket.sendto`{.interpreted-text role="meth"};
- `signal.sigtimedwait`{.interpreted-text role="func"} and `signal.sigwaitinfo`{.interpreted-text role="func"};
- `time.sleep`{.interpreted-text role="func"}.

::: seealso

`475`{.interpreted-text role="pep"} \-- Retry system calls failing with EINTR

:   PEP and implementation written by Charles-François Natali and Victor Stinner, with the help of Antoine Pitrou (the French connection).
:::

### PEP 479: Change StopIteration handling inside generators {#whatsnew-pep-479}

The interaction of generators and `StopIteration`{.interpreted-text role="exc"} in Python 3.4 and earlier was sometimes surprising, and could conceal obscure bugs. Previously, `StopIteration` raised accidentally inside a generator function was interpreted as the end of the iteration by the loop construct driving the generator.

`479`{.interpreted-text role="pep"} changes the behavior of generators: when a `StopIteration` exception is raised inside a generator, it is replaced with a `RuntimeError`{.interpreted-text role="exc"} before it exits the generator frame. The main goal of this change is to ease debugging in the situation where an unguarded `next`{.interpreted-text role="func"} call raises `StopIteration` and causes the iteration controlled by the generator to terminate silently. This is particularly pernicious in combination with the `yield from` construct.

This is a backwards incompatible change, so to enable the new behavior, a `__future__`{.interpreted-text role="term"} import is necessary:

    >>> from __future__ import generator_stop

    >>> def gen():
    ...     next(iter([]))
    ...     yield
    ...
    >>> next(gen())
    Traceback (most recent call last):
      File "<stdin>", line 2, in gen
    StopIteration

    The above exception was the direct cause of the following exception:

    Traceback (most recent call last):
      File "<stdin>", line 1, in <module>
    RuntimeError: generator raised StopIteration

Without a `__future__` import, a `PendingDeprecationWarning`{.interpreted-text role="exc"} will be raised whenever a `StopIteration`{.interpreted-text role="exc"} exception is raised inside a generator.

::: seealso

`479`{.interpreted-text role="pep"} \-- Change StopIteration handling inside generators

:   PEP written by Chris Angelico and Guido van Rossum. Implemented by Chris Angelico, Yury Selivanov and Nick Coghlan.
:::

### PEP 485: A function for testing approximate equality {#whatsnew-pep-485}

`485`{.interpreted-text role="pep"} adds the `math.isclose`{.interpreted-text role="func"} and `cmath.isclose`{.interpreted-text role="func"} functions which tell whether two values are approximately equal or \"close\" to each other. Whether or not two values are considered close is determined according to given absolute and relative tolerances. Relative tolerance is the maximum allowed difference between `isclose` arguments, relative to the larger absolute value:

    >>> import math
    >>> a = 5.0
    >>> b = 4.99998
    >>> math.isclose(a, b, rel_tol=1e-5)
    True
    >>> math.isclose(a, b, rel_tol=1e-6)
    False

It is also possible to compare two values using absolute tolerance, which must be a non-negative value:

    >>> import math
    >>> a = 5.0
    >>> b = 4.99998
    >>> math.isclose(a, b, abs_tol=0.00003)
    True
    >>> math.isclose(a, b, abs_tol=0.00001)
    False

::: seealso

`485`{.interpreted-text role="pep"} \-- A function for testing approximate equality

:   PEP written by Christopher Barker; implemented by Chris Barker and Tal Einat.
:::

### PEP 486: Make the Python Launcher aware of virtual environments {#whatsnew-pep-486}

`486`{.interpreted-text role="pep"} makes the Windows launcher (see `397`{.interpreted-text role="pep"}) aware of an active virtual environment. When the default interpreter would be used and the `VIRTUAL_ENV` environment variable is set, the interpreter in the virtual environment will be used.

::: seealso

`486`{.interpreted-text role="pep"} \-- Make the Python Launcher aware of virtual environments

:   PEP written and implemented by Paul Moore.
:::

### PEP 488: Elimination of PYO files {#whatsnew-pep-488}

`488`{.interpreted-text role="pep"} does away with the concept of `.pyo` files. This means that `.pyc` files represent both unoptimized and optimized bytecode. To prevent the need to constantly regenerate bytecode files, `.pyc` files now have an optional `opt-` tag in their name when the bytecode is optimized. This has the side-effect of no more bytecode file name clashes when running under either `-O`{.interpreted-text role="option"} or `-OO`{.interpreted-text role="option"}. Consequently, bytecode files generated from `-O`{.interpreted-text role="option"}, and `-OO`{.interpreted-text role="option"} may now exist simultaneously. `importlib.util.cache_from_source`{.interpreted-text role="func"} has an updated API to help with this change.

::: seealso

`488`{.interpreted-text role="pep"} \-- Elimination of PYO files

:   PEP written and implemented by Brett Cannon.
:::

### PEP 489: Multi-phase extension module initialization {#whatsnew-pep-489}

`489`{.interpreted-text role="pep"} updates extension module initialization to take advantage of the two step module loading mechanism introduced by `451`{.interpreted-text role="pep"} in Python 3.4.

This change brings the import semantics of extension modules that opt-in to using the new mechanism much closer to those of Python source and bytecode modules, including the ability to use any valid identifier as a module name, rather than being restricted to ASCII.

::: seealso

`489`{.interpreted-text role="pep"} \-- Multi-phase extension module initialization

:   PEP written by Petr Viktorin, Stefan Behnel, and Nick Coghlan; implemented by Petr Viktorin.
:::

## Other Language Changes

Some smaller changes made to the core Python language are:

- Added the `"namereplace"` error handlers. The `"backslashreplace"` error handlers now work with decoding and translating. (Contributed by Serhiy Storchaka in `19676`{.interpreted-text role="issue"} and `22286`{.interpreted-text role="issue"}.)
- The `-b`{.interpreted-text role="option"} option now affects comparisons of `bytes`{.interpreted-text role="class"} with `int`{.interpreted-text role="class"}. (Contributed by Serhiy Storchaka in `23681`{.interpreted-text role="issue"}.)
- New Kazakh `kz1048` and Tajik `koi8_t` `codecs <standard-encodings>`{.interpreted-text role="ref"}. (Contributed by Serhiy Storchaka in `22682`{.interpreted-text role="issue"} and `22681`{.interpreted-text role="issue"}.)
- Property docstrings are now writable. This is especially useful for `collections.namedtuple`{.interpreted-text role="func"} docstrings. (Contributed by Berker Peksag in `24064`{.interpreted-text role="issue"}.)
- Circular imports involving relative imports are now supported. (Contributed by Brett Cannon and Antoine Pitrou in `17636`{.interpreted-text role="issue"}.)

## New Modules

### typing

The new `typing`{.interpreted-text role="mod"} `provisional <provisional API>`{.interpreted-text role="term"} module provides standard definitions and tools for function type annotations. See `Type Hints <whatsnew-pep-484>`{.interpreted-text role="ref"} for more information.

### zipapp {#whatsnew-zipapp}

The new `zipapp`{.interpreted-text role="mod"} module (specified in `441`{.interpreted-text role="pep"}) provides an API and command line tool for creating executable Python Zip Applications, which were introduced in Python 2.6 in `1739468`{.interpreted-text role="issue"}, but which were not well publicized, either at the time or since.

With the new module, bundling your application is as simple as putting all the files, including a `__main__.py` file, into a directory `myapp` and running:

``` shell-session
$ python -m zipapp myapp
$ python myapp.pyz
```

The module implementation has been contributed by Paul Moore in `23491`{.interpreted-text role="issue"}.

::: seealso
`441`{.interpreted-text role="pep"} \-- Improving Python ZIP Application Support
:::

## Improved Modules

### argparse

The `~argparse.ArgumentParser`{.interpreted-text role="class"} class now allows disabling `abbreviated usage <prefix-matching>`{.interpreted-text role="ref"} of long options by setting `allow_abbrev`{.interpreted-text role="ref"} to `False`. (Contributed by Jonathan Paugh, Steven Bethard, paul j3 and Daniel Eriksson in `14910`{.interpreted-text role="issue"}.)

### asyncio

Since the `asyncio`{.interpreted-text role="mod"} module is `provisional <provisional API>`{.interpreted-text role="term"}, all changes introduced in Python 3.5 have also been backported to Python 3.4.x.

Notable changes in the `asyncio`{.interpreted-text role="mod"} module since Python 3.4.0:

- New debugging APIs: `loop.set_debug() <asyncio.loop.set_debug>`{.interpreted-text role="meth"} and `loop.get_debug() <asyncio.loop.get_debug>`{.interpreted-text role="meth"} methods. (Contributed by Victor Stinner.)
- The proactor event loop now supports SSL. (Contributed by Antoine Pitrou and Victor Stinner in `22560`{.interpreted-text role="issue"}.)
- A new `loop.is_closed() <asyncio.loop.is_closed>`{.interpreted-text role="meth"} method to check if the event loop is closed. (Contributed by Victor Stinner in `21326`{.interpreted-text role="issue"}.)
- A new `loop.create_task() <asyncio.loop.create_task>`{.interpreted-text role="meth"} to conveniently create and schedule a new `~asyncio.Task`{.interpreted-text role="class"} for a coroutine. The `create_task` method is also used by all asyncio functions that wrap coroutines into tasks, such as `asyncio.wait`{.interpreted-text role="func"}, `asyncio.gather`{.interpreted-text role="func"}, etc. (Contributed by Victor Stinner.)
- A new `transport.get_write_buffer_limits() <asyncio.WriteTransport.get_write_buffer_limits>`{.interpreted-text role="meth"} method to inquire for *high-* and *low-* water limits of the flow control. (Contributed by Victor Stinner.)
- The `!async`{.interpreted-text role="func"} function is deprecated in favor of `~asyncio.ensure_future`{.interpreted-text role="func"}. (Contributed by Yury Selivanov.)
- New `loop.set_task_factory()
  <asyncio.loop.set_task_factory>`{.interpreted-text role="meth"} and `loop.get_task_factory() <asyncio.loop.get_task_factory>`{.interpreted-text role="meth"} methods to customize the task factory that `loop.create_task()
  <asyncio.loop.create_task>`{.interpreted-text role="meth"} method uses. (Contributed by Yury Selivanov.)
- New `Queue.join() <asyncio.Queue.join>`{.interpreted-text role="meth"} and `Queue.task_done() <asyncio.Queue.task_done>`{.interpreted-text role="meth"} queue methods. (Contributed by Victor Stinner.)
- The `JoinableQueue` class was removed, in favor of the `asyncio.Queue`{.interpreted-text role="class"} class. (Contributed by Victor Stinner.)

Updates in 3.5.1:

- The `~asyncio.ensure_future`{.interpreted-text role="func"} function and all functions that use it, such as `loop.run_until_complete() <asyncio.loop.run_until_complete>`{.interpreted-text role="meth"}, now accept all kinds of `awaitable objects <awaitable>`{.interpreted-text role="term"}. (Contributed by Yury Selivanov.)
- New `~asyncio.run_coroutine_threadsafe`{.interpreted-text role="func"} function to submit coroutines to event loops from other threads. (Contributed by Vincent Michel.)
- New `Transport.is_closing() <asyncio.BaseTransport.is_closing>`{.interpreted-text role="meth"} method to check if the transport is closing or closed. (Contributed by Yury Selivanov.)
- The `loop.create_server() <asyncio.loop.create_server>`{.interpreted-text role="meth"} method can now accept a list of hosts. (Contributed by Yann Sionneau.)

Updates in 3.5.2:

- New `loop.create_future() <asyncio.loop.create_future>`{.interpreted-text role="meth"} method to create Future objects. This allows alternative event loop implementations, such as [uvloop](https://github.com/MagicStack/uvloop), to provide a faster `asyncio.Future`{.interpreted-text role="class"} implementation. (Contributed by Yury Selivanov.)
- New `loop.get_exception_handler() <asyncio.loop.get_exception_handler>`{.interpreted-text role="meth"} method to get the current exception handler. (Contributed by Yury Selivanov.)
- New `StreamReader.readuntil() <asyncio.StreamReader.readuntil>`{.interpreted-text role="meth"} method to read data from the stream until a separator bytes sequence appears. (Contributed by Mark Korenberg.)
- The `loop.create_connection() <asyncio.loop.create_connection>`{.interpreted-text role="meth"} and `loop.create_server() <asyncio.loop.create_server>`{.interpreted-text role="meth"} methods are optimized to avoid calling the system `getaddrinfo` function if the address is already resolved. (Contributed by A. Jesse Jiryu Davis.)
- The `loop.sock_connect(sock, address) <asyncio.loop.sock_connect>`{.interpreted-text role="meth"} no longer requires the *address* to be resolved prior to the call. (Contributed by A. Jesse Jiryu Davis.)

### bz2

The `BZ2Decompressor.decompress <bz2.BZ2Decompressor.decompress>`{.interpreted-text role="meth"} method now accepts an optional *max_length* argument to limit the maximum size of decompressed data. (Contributed by Nikolaus Rath in `15955`{.interpreted-text role="issue"}.)

### cgi

The `!FieldStorage`{.interpreted-text role="class"} class now supports the `context manager`{.interpreted-text role="term"} protocol. (Contributed by Berker Peksag in `20289`{.interpreted-text role="issue"}.)

### cmath

A new function `~cmath.isclose`{.interpreted-text role="func"} provides a way to test for approximate equality. (Contributed by Chris Barker and Tal Einat in `24270`{.interpreted-text role="issue"}.)

### code

The `InteractiveInterpreter.showtraceback() <code.InteractiveInterpreter.showtraceback>`{.interpreted-text role="func"} method now prints the full chained traceback, just like the interactive interpreter. (Contributed by Claudiu Popa in `17442`{.interpreted-text role="issue"}.)

### collections

::: {#whatsnew-ordereddict}
The `~collections.OrderedDict`{.interpreted-text role="class"} class is now implemented in C, which makes it 4 to 100 times faster. (Contributed by Eric Snow in `16991`{.interpreted-text role="issue"}.)
:::

`!OrderedDict.items`{.interpreted-text role="meth"}, `!OrderedDict.keys`{.interpreted-text role="meth"}, and `!OrderedDict.values`{.interpreted-text role="meth"} views now support `reversed`{.interpreted-text role="func"} iteration. (Contributed by Serhiy Storchaka in `19505`{.interpreted-text role="issue"}.)

The `~collections.deque`{.interpreted-text role="class"} class now defines `~collections.deque.index`{.interpreted-text role="meth"}, `~collections.deque.insert`{.interpreted-text role="meth"}, and `~collections.deque.copy`{.interpreted-text role="meth"}, and supports the `+` and `*` operators. This allows deques to be recognized as a `~collections.abc.MutableSequence`{.interpreted-text role="class"} and improves their substitutability for lists. (Contributed by Raymond Hettinger in `23704`{.interpreted-text role="issue"}.)

Docstrings produced by `~collections.namedtuple`{.interpreted-text role="func"} can now be updated:

    Point = namedtuple('Point', ['x', 'y'])
    Point.__doc__ += ': Cartesian coordinate'
    Point.x.__doc__ = 'abscissa'
    Point.y.__doc__ = 'ordinate'

(Contributed by Berker Peksag in `24064`{.interpreted-text role="issue"}.)

The `~collections.UserString`{.interpreted-text role="class"} class now implements the `~object.__getnewargs__`{.interpreted-text role="meth"}, `~object.__rmod__`{.interpreted-text role="meth"}, `~str.casefold`{.interpreted-text role="meth"}, `~str.format_map`{.interpreted-text role="meth"}, `~str.isprintable`{.interpreted-text role="meth"}, and `~str.maketrans`{.interpreted-text role="meth"} methods to match the corresponding methods of `str`{.interpreted-text role="class"}. (Contributed by Joe Jevnik in `22189`{.interpreted-text role="issue"}.)

### collections.abc

The `!Sequence.index`{.interpreted-text role="meth"} method now accepts *start* and *stop* arguments to match the corresponding methods of `tuple`{.interpreted-text role="class"}, `list`{.interpreted-text role="class"}, etc. (Contributed by Devin Jeanpierre in `23086`{.interpreted-text role="issue"}.)

A new `~collections.abc.Generator`{.interpreted-text role="class"} abstract base class. (Contributed by Stefan Behnel in `24018`{.interpreted-text role="issue"}.)

New `~collections.abc.Awaitable`{.interpreted-text role="class"}, `~collections.abc.Coroutine`{.interpreted-text role="class"}, `~collections.abc.AsyncIterator`{.interpreted-text role="class"}, and `~collections.abc.AsyncIterable`{.interpreted-text role="class"} abstract base classes. (Contributed by Yury Selivanov in `24184`{.interpreted-text role="issue"}.)

For earlier Python versions, a backport of the new ABCs is available in an external `PyPI package <backports_abc>`{.interpreted-text role="pypi"}.

### compileall

A new `compileall`{.interpreted-text role="mod"} option, `-j {N}`{.interpreted-text role="samp"}, allows running *N* workers simultaneously to perform parallel bytecode compilation. The `~compileall.compile_dir`{.interpreted-text role="func"} function has a corresponding `workers` parameter. (Contributed by Claudiu Popa in `16104`{.interpreted-text role="issue"}.)

Another new option, `-r`, allows controlling the maximum recursion level for subdirectories. (Contributed by Claudiu Popa in `19628`{.interpreted-text role="issue"}.)

The `-q` command line option can now be specified more than once, in which case all output, including errors, will be suppressed. The corresponding `quiet` parameter in `~compileall.compile_dir`{.interpreted-text role="func"}, `~compileall.compile_file`{.interpreted-text role="func"}, and `~compileall.compile_path`{.interpreted-text role="func"} can now accept an integer value indicating the level of output suppression. (Contributed by Thomas Kluyver in `21338`{.interpreted-text role="issue"}.)

### concurrent.futures

The `Executor.map() <concurrent.futures.Executor.map>`{.interpreted-text role="meth"} method now accepts a *chunksize* argument to allow batching of tasks to improve performance when `~concurrent.futures.ProcessPoolExecutor`{.interpreted-text role="meth"} is used. (Contributed by Dan O\'Reilly in `11271`{.interpreted-text role="issue"}.)

The number of workers in the `~concurrent.futures.ThreadPoolExecutor`{.interpreted-text role="class"} constructor is optional now. The default value is 5 times the number of CPUs. (Contributed by Claudiu Popa in `21527`{.interpreted-text role="issue"}.)

### configparser

`configparser`{.interpreted-text role="mod"} now provides a way to customize the conversion of values by specifying a dictionary of converters in the `~configparser.ConfigParser`{.interpreted-text role="class"} constructor, or by defining them as methods in `ConfigParser` subclasses. Converters defined in a parser instance are inherited by its section proxies.

Example:

    >>> import configparser
    >>> conv = {}
    >>> conv['list'] = lambda v: [e.strip() for e in v.split() if e.strip()]
    >>> cfg = configparser.ConfigParser(converters=conv)
    >>> cfg.read_string("""
    ... [s]
    ... list = a b c d e f g
    ... """)
    >>> cfg.get('s', 'list')
    'a b c d e f g'
    >>> cfg.getlist('s', 'list')
    ['a', 'b', 'c', 'd', 'e', 'f', 'g']
    >>> section = cfg['s']
    >>> section.getlist('list')
    ['a', 'b', 'c', 'd', 'e', 'f', 'g']

(Contributed by Łukasz Langa in `18159`{.interpreted-text role="issue"}.)

### contextlib

The new `~contextlib.redirect_stderr`{.interpreted-text role="func"} `context manager`{.interpreted-text role="term"} (similar to `~contextlib.redirect_stdout`{.interpreted-text role="func"}) makes it easier for utility scripts to handle inflexible APIs that write their output to `sys.stderr`{.interpreted-text role="data"} and don\'t provide any options to redirect it:

    >>> import contextlib, io, logging
    >>> f = io.StringIO()
    >>> with contextlib.redirect_stderr(f):
    ...     logging.warning('warning')
    ...
    >>> f.getvalue()
    'WARNING:root:warning\n'

(Contributed by Berker Peksag in `22389`{.interpreted-text role="issue"}.)

### csv

The `~csv.csvwriter.writerow`{.interpreted-text role="meth"} method now supports arbitrary iterables, not just sequences. (Contributed by Serhiy Storchaka in `23171`{.interpreted-text role="issue"}.)

### curses

The new `~curses.update_lines_cols`{.interpreted-text role="func"} function updates the `~curses.LINES`{.interpreted-text role="data"} and `~curses.COLS`{.interpreted-text role="data"} module variables. This is useful for detecting manual screen resizing. (Contributed by Arnon Yaari in `4254`{.interpreted-text role="issue"}.)

### dbm

`dumb.open <dbm.dumb.open>`{.interpreted-text role="func"} always creates a new database when the flag has the value `"n"`. (Contributed by Claudiu Popa in `18039`{.interpreted-text role="issue"}.)

### difflib

The charset of HTML documents generated by `HtmlDiff.make_file() <difflib.HtmlDiff.make_file>`{.interpreted-text role="meth"} can now be customized by using a new *charset* keyword-only argument. The default charset of HTML document changed from `"ISO-8859-1"` to `"utf-8"`. (Contributed by Berker Peksag in `2052`{.interpreted-text role="issue"}.)

The `~difflib.diff_bytes`{.interpreted-text role="func"} function can now compare lists of byte strings. This fixes a regression from Python 2. (Contributed by Terry J. Reedy and Greg Ward in `17445`{.interpreted-text role="issue"}.)

### distutils

Both the `build` and `build_ext` commands now accept a `-j` option to enable parallel building of extension modules. (Contributed by Antoine Pitrou in `5309`{.interpreted-text role="issue"}.)

The `distutils` module now supports `xz` compression, and can be enabled by passing `xztar` as an argument to `bdist --format`. (Contributed by Serhiy Storchaka in `16314`{.interpreted-text role="issue"}.)

### doctest

The `~doctest.DocTestSuite`{.interpreted-text role="func"} function returns an empty `unittest.TestSuite`{.interpreted-text role="class"} if *module* contains no docstrings, instead of raising `ValueError`{.interpreted-text role="exc"}. (Contributed by Glenn Jones in `15916`{.interpreted-text role="issue"}.)

### email

A new policy option `Policy.mangle_from_ <email.policy.Policy.mangle_from_>`{.interpreted-text role="attr"} controls whether or not lines that start with `"From "` in email bodies are prefixed with a `">"` character by generators. The default is `True` for `~email.policy.compat32`{.interpreted-text role="attr"} and `False` for all other policies. (Contributed by Milan Oberkirch in `20098`{.interpreted-text role="issue"}.)

A new `Message.get_content_disposition() <email.message.Message.get_content_disposition>`{.interpreted-text role="meth"} method provides easy access to a canonical value for the `Content-Disposition`{.interpreted-text role="mailheader"} header. (Contributed by Abhilash Raj in `21083`{.interpreted-text role="issue"}.)

A new policy option `EmailPolicy.utf8 <email.policy.EmailPolicy.utf8>`{.interpreted-text role="attr"} can be set to `True` to encode email headers using the UTF-8 charset instead of using encoded words. This allows `Messages` to be formatted according to `6532`{.interpreted-text role="rfc"} and used with an SMTP server that supports the `6531`{.interpreted-text role="rfc"} `SMTPUTF8` extension. (Contributed by R. David Murray in `24211`{.interpreted-text role="issue"}.)

The `mime.text.MIMEText <email.mime.text.MIMEText>`{.interpreted-text role="class"} constructor now accepts a `charset.Charset <email.charset.Charset>`{.interpreted-text role="class"} instance. (Contributed by Claude Paroz and Berker Peksag in `16324`{.interpreted-text role="issue"}.)

### enum

The `~enum.Enum`{.interpreted-text role="class"} callable has a new parameter *start* to specify the initial number of enum values if only *names* are provided:

    >>> Animal = enum.Enum('Animal', 'cat dog', start=10)
    >>> Animal.cat
    <Animal.cat: 10>
    >>> Animal.dog
    <Animal.dog: 11>

(Contributed by Ethan Furman in `21706`{.interpreted-text role="issue"}.)

### faulthandler

The `~faulthandler.enable`{.interpreted-text role="func"}, `~faulthandler.register`{.interpreted-text role="func"}, `~faulthandler.dump_traceback`{.interpreted-text role="func"} and `~faulthandler.dump_traceback_later`{.interpreted-text role="func"} functions now accept file descriptors in addition to file-like objects. (Contributed by Wei Wu in `23566`{.interpreted-text role="issue"}.)

### functools

::: {#whatsnew-lrucache}
Most of the `~functools.lru_cache`{.interpreted-text role="func"} machinery is now implemented in C, making it significantly faster. (Contributed by Matt Joiner, Alexey Kachayev, and Serhiy Storchaka in `14373`{.interpreted-text role="issue"}.)
:::

### glob

The `~glob.iglob`{.interpreted-text role="func"} and `~glob.glob`{.interpreted-text role="func"} functions now support recursive search in subdirectories, using the `"**"` pattern. (Contributed by Serhiy Storchaka in `13968`{.interpreted-text role="issue"}.)

### gzip

The *mode* argument of the `~gzip.GzipFile`{.interpreted-text role="class"} constructor now accepts `"x"` to request exclusive creation. (Contributed by Tim Heaney in `19222`{.interpreted-text role="issue"}.)

### heapq

Element comparison in `~heapq.merge`{.interpreted-text role="func"} can now be customized by passing a `key function`{.interpreted-text role="term"} in a new optional *key* keyword argument, and a new optional *reverse* keyword argument can be used to reverse element comparison:

    >>> import heapq
    >>> a = ['9', '777', '55555']
    >>> b = ['88', '6666']
    >>> list(heapq.merge(a, b, key=len))
    ['9', '88', '777', '6666', '55555']
    >>> list(heapq.merge(reversed(a), reversed(b), key=len, reverse=True))
    ['55555', '6666', '777', '88', '9']

(Contributed by Raymond Hettinger in `13742`{.interpreted-text role="issue"}.)

### http

A new `HTTPStatus <http.HTTPStatus>`{.interpreted-text role="class"} enum that defines a set of HTTP status codes, reason phrases and long descriptions written in English. (Contributed by Demian Brecht in `21793`{.interpreted-text role="issue"}.)

### http.client

`HTTPConnection.getresponse() <http.client.HTTPConnection.getresponse>`{.interpreted-text role="meth"} now raises a `~http.client.RemoteDisconnected`{.interpreted-text role="exc"} exception when a remote server connection is closed unexpectedly. Additionally, if a `ConnectionError`{.interpreted-text role="exc"} (of which `RemoteDisconnected` is a subclass) is raised, the client socket is now closed automatically, and will reconnect on the next request:

    import http.client
    conn = http.client.HTTPConnection('www.python.org')
    for retries in range(3):
        try:
            conn.request('GET', '/')
            resp = conn.getresponse()
        except http.client.RemoteDisconnected:
            pass

(Contributed by Martin Panter in `3566`{.interpreted-text role="issue"}.)

### idlelib and IDLE

Since idlelib implements the IDLE shell and editor and is not intended for import by other programs, it gets improvements with every release. See `Lib/idlelib/NEWS.txt`{.interpreted-text role="file"} for a cumulative list of changes since 3.4.0, as well as changes made in future 3.5.x releases. This file is also available from the IDLE `Help --> About IDLE`{.interpreted-text role="menuselection"} dialog.

### imaplib

The `~imaplib.IMAP4`{.interpreted-text role="class"} class now supports the `context manager`{.interpreted-text role="term"} protocol. When used in a `with`{.interpreted-text role="keyword"} statement, the IMAP4 `LOGOUT` command will be called automatically at the end of the block. (Contributed by Tarek Ziadé and Serhiy Storchaka in `4972`{.interpreted-text role="issue"}.)

The `imaplib`{.interpreted-text role="mod"} module now supports `5161`{.interpreted-text role="rfc"} (ENABLE Extension) and `6855`{.interpreted-text role="rfc"} (UTF-8 Support) via the `IMAP4.enable() <imaplib.IMAP4.enable>`{.interpreted-text role="meth"} method. A new `IMAP4.utf8_enabled <imaplib.IMAP4.utf8_enabled>`{.interpreted-text role="attr"} attribute tracks whether or not `6855`{.interpreted-text role="rfc"} support is enabled. (Contributed by Milan Oberkirch, R. David Murray, and Maciej Szulik in `21800`{.interpreted-text role="issue"}.)

The `imaplib`{.interpreted-text role="mod"} module now automatically encodes non-ASCII string usernames and passwords using UTF-8, as recommended by the RFCs. (Contributed by Milan Oberkirch in `21800`{.interpreted-text role="issue"}.)

### imghdr

The `!what`{.interpreted-text role="func"} function now recognizes the [OpenEXR](https://www.openexr.com) format (contributed by Martin Vignali and Claudiu Popa in `20295`{.interpreted-text role="issue"}), and the [WebP](https://en.wikipedia.org/wiki/WebP) format (contributed by Fabrice Aneche and Claudiu Popa in `20197`{.interpreted-text role="issue"}.)

### importlib

The `util.LazyLoader <importlib.util.LazyLoader>`{.interpreted-text role="class"} class allows for lazy loading of modules in applications where startup time is important. (Contributed by Brett Cannon in `17621`{.interpreted-text role="issue"}.)

The `abc.InspectLoader.source_to_code() <importlib.abc.InspectLoader.source_to_code>`{.interpreted-text role="func"} method is now a static method. This makes it easier to initialize a module object with code compiled from a string by running `exec(code, module.__dict__)`. (Contributed by Brett Cannon in `21156`{.interpreted-text role="issue"}.)

The new `util.module_from_spec() <importlib.util.module_from_spec>`{.interpreted-text role="func"} function is now the preferred way to create a new module. As opposed to creating a `types.ModuleType`{.interpreted-text role="class"} instance directly, this new function will set the various import-controlled attributes based on the passed-in spec object. (Contributed by Brett Cannon in `20383`{.interpreted-text role="issue"}.)

### inspect

Both the `~inspect.Signature`{.interpreted-text role="class"} and `~inspect.Parameter`{.interpreted-text role="class"} classes are now picklable and hashable. (Contributed by Yury Selivanov in `20726`{.interpreted-text role="issue"} and `20334`{.interpreted-text role="issue"}.)

A new `BoundArguments.apply_defaults() <inspect.BoundArguments.apply_defaults>`{.interpreted-text role="meth"} method provides a way to set default values for missing arguments:

    >>> def foo(a, b='ham', *args): pass
    >>> ba = inspect.signature(foo).bind('spam')
    >>> ba.apply_defaults()
    >>> ba.arguments
    OrderedDict([('a', 'spam'), ('b', 'ham'), ('args', ())])

(Contributed by Yury Selivanov in `24190`{.interpreted-text role="issue"}.)

A new class method `Signature.from_callable() <inspect.Signature.from_callable>`{.interpreted-text role="meth"} makes subclassing of `~inspect.Signature`{.interpreted-text role="class"} easier. (Contributed by Yury Selivanov and Eric Snow in `17373`{.interpreted-text role="issue"}.)

The `~inspect.signature`{.interpreted-text role="func"} function now accepts a *follow_wrapped* optional keyword argument, which, when set to `False`, disables automatic following of `__wrapped__` links. (Contributed by Yury Selivanov in `20691`{.interpreted-text role="issue"}.)

A set of new functions to inspect `coroutine functions <coroutine function>`{.interpreted-text role="term"} and `coroutine objects <coroutine>`{.interpreted-text role="term"} has been added: `~inspect.iscoroutine`{.interpreted-text role="func"}, `~inspect.iscoroutinefunction`{.interpreted-text role="func"}, `~inspect.isawaitable`{.interpreted-text role="func"}, `~inspect.getcoroutinelocals`{.interpreted-text role="func"}, and `~inspect.getcoroutinestate`{.interpreted-text role="func"}. (Contributed by Yury Selivanov in `24017`{.interpreted-text role="issue"} and `24400`{.interpreted-text role="issue"}.)

The `~inspect.stack`{.interpreted-text role="func"}, `~inspect.trace`{.interpreted-text role="func"}, `~inspect.getouterframes`{.interpreted-text role="func"}, and `~inspect.getinnerframes`{.interpreted-text role="func"} functions now return a list of named tuples. (Contributed by Daniel Shahaf in `16808`{.interpreted-text role="issue"}.)

### io

A new `BufferedIOBase.readinto1() <io.BufferedIOBase.readinto1>`{.interpreted-text role="meth"} method, that uses at most one call to the underlying raw stream\'s `RawIOBase.read() <io.RawIOBase.read>`{.interpreted-text role="meth"} or `RawIOBase.readinto() <io.RawIOBase.readinto>`{.interpreted-text role="meth"} methods. (Contributed by Nikolaus Rath in `20578`{.interpreted-text role="issue"}.)

### ipaddress

Both the `~ipaddress.IPv4Network`{.interpreted-text role="class"} and `~ipaddress.IPv6Network`{.interpreted-text role="class"} classes now accept an `(address, netmask)` tuple argument, so as to easily construct network objects from existing addresses:

    >>> import ipaddress
    >>> ipaddress.IPv4Network(('127.0.0.0', 8))
    IPv4Network('127.0.0.0/8')
    >>> ipaddress.IPv4Network(('127.0.0.0', '255.0.0.0'))
    IPv4Network('127.0.0.0/8')

(Contributed by Peter Moody and Antoine Pitrou in `16531`{.interpreted-text role="issue"}.)

A new `~ipaddress.IPv4Address.reverse_pointer`{.interpreted-text role="attr"} attribute for the `~ipaddress.IPv4Address`{.interpreted-text role="class"} and `~ipaddress.IPv6Address`{.interpreted-text role="class"} classes returns the name of the reverse DNS PTR record:

    >>> import ipaddress
    >>> addr = ipaddress.IPv4Address('127.0.0.1')
    >>> addr.reverse_pointer
    '1.0.0.127.in-addr.arpa'
    >>> addr6 = ipaddress.IPv6Address('::1')
    >>> addr6.reverse_pointer
    '1.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.ip6.arpa'

(Contributed by Leon Weber in `20480`{.interpreted-text role="issue"}.)

### json

The `json.tool`{.interpreted-text role="mod"} command line interface now preserves the order of keys in JSON objects passed in input. The new `--sort-keys` option can be used to sort the keys alphabetically. (Contributed by Berker Peksag in `21650`{.interpreted-text role="issue"}.)

JSON decoder now raises `~json.JSONDecodeError`{.interpreted-text role="exc"} instead of `ValueError`{.interpreted-text role="exc"} to provide better context information about the error. (Contributed by Serhiy Storchaka in `19361`{.interpreted-text role="issue"}.)

### linecache

A new `~linecache.lazycache`{.interpreted-text role="func"} function can be used to capture information about a non-file-based module to permit getting its lines later via `~linecache.getline`{.interpreted-text role="func"}. This avoids doing I/O until a line is actually needed, without having to carry the module globals around indefinitely. (Contributed by Robert Collins in `17911`{.interpreted-text role="issue"}.)

### locale

A new `~locale.delocalize`{.interpreted-text role="func"} function can be used to convert a string into a normalized number string, taking the `LC_NUMERIC` settings into account:

    >>> import locale
    >>> locale.setlocale(locale.LC_NUMERIC, 'de_DE.UTF-8')
    'de_DE.UTF-8'
    >>> locale.delocalize('1.234,56')
    '1234.56'
    >>> locale.setlocale(locale.LC_NUMERIC, 'en_US.UTF-8')
    'en_US.UTF-8'
    >>> locale.delocalize('1,234.56')
    '1234.56'

(Contributed by Cédric Krier in `13918`{.interpreted-text role="issue"}.)

### logging

All logging methods (`~logging.Logger`{.interpreted-text role="class"} `~logging.Logger.log`{.interpreted-text role="meth"}, `~logging.Logger.exception`{.interpreted-text role="meth"}, `~logging.Logger.critical`{.interpreted-text role="meth"}, `~logging.Logger.debug`{.interpreted-text role="meth"}, etc.), now accept exception instances as an *exc_info* argument, in addition to boolean values and exception tuples:

    >>> import logging
    >>> try:
    ...     1/0
    ... except ZeroDivisionError as ex:
    ...     logging.error('exception', exc_info=ex)
    ERROR:root:exception

(Contributed by Yury Selivanov in `20537`{.interpreted-text role="issue"}.)

The `handlers.HTTPHandler <logging.handlers.HTTPHandler>`{.interpreted-text role="class"} class now accepts an optional `ssl.SSLContext`{.interpreted-text role="class"} instance to configure SSL settings used in an HTTP connection. (Contributed by Alex Gaynor in `22788`{.interpreted-text role="issue"}.)

The `handlers.QueueListener <logging.handlers.QueueListener>`{.interpreted-text role="class"} class now takes a *respect_handler_level* keyword argument which, if set to `True`, will pass messages to handlers taking handler levels into account. (Contributed by Vinay Sajip.)

### lzma

The `LZMADecompressor.decompress() <lzma.LZMADecompressor.decompress>`{.interpreted-text role="meth"} method now accepts an optional *max_length* argument to limit the maximum size of decompressed data. (Contributed by Martin Panter in `15955`{.interpreted-text role="issue"}.)

### math

Two new constants have been added to the `math`{.interpreted-text role="mod"} module: `~math.inf`{.interpreted-text role="data"} and `~math.nan`{.interpreted-text role="data"}. (Contributed by Mark Dickinson in `23185`{.interpreted-text role="issue"}.)

A new function `~math.isclose`{.interpreted-text role="func"} provides a way to test for approximate equality. (Contributed by Chris Barker and Tal Einat in `24270`{.interpreted-text role="issue"}.)

A new `~math.gcd`{.interpreted-text role="func"} function has been added. The `!fractions.gcd`{.interpreted-text role="func"} function is now deprecated. (Contributed by Mark Dickinson and Serhiy Storchaka in `22486`{.interpreted-text role="issue"}.)

### multiprocessing

`sharedctypes.synchronized() <multiprocessing.sharedctypes.synchronized>`{.interpreted-text role="func"} objects now support the `context manager`{.interpreted-text role="term"} protocol. (Contributed by Charles-François Natali in `21565`{.interpreted-text role="issue"}.)

### operator

`~operator.attrgetter`{.interpreted-text role="func"}, `~operator.itemgetter`{.interpreted-text role="func"}, and `~operator.methodcaller`{.interpreted-text role="func"} objects now support pickling. (Contributed by Josh Rosenberg and Serhiy Storchaka in `22955`{.interpreted-text role="issue"}.)

New `~operator.matmul`{.interpreted-text role="func"} and `~operator.imatmul`{.interpreted-text role="func"} functions to perform matrix multiplication. (Contributed by Benjamin Peterson in `21176`{.interpreted-text role="issue"}.)

### os

The new `~os.scandir`{.interpreted-text role="func"} function returning an iterator of `~os.DirEntry`{.interpreted-text role="class"} objects has been added. If possible, `~os.scandir`{.interpreted-text role="func"} extracts file attributes while scanning a directory, removing the need to perform subsequent system calls to determine file type or attributes, which may significantly improve performance. (Contributed by Ben Hoyt with the help of Victor Stinner in `22524`{.interpreted-text role="issue"}.)

On Windows, a new `stat_result.st_file_attributes <os.stat_result.st_file_attributes>`{.interpreted-text role="attr"} attribute is now available. It corresponds to the `dwFileAttributes` member of the `BY_HANDLE_FILE_INFORMATION` structure returned by `GetFileInformationByHandle()`. (Contributed by Ben Hoyt in `21719`{.interpreted-text role="issue"}.)

The `~os.urandom`{.interpreted-text role="func"} function now uses the `getrandom()` syscall on Linux 3.17 or newer, and `getentropy()` on OpenBSD 5.6 and newer, removing the need to use `/dev/urandom` and avoiding failures due to potential file descriptor exhaustion. (Contributed by Victor Stinner in `22181`{.interpreted-text role="issue"}.)

New `~os.get_blocking`{.interpreted-text role="func"} and `~os.set_blocking`{.interpreted-text role="func"} functions allow getting and setting a file descriptor\'s blocking mode (`~os.O_NONBLOCK`{.interpreted-text role="const"}.) (Contributed by Victor Stinner in `22054`{.interpreted-text role="issue"}.)

The `~os.truncate`{.interpreted-text role="func"} and `~os.ftruncate`{.interpreted-text role="func"} functions are now supported on Windows. (Contributed by Steve Dower in `23668`{.interpreted-text role="issue"}.)

There is a new `os.path.commonpath`{.interpreted-text role="func"} function returning the longest common sub-path of each passed pathname. Unlike the `os.path.commonprefix`{.interpreted-text role="func"} function, it always returns a valid path:

    >>> os.path.commonprefix(['/usr/lib', '/usr/local/lib'])
    '/usr/l'

    >>> os.path.commonpath(['/usr/lib', '/usr/local/lib'])
    '/usr'

(Contributed by Rafik Draoui and Serhiy Storchaka in `10395`{.interpreted-text role="issue"}.)

### pathlib

The new `Path.samefile() <pathlib.Path.samefile>`{.interpreted-text role="meth"} method can be used to check whether the path points to the same file as another path, which can be either another `~pathlib.Path`{.interpreted-text role="class"} object, or a string:

    >>> import pathlib
    >>> p1 = pathlib.Path('/etc/hosts')
    >>> p2 = pathlib.Path('/etc/../etc/hosts')
    >>> p1.samefile(p2)
    True

(Contributed by Vajrasky Kok and Antoine Pitrou in `19775`{.interpreted-text role="issue"}.)

The `Path.mkdir() <pathlib.Path.mkdir>`{.interpreted-text role="meth"} method now accepts a new optional *exist_ok* argument to match `mkdir -p` and `os.makedirs`{.interpreted-text role="func"} functionality. (Contributed by Berker Peksag in `21539`{.interpreted-text role="issue"}.)

There is a new `Path.expanduser() <pathlib.Path.expanduser>`{.interpreted-text role="meth"} method to expand `~` and `~user` prefixes. (Contributed by Serhiy Storchaka and Claudiu Popa in `19776`{.interpreted-text role="issue"}.)

A new `Path.home() <pathlib.Path.home>`{.interpreted-text role="meth"} class method can be used to get a `~pathlib.Path`{.interpreted-text role="class"} instance representing the user's home directory. (Contributed by Victor Salgado and Mayank Tripathi in `19777`{.interpreted-text role="issue"}.)

New `Path.write_text() <pathlib.Path.write_text>`{.interpreted-text role="meth"}, `Path.read_text() <pathlib.Path.read_text>`{.interpreted-text role="meth"}, `Path.write_bytes() <pathlib.Path.write_bytes>`{.interpreted-text role="meth"}, `Path.read_bytes() <pathlib.Path.read_bytes>`{.interpreted-text role="meth"} methods to simplify read/write operations on files.

The following code snippet will create or rewrite existing file `~/spam42`:

    >>> import pathlib
    >>> p = pathlib.Path('~/spam42')
    >>> p.expanduser().write_text('ham')
    3

(Contributed by Christopher Welborn in `20218`{.interpreted-text role="issue"}.)

### pickle

Nested objects, such as unbound methods or nested classes, can now be pickled using `pickle protocols <pickle-protocols>`{.interpreted-text role="ref"} older than protocol version 4. Protocol version 4 already supports these cases. (Contributed by Serhiy Storchaka in `23611`{.interpreted-text role="issue"}.)

### poplib

A new `POP3.utf8() <poplib.POP3.utf8>`{.interpreted-text role="meth"} command enables `6856`{.interpreted-text role="rfc"} (Internationalized Email) support, if a POP server supports it. (Contributed by Milan OberKirch in `21804`{.interpreted-text role="issue"}.)

### re

References and conditional references to groups with fixed length are now allowed in lookbehind assertions:

    >>> import re
    >>> pat = re.compile(r'(a|b).(?<=\1)c')
    >>> pat.match('aac')
    <_sre.SRE_Match object; span=(0, 3), match='aac'>
    >>> pat.match('bbc')
    <_sre.SRE_Match object; span=(0, 3), match='bbc'>

(Contributed by Serhiy Storchaka in `9179`{.interpreted-text role="issue"}.)

The number of capturing groups in regular expressions is no longer limited to 100. (Contributed by Serhiy Storchaka in `22437`{.interpreted-text role="issue"}.)

The `~re.sub`{.interpreted-text role="func"} and `~re.subn`{.interpreted-text role="func"} functions now replace unmatched groups with empty strings instead of raising an exception. (Contributed by Serhiy Storchaka in `1519638`{.interpreted-text role="issue"}.)

The `re.error <re.PatternError>`{.interpreted-text role="class"} exceptions have new attributes, `~re.PatternError.msg`{.interpreted-text role="attr"}, `~re.PatternError.pattern`{.interpreted-text role="attr"}, `~re.PatternError.pos`{.interpreted-text role="attr"}, `~re.PatternError.lineno`{.interpreted-text role="attr"}, and `~re.PatternError.colno`{.interpreted-text role="attr"}, that provide better context information about the error:

    >>> re.compile("""
    ...     (?x)
    ...     .++
    ... """)
    Traceback (most recent call last):
       ...
    sre_constants.error: multiple repeat at position 16 (line 3, column 7)

(Contributed by Serhiy Storchaka in `22578`{.interpreted-text role="issue"}.)

### readline

A new `~readline.append_history_file`{.interpreted-text role="func"} function can be used to append the specified number of trailing elements in history to the given file. (Contributed by Bruno Cauet in `22940`{.interpreted-text role="issue"}.)

### selectors

The new `~selectors.DevpollSelector`{.interpreted-text role="class"} supports efficient `/dev/poll` polling on Solaris. (Contributed by Giampaolo Rodola\' in `18931`{.interpreted-text role="issue"}.)

### shutil

The `~shutil.move`{.interpreted-text role="func"} function now accepts a *copy_function* argument, allowing, for example, the `~shutil.copy`{.interpreted-text role="func"} function to be used instead of the default `~shutil.copy2`{.interpreted-text role="func"} if there is a need to ignore file metadata when moving. (Contributed by Claudiu Popa in `19840`{.interpreted-text role="issue"}.)

The `~shutil.make_archive`{.interpreted-text role="func"} function now supports the *xztar* format. (Contributed by Serhiy Storchaka in `5411`{.interpreted-text role="issue"}.)

### signal

On Windows, the `~signal.set_wakeup_fd`{.interpreted-text role="func"} function now also supports socket handles. (Contributed by Victor Stinner in `22018`{.interpreted-text role="issue"}.)

Various `SIG*` constants in the `signal`{.interpreted-text role="mod"} module have been converted into `Enums <enum>`{.interpreted-text role="mod"}. This allows meaningful names to be printed during debugging, instead of integer \"magic numbers\". (Contributed by Giampaolo Rodola\' in `21076`{.interpreted-text role="issue"}.)

### smtpd

Both the `!SMTPServer`{.interpreted-text role="class"} and `!SMTPChannel`{.interpreted-text role="class"} classes now accept a *decode_data* keyword argument to determine if the `DATA` portion of the SMTP transaction is decoded using the `"utf-8"` codec or is instead provided to the `!SMTPServer.process_message`{.interpreted-text role="meth"} method as a byte string. The default is `True` for backward compatibility reasons, but will change to `False` in Python 3.6. If *decode_data* is set to `False`, the `process_message` method must be prepared to accept keyword arguments. (Contributed by Maciej Szulik in `19662`{.interpreted-text role="issue"}.)

The `!SMTPServer`{.interpreted-text role="class"} class now advertises the `8BITMIME` extension (`6152`{.interpreted-text role="rfc"}) if *decode_data* has been set `True`. If the client specifies `BODY=8BITMIME` on the `MAIL` command, it is passed to `!SMTPServer.process_message`{.interpreted-text role="meth"} via the *mail_options* keyword. (Contributed by Milan Oberkirch and R. David Murray in `21795`{.interpreted-text role="issue"}.)

The `!SMTPServer`{.interpreted-text role="class"} class now also supports the `SMTPUTF8` extension (`6531`{.interpreted-text role="rfc"}: Internationalized Email). If the client specified `SMTPUTF8 BODY=8BITMIME` on the `MAIL` command, they are passed to `!SMTPServer.process_message`{.interpreted-text role="meth"} via the *mail_options* keyword. It is the responsibility of the `process_message` method to correctly handle the `SMTPUTF8` data. (Contributed by Milan Oberkirch in `21725`{.interpreted-text role="issue"}.)

It is now possible to provide, directly or via name resolution, IPv6 addresses in the `!SMTPServer`{.interpreted-text role="class"} constructor, and have it successfully connect. (Contributed by Milan Oberkirch in `14758`{.interpreted-text role="issue"}.)

### smtplib

A new `SMTP.auth() <smtplib.SMTP.auth>`{.interpreted-text role="meth"} method provides a convenient way to implement custom authentication mechanisms. (Contributed by Milan Oberkirch in `15014`{.interpreted-text role="issue"}.)

The `SMTP.set_debuglevel() <smtplib.SMTP.set_debuglevel>`{.interpreted-text role="meth"} method now accepts an additional debuglevel (2), which enables timestamps in debug messages. (Contributed by Gavin Chappell and Maciej Szulik in `16914`{.interpreted-text role="issue"}.)

Both the `SMTP.sendmail() <smtplib.SMTP.sendmail>`{.interpreted-text role="meth"} and `SMTP.send_message() <smtplib.SMTP.send_message>`{.interpreted-text role="meth"} methods now support `6531`{.interpreted-text role="rfc"} (SMTPUTF8). (Contributed by Milan Oberkirch and R. David Murray in `22027`{.interpreted-text role="issue"}.)

### sndhdr

The `!what`{.interpreted-text role="func"} and `!whathdr`{.interpreted-text role="func"} functions now return a `~collections.namedtuple`{.interpreted-text role="func"}. (Contributed by Claudiu Popa in `18615`{.interpreted-text role="issue"}.)

### socket

Functions with timeouts now use a monotonic clock, instead of a system clock. (Contributed by Victor Stinner in `22043`{.interpreted-text role="issue"}.)

A new `socket.sendfile() <socket.socket.sendfile>`{.interpreted-text role="meth"} method allows sending a file over a socket by using the high-performance `os.sendfile`{.interpreted-text role="func"} function on UNIX, resulting in uploads being from 2 to 3 times faster than when using plain `socket.send() <socket.socket.send>`{.interpreted-text role="meth"}. (Contributed by Giampaolo Rodola\' in `17552`{.interpreted-text role="issue"}.)

The `socket.sendall() <socket.socket.sendall>`{.interpreted-text role="meth"} method no longer resets the socket timeout every time bytes are received or sent. The socket timeout is now the maximum total duration to send all data. (Contributed by Victor Stinner in `23853`{.interpreted-text role="issue"}.)

The *backlog* argument of the `socket.listen() <socket.socket.listen>`{.interpreted-text role="meth"} method is now optional. By default it is set to `SOMAXCONN <socket.SOMAXCONN>`{.interpreted-text role="data"} or to `128`, whichever is less. (Contributed by Charles-François Natali in `21455`{.interpreted-text role="issue"}.)

### ssl

#### Memory BIO Support {#whatsnew-sslmemorybio}

(Contributed by Geert Jansen in `21965`{.interpreted-text role="issue"}.)

The new `~ssl.SSLObject`{.interpreted-text role="class"} class has been added to provide SSL protocol support for cases when the network I/O capabilities of `~ssl.SSLSocket`{.interpreted-text role="class"} are not necessary or are suboptimal. `SSLObject` represents an SSL protocol instance, but does not implement any network I/O methods, and instead provides a memory buffer interface. The new `~ssl.MemoryBIO`{.interpreted-text role="class"} class can be used to pass data between Python and an SSL protocol instance.

The memory BIO SSL support is primarily intended to be used in frameworks implementing asynchronous I/O for which `~ssl.SSLSocket`{.interpreted-text role="class"}\'s readiness model (\"select/poll\") is inefficient.

A new `SSLContext.wrap_bio() <ssl.SSLContext.wrap_bio>`{.interpreted-text role="meth"} method can be used to create a new `SSLObject` instance.

#### Application-Layer Protocol Negotiation Support

(Contributed by Benjamin Peterson in `20188`{.interpreted-text role="issue"}.)

Where OpenSSL support is present, the `ssl`{.interpreted-text role="mod"} module now implements the *Application-Layer Protocol Negotiation* TLS extension as described in `7301`{.interpreted-text role="rfc"}.

The new `SSLContext.set_alpn_protocols() <ssl.SSLContext.set_alpn_protocols>`{.interpreted-text role="meth"} can be used to specify which protocols a socket should advertise during the TLS handshake.

The new `SSLSocket.selected_alpn_protocol() <ssl.SSLSocket.selected_alpn_protocol>`{.interpreted-text role="meth"} returns the protocol that was selected during the TLS handshake. The `~ssl.HAS_ALPN`{.interpreted-text role="const"} flag indicates whether ALPN support is present.

#### Other Changes

There is a new `SSLSocket.version() <ssl.SSLSocket.version>`{.interpreted-text role="meth"} method to query the actual protocol version in use. (Contributed by Antoine Pitrou in `20421`{.interpreted-text role="issue"}.)

The `~ssl.SSLSocket`{.interpreted-text role="class"} class now implements a `!SSLSocket.sendfile`{.interpreted-text role="meth"} method. (Contributed by Giampaolo Rodola\' in `17552`{.interpreted-text role="issue"}.)

The `!SSLSocket.send`{.interpreted-text role="meth"} method now raises either the `ssl.SSLWantReadError`{.interpreted-text role="exc"} or `ssl.SSLWantWriteError`{.interpreted-text role="exc"} exception on a non-blocking socket if the operation would block. Previously, it would return `0`. (Contributed by Nikolaus Rath in `20951`{.interpreted-text role="issue"}.)

The `~ssl.cert_time_to_seconds`{.interpreted-text role="func"} function now interprets the input time as UTC and not as local time, per `5280`{.interpreted-text role="rfc"}. Additionally, the return value is always an `int`{.interpreted-text role="class"}. (Contributed by Akira Li in `19940`{.interpreted-text role="issue"}.)

New `!SSLObject.shared_ciphers`{.interpreted-text role="meth"} and `SSLSocket.shared_ciphers() <ssl.SSLSocket.shared_ciphers>`{.interpreted-text role="meth"} methods return the list of ciphers sent by the client during the handshake. (Contributed by Benjamin Peterson in `23186`{.interpreted-text role="issue"}.)

The `SSLSocket.do_handshake() <ssl.SSLSocket.do_handshake>`{.interpreted-text role="meth"}, `SSLSocket.read() <ssl.SSLSocket.read>`{.interpreted-text role="meth"}, `!SSLSocket.shutdown`{.interpreted-text role="meth"}, and `SSLSocket.write() <ssl.SSLSocket.write>`{.interpreted-text role="meth"} methods of the `~ssl.SSLSocket`{.interpreted-text role="class"} class no longer reset the socket timeout every time bytes are received or sent. The socket timeout is now the maximum total duration of the method. (Contributed by Victor Stinner in `23853`{.interpreted-text role="issue"}.)

The `!match_hostname`{.interpreted-text role="func"} function now supports matching of IP addresses. (Contributed by Antoine Pitrou in `23239`{.interpreted-text role="issue"}.)

### sqlite3

The `~sqlite3.Row`{.interpreted-text role="class"} class now fully supports the sequence protocol, in particular `reversed`{.interpreted-text role="func"} iteration and slice indexing. (Contributed by Claudiu Popa in `10203`{.interpreted-text role="issue"}; by Lucas Sinclair, Jessica McKellar, and Serhiy Storchaka in `13583`{.interpreted-text role="issue"}.)

### subprocess {#whatsnew-subprocess}

The new `~subprocess.run`{.interpreted-text role="func"} function has been added. It runs the specified command and returns a `~subprocess.CompletedProcess`{.interpreted-text role="class"} object, which describes a finished process. The new API is more consistent and is the recommended approach to invoking subprocesses in Python code that does not need to maintain compatibility with earlier Python versions. (Contributed by Thomas Kluyver in `23342`{.interpreted-text role="issue"}.)

Examples:

    >>> subprocess.run(["ls", "-l"])  # doesn't capture output
    CompletedProcess(args=['ls', '-l'], returncode=0)

    >>> subprocess.run("exit 1", shell=True, check=True)
    Traceback (most recent call last):
      ...
    subprocess.CalledProcessError: Command 'exit 1' returned non-zero exit status 1

    >>> subprocess.run(["ls", "-l", "/dev/null"], stdout=subprocess.PIPE)
    CompletedProcess(args=['ls', '-l', '/dev/null'], returncode=0,
    stdout=b'crw-rw-rw- 1 root root 1, 3 Jan 23 16:23 /dev/null\n')

### sys

A new `!set_coroutine_wrapper`{.interpreted-text role="func"} function allows setting a global hook that will be called whenever a `coroutine object <coroutine>`{.interpreted-text role="term"} is created by an `async def`{.interpreted-text role="keyword"} function. A corresponding `!get_coroutine_wrapper`{.interpreted-text role="func"} can be used to obtain a currently set wrapper. Both functions are `provisional <provisional API>`{.interpreted-text role="term"}, and are intended for debugging purposes only. (Contributed by Yury Selivanov in `24017`{.interpreted-text role="issue"}.)

A new `~sys.is_finalizing`{.interpreted-text role="func"} function can be used to check if the Python interpreter is `shutting down <interpreter shutdown>`{.interpreted-text role="term"}. (Contributed by Antoine Pitrou in `22696`{.interpreted-text role="issue"}.)

### sysconfig

The name of the user scripts directory on Windows now includes the first two components of the Python version. (Contributed by Paul Moore in `23437`{.interpreted-text role="issue"}.)

### tarfile

The *mode* argument of the `~tarfile.open`{.interpreted-text role="func"} function now accepts `"x"` to request exclusive creation. (Contributed by Berker Peksag in `21717`{.interpreted-text role="issue"}.)

The `TarFile.extractall() <tarfile.TarFile.extractall>`{.interpreted-text role="meth"} and `TarFile.extract() <tarfile.TarFile.extract>`{.interpreted-text role="meth"} methods now take a keyword argument *numeric_owner*. If set to `True`, the extracted files and directories will be owned by the numeric `uid` and `gid` from the tarfile. If set to `False` (the default, and the behavior in versions prior to 3.5), they will be owned by the named user and group in the tarfile. (Contributed by Michael Vogt and Eric Smith in `23193`{.interpreted-text role="issue"}.)

The `TarFile.list() <tarfile.TarFile.list>`{.interpreted-text role="meth"} now accepts an optional *members* keyword argument that can be set to a subset of the list returned by `TarFile.getmembers() <tarfile.TarFile.getmembers>`{.interpreted-text role="meth"}. (Contributed by Serhiy Storchaka in `21549`{.interpreted-text role="issue"}.)

### threading

Both the `Lock.acquire() <threading.Lock.acquire>`{.interpreted-text role="meth"} and `RLock.acquire() <threading.RLock.acquire>`{.interpreted-text role="meth"} methods now use a monotonic clock for timeout management. (Contributed by Victor Stinner in `22043`{.interpreted-text role="issue"}.)

### time

The `~time.monotonic`{.interpreted-text role="func"} function is now always available. (Contributed by Victor Stinner in `22043`{.interpreted-text role="issue"}.)

### timeit

A new command line option `-u` or `--unit={U}`{.interpreted-text role="samp"} can be used to specify the time unit for the timer output. Supported options are `usec`, `msec`, or `sec`. (Contributed by Julian Gindi in `18983`{.interpreted-text role="issue"}.)

The `~timeit.timeit`{.interpreted-text role="func"} function has a new *globals* parameter for specifying the namespace in which the code will be running. (Contributed by Ben Roberts in `2527`{.interpreted-text role="issue"}.)

### tkinter

The `!tkinter._fix`{.interpreted-text role="mod"} module used for setting up the Tcl/Tk environment on Windows has been replaced by a private function in the `!_tkinter`{.interpreted-text role="mod"} module which makes no permanent changes to environment variables. (Contributed by Zachary Ware in `20035`{.interpreted-text role="issue"}.)

### traceback {#whatsnew-traceback}

New `~traceback.walk_stack`{.interpreted-text role="func"} and `~traceback.walk_tb`{.interpreted-text role="func"} functions to conveniently traverse frame and `traceback objects <traceback-objects>`{.interpreted-text role="ref"}. (Contributed by Robert Collins in `17911`{.interpreted-text role="issue"}.)

New lightweight classes: `~traceback.TracebackException`{.interpreted-text role="class"}, `~traceback.StackSummary`{.interpreted-text role="class"}, and `~traceback.FrameSummary`{.interpreted-text role="class"}. (Contributed by Robert Collins in `17911`{.interpreted-text role="issue"}.)

Both the `~traceback.print_tb`{.interpreted-text role="func"} and `~traceback.print_stack`{.interpreted-text role="func"} functions now support negative values for the *limit* argument. (Contributed by Dmitry Kazakov in `22619`{.interpreted-text role="issue"}.)

### types

A new `~types.coroutine`{.interpreted-text role="func"} function to transform `generator <generator iterator>`{.interpreted-text role="term"} and `generator-like <collections.abc.Generator>`{.interpreted-text role="class"} objects into `awaitables <awaitable>`{.interpreted-text role="term"}. (Contributed by Yury Selivanov in `24017`{.interpreted-text role="issue"}.)

A new type called `~types.CoroutineType`{.interpreted-text role="class"}, which is used for `coroutine`{.interpreted-text role="term"} objects created by `async def`{.interpreted-text role="keyword"} functions. (Contributed by Yury Selivanov in `24400`{.interpreted-text role="issue"}.)

### unicodedata

The `unicodedata`{.interpreted-text role="mod"} module now uses data from [Unicode 8.0.0](https://unicode.org/versions/Unicode8.0.0/).

### unittest

The `TestLoader.loadTestsFromModule() <unittest.TestLoader.loadTestsFromModule>`{.interpreted-text role="meth"} method now accepts a keyword-only argument *pattern* which is passed to `load_tests` as the third argument. Found packages are now checked for `load_tests` regardless of whether their path matches *pattern*, because it is impossible for a package name to match the default pattern. (Contributed by Robert Collins and Barry A. Warsaw in `16662`{.interpreted-text role="issue"}.)

Unittest discovery errors now are exposed in the `TestLoader.errors <unittest.TestLoader.errors>`{.interpreted-text role="data"} attribute of the `~unittest.TestLoader`{.interpreted-text role="class"} instance. (Contributed by Robert Collins in `19746`{.interpreted-text role="issue"}.)

A new command line option `--locals` to show local variables in tracebacks. (Contributed by Robert Collins in `22936`{.interpreted-text role="issue"}.)

### unittest.mock

The `~unittest.mock.Mock`{.interpreted-text role="class"} class has the following improvements:

- The class constructor has a new *unsafe* parameter, which causes mock objects to raise `AttributeError`{.interpreted-text role="exc"} on attribute names starting with `"assert"`. (Contributed by Kushal Das in `21238`{.interpreted-text role="issue"}.)
- A new `Mock.assert_not_called() <unittest.mock.Mock.assert_not_called>`{.interpreted-text role="meth"} method to check if the mock object was called. (Contributed by Kushal Das in `21262`{.interpreted-text role="issue"}.)

The `~unittest.mock.MagicMock`{.interpreted-text role="class"} class now supports `~object.__truediv__`{.interpreted-text role="meth"}, `~object.__divmod__`{.interpreted-text role="meth"} and `~object.__matmul__`{.interpreted-text role="meth"} operators. (Contributed by Johannes Baiter in `20968`{.interpreted-text role="issue"}, and Håkan Lövdahl in `23581`{.interpreted-text role="issue"} and `23568`{.interpreted-text role="issue"}.)

It is no longer necessary to explicitly pass `create=True` to the `~unittest.mock.patch`{.interpreted-text role="func"} function when patching builtin names. (Contributed by Kushal Das in `17660`{.interpreted-text role="issue"}.)

### urllib

A new `request.HTTPPasswordMgrWithPriorAuth <urllib.request.HTTPPasswordMgrWithPriorAuth>`{.interpreted-text role="class"} class allows HTTP Basic Authentication credentials to be managed so as to eliminate unnecessary `401` response handling, or to unconditionally send credentials on the first request in order to communicate with servers that return a `404` response instead of a `401` if the `Authorization` header is not sent. (Contributed by Matej Cepl in `19494`{.interpreted-text role="issue"} and Akshit Khurana in `7159`{.interpreted-text role="issue"}.)

A new *quote_via* argument for the `parse.urlencode() <urllib.parse.urlencode>`{.interpreted-text role="func"} function provides a way to control the encoding of query parts if needed. (Contributed by Samwyse and Arnon Yaari in `13866`{.interpreted-text role="issue"}.)

The `request.urlopen() <urllib.request.urlopen>`{.interpreted-text role="func"} function accepts an `ssl.SSLContext`{.interpreted-text role="class"} object as a *context* argument, which will be used for the HTTPS connection. (Contributed by Alex Gaynor in `22366`{.interpreted-text role="issue"}.)

The `parse.urljoin() <urllib.parse.urljoin>`{.interpreted-text role="func"} was updated to use the `3986`{.interpreted-text role="rfc"} semantics for the resolution of relative URLs, rather than `1808`{.interpreted-text role="rfc"} and `2396`{.interpreted-text role="rfc"}. (Contributed by Demian Brecht and Senthil Kumaran in `22118`{.interpreted-text role="issue"}.)

### wsgiref

The *headers* argument of the `headers.Headers <wsgiref.headers.Headers>`{.interpreted-text role="class"} class constructor is now optional. (Contributed by Pablo Torres Navarrete and SilentGhost in `5800`{.interpreted-text role="issue"}.)

### xmlrpc

The `client.ServerProxy <xmlrpc.client.ServerProxy>`{.interpreted-text role="class"} class now supports the `context manager`{.interpreted-text role="term"} protocol. (Contributed by Claudiu Popa in `20627`{.interpreted-text role="issue"}.)

The `client.ServerProxy <xmlrpc.client.ServerProxy>`{.interpreted-text role="class"} constructor now accepts an optional `ssl.SSLContext`{.interpreted-text role="class"} instance. (Contributed by Alex Gaynor in `22960`{.interpreted-text role="issue"}.)

### xml.sax

SAX parsers now support a character stream of the `xmlreader.InputSource <xml.sax.xmlreader.InputSource>`{.interpreted-text role="class"} object. (Contributed by Serhiy Storchaka in `2175`{.interpreted-text role="issue"}.)

`~xml.sax.parseString`{.interpreted-text role="func"} now accepts a `str`{.interpreted-text role="class"} instance. (Contributed by Serhiy Storchaka in `10590`{.interpreted-text role="issue"}.)

### zipfile

ZIP output can now be written to unseekable streams. (Contributed by Serhiy Storchaka in `23252`{.interpreted-text role="issue"}.)

The *mode* argument of `ZipFile.open() <zipfile.ZipFile.open>`{.interpreted-text role="meth"} method now accepts `"x"` to request exclusive creation. (Contributed by Serhiy Storchaka in `21717`{.interpreted-text role="issue"}.)

## Other module-level changes

Many functions in the `mmap`{.interpreted-text role="mod"}, `!ossaudiodev`{.interpreted-text role="mod"}, `socket`{.interpreted-text role="mod"}, `ssl`{.interpreted-text role="mod"}, and `codecs`{.interpreted-text role="mod"} modules now accept writable `bytes-like objects <bytes-like object>`{.interpreted-text role="term"}. (Contributed by Serhiy Storchaka in `23001`{.interpreted-text role="issue"}.)

## Optimizations

The `os.walk`{.interpreted-text role="func"} function has been sped up by 3 to 5 times on POSIX systems, and by 7 to 20 times on Windows. This was done using the new `os.scandir`{.interpreted-text role="func"} function, which exposes file information from the underlying `readdir` or `FindFirstFile`/`FindNextFile` system calls. (Contributed by Ben Hoyt with help from Victor Stinner in `23605`{.interpreted-text role="issue"}.)

Construction of `bytes(int)` (filled by zero bytes) is faster and uses less memory for large objects. `calloc()` is used instead of `malloc()` to allocate memory for these objects. (Contributed by Victor Stinner in `21233`{.interpreted-text role="issue"}.)

Some operations on `ipaddress`{.interpreted-text role="mod"} `~ipaddress.IPv4Network`{.interpreted-text role="class"} and `~ipaddress.IPv6Network`{.interpreted-text role="class"} have been massively sped up, such as `~ipaddress.IPv4Network.subnets`{.interpreted-text role="meth"}, `~ipaddress.IPv4Network.supernet`{.interpreted-text role="meth"}, `~ipaddress.summarize_address_range`{.interpreted-text role="func"}, `~ipaddress.collapse_addresses`{.interpreted-text role="func"}. The speed up can range from 3 to 15 times. (Contributed by Antoine Pitrou, Michel Albert, and Markus in `21486`{.interpreted-text role="issue"}, `21487`{.interpreted-text role="issue"}, `20826`{.interpreted-text role="issue"}, `23266`{.interpreted-text role="issue"}.)

Pickling of `ipaddress`{.interpreted-text role="mod"} objects was optimized to produce significantly smaller output. (Contributed by Serhiy Storchaka in `23133`{.interpreted-text role="issue"}.)

Many operations on `io.BytesIO`{.interpreted-text role="class"} are now 50% to 100% faster. (Contributed by Serhiy Storchaka in `15381`{.interpreted-text role="issue"} and David Wilson in `22003`{.interpreted-text role="issue"}.)

The `marshal.dumps`{.interpreted-text role="func"} function is now faster: 65\--85% with versions 3 and 4, 20\--25% with versions 0 to 2 on typical data, and up to 5 times in best cases. (Contributed by Serhiy Storchaka in `20416`{.interpreted-text role="issue"} and `23344`{.interpreted-text role="issue"}.)

The UTF-32 encoder is now 3 to 7 times faster. (Contributed by Serhiy Storchaka in `15027`{.interpreted-text role="issue"}.)

Regular expressions are now parsed up to 10% faster. (Contributed by Serhiy Storchaka in `19380`{.interpreted-text role="issue"}.)

The `json.dumps`{.interpreted-text role="func"} function was optimized to run with `ensure_ascii=False` as fast as with `ensure_ascii=True`. (Contributed by Naoki Inada in `23206`{.interpreted-text role="issue"}.)

The `PyObject_IsInstance`{.interpreted-text role="c:func"} and `PyObject_IsSubclass`{.interpreted-text role="c:func"} functions have been sped up in the common case that the second argument has `type`{.interpreted-text role="class"} as its metaclass. (Contributed Georg Brandl by in `22540`{.interpreted-text role="issue"}.)

Method caching was slightly improved, yielding up to 5% performance improvement in some benchmarks. (Contributed by Antoine Pitrou in `22847`{.interpreted-text role="issue"}.)

Objects from the `random`{.interpreted-text role="mod"} module now use 50% less memory on 64-bit builds. (Contributed by Serhiy Storchaka in `23488`{.interpreted-text role="issue"}.)

The `property`{.interpreted-text role="func"} getter calls are up to 25% faster. (Contributed by Joe Jevnik in `23910`{.interpreted-text role="issue"}.)

Instantiation of `fractions.Fraction`{.interpreted-text role="class"} is now up to 30% faster. (Contributed by Stefan Behnel in `22464`{.interpreted-text role="issue"}.)

String methods `~str.find`{.interpreted-text role="meth"}, `~str.rfind`{.interpreted-text role="meth"}, `~str.split`{.interpreted-text role="meth"}, `~str.partition`{.interpreted-text role="meth"} and the `in`{.interpreted-text role="keyword"} string operator are now significantly faster for searching 1-character substrings. (Contributed by Serhiy Storchaka in `23573`{.interpreted-text role="issue"}.)

## Build and C API Changes

New `calloc` functions were added:

- `PyMem_RawCalloc`{.interpreted-text role="c:func"},
- `PyMem_Calloc`{.interpreted-text role="c:func"},
- `PyObject_Calloc`{.interpreted-text role="c:func"}.

(Contributed by Victor Stinner in `21233`{.interpreted-text role="issue"}.)

New encoding/decoding helper functions:

- `Py_DecodeLocale`{.interpreted-text role="c:func"} (replaced `_Py_char2wchar()`),
- `Py_EncodeLocale`{.interpreted-text role="c:func"} (replaced `_Py_wchar2char()`).

(Contributed by Victor Stinner in `18395`{.interpreted-text role="issue"}.)

A new `PyCodec_NameReplaceErrors`{.interpreted-text role="c:func"} function to replace the unicode encode error with `\N{...}` escapes. (Contributed by Serhiy Storchaka in `19676`{.interpreted-text role="issue"}.)

A new `PyErr_FormatV`{.interpreted-text role="c:func"} function similar to `PyErr_Format`{.interpreted-text role="c:func"}, but accepts a `va_list`{.interpreted-text role="c:type"} argument. (Contributed by Antoine Pitrou in `18711`{.interpreted-text role="issue"}.)

A new `PyExc_RecursionError`{.interpreted-text role="c:data"} exception. (Contributed by Georg Brandl in `19235`{.interpreted-text role="issue"}.)

New `PyModule_FromDefAndSpec`{.interpreted-text role="c:func"}, `PyModule_FromDefAndSpec2`{.interpreted-text role="c:func"}, and `PyModule_ExecDef`{.interpreted-text role="c:func"} functions introduced by `489`{.interpreted-text role="pep"} \-- multi-phase extension module initialization. (Contributed by Petr Viktorin in `24268`{.interpreted-text role="issue"}.)

New `PyNumber_MatrixMultiply`{.interpreted-text role="c:func"} and `PyNumber_InPlaceMatrixMultiply`{.interpreted-text role="c:func"} functions to perform matrix multiplication. (Contributed by Benjamin Peterson in `21176`{.interpreted-text role="issue"}. See also `465`{.interpreted-text role="pep"} for details.)

The `PyTypeObject.tp_finalize`{.interpreted-text role="c:member"} slot is now part of the stable ABI.

Windows builds now require Microsoft Visual C++ 14.0, which is available as part of [Visual Studio 2015](https://visualstudio.microsoft.com/en/vs/older-downloads/#visual-studio-2015-and-other-products).

Extension modules now include a platform information tag in their filename on some platforms (the tag is optional, and CPython will import extensions without it, although if the tag is present and mismatched, the extension won\'t be loaded):

- On Linux, extension module filenames end with `.cpython-<major><minor>m-<architecture>-<os>.pyd`:
  - `<major>` is the major number of the Python version; for Python 3.5 this is `3`.
  - `<minor>` is the minor number of the Python version; for Python 3.5 this is `5`.
  - `<architecture>` is the hardware architecture the extension module was built to run on. It\'s most commonly either `i386` for 32-bit Intel platforms or `x86_64` for 64-bit Intel (and AMD) platforms.
  - `<os>` is always `linux-gnu`, except for extensions built to talk to the 32-bit ABI on 64-bit platforms, in which case it is `linux-gnu32` (and `<architecture>` will be `x86_64`).
- On Windows, extension module filenames end with `<debug>.cp<major><minor>-<platform>.pyd`:
  - `<major>` is the major number of the Python version; for Python 3.5 this is `3`.
  - `<minor>` is the minor number of the Python version; for Python 3.5 this is `5`.
  - `<platform>` is the platform the extension module was built for, either `win32` for Win32, `win_amd64` for Win64, `win_ia64` for Windows Itanium 64, and `win_arm` for Windows on ARM.
  - If built in debug mode, `<debug>` will be `_d`, otherwise it will be blank.
- On OS X platforms, extension module filenames now end with `-darwin.so`.
- On all other platforms, extension module filenames are the same as they were with Python 3.4.

## Deprecated

### New Keywords

`async` and `await` are not recommended to be used as variable, class, function or module names. Introduced by `492`{.interpreted-text role="pep"} in Python 3.5, they will become proper keywords in Python 3.7.

### Deprecated Python Behavior

Raising the `StopIteration`{.interpreted-text role="exc"} exception inside a generator will now generate a silent `PendingDeprecationWarning`{.interpreted-text role="exc"}, which will become a non-silent deprecation warning in Python 3.6 and will trigger a `RuntimeError`{.interpreted-text role="exc"} in Python 3.7. See `PEP 479: Change StopIteration handling inside generators <whatsnew-pep-479>`{.interpreted-text role="ref"} for details.

### Unsupported Operating Systems

Windows XP is no longer supported by Microsoft, thus, per [PEP 11](http://www.python.org/dev/peps/pep-0011/ "PEP 11"), CPython 3.5 is no longer officially supported on this OS.

### Deprecated Python modules, functions and methods

The `!formatter`{.interpreted-text role="mod"} module has now graduated to full deprecation and is still slated for removal in Python 3.6.

The `!asyncio.async`{.interpreted-text role="func"} function is deprecated in favor of `~asyncio.ensure_future`{.interpreted-text role="func"}.

The `!smtpd`{.interpreted-text role="mod"} module has in the past always decoded the DATA portion of email messages using the `utf-8` codec. This can now be controlled by the new *decode_data* keyword to `!SMTPServer`{.interpreted-text role="class"}. The default value is `True`, but this default is deprecated. Specify the *decode_data* keyword with an appropriate value to avoid the deprecation warning.

Directly assigning values to the `~http.cookies.Morsel.key`{.interpreted-text role="attr"}, `~http.cookies.Morsel.value`{.interpreted-text role="attr"} and `~http.cookies.Morsel.coded_value`{.interpreted-text role="attr"} of `http.cookies.Morsel`{.interpreted-text role="class"} objects is deprecated. Use the `~http.cookies.Morsel.set`{.interpreted-text role="meth"} method instead. In addition, the undocumented *LegalChars* parameter of `~http.cookies.Morsel.set`{.interpreted-text role="meth"} is deprecated, and is now ignored.

Passing a format string as keyword argument *format_string* to the `~string.Formatter.format`{.interpreted-text role="meth"} method of the `string.Formatter`{.interpreted-text role="class"} class has been deprecated. (Contributed by Serhiy Storchaka in `23671`{.interpreted-text role="issue"}.)

The `!platform.dist`{.interpreted-text role="func"} and `!platform.linux_distribution`{.interpreted-text role="func"} functions are now deprecated. Linux distributions use too many different ways of describing themselves, so the functionality is left to a package. (Contributed by Vajrasky Kok and Berker Peksag in `1322`{.interpreted-text role="issue"}.)

The previously undocumented `from_function` and `from_builtin` methods of `inspect.Signature`{.interpreted-text role="class"} are deprecated. Use the new `Signature.from_callable() <inspect.Signature.from_callable>`{.interpreted-text role="meth"} method instead. (Contributed by Yury Selivanov in `24248`{.interpreted-text role="issue"}.)

The `!inspect.getargspec`{.interpreted-text role="func"} function is deprecated and scheduled to be removed in Python 3.6. (See `20438`{.interpreted-text role="issue"} for details.)

The `inspect`{.interpreted-text role="mod"} `~inspect.getfullargspec`{.interpreted-text role="func"}, `~inspect.getcallargs`{.interpreted-text role="func"}, and `!formatargspec`{.interpreted-text role="func"} functions are deprecated in favor of the `inspect.signature`{.interpreted-text role="func"} API. (Contributed by Yury Selivanov in `20438`{.interpreted-text role="issue"}.)

`~inspect.getargvalues`{.interpreted-text role="func"} and `~inspect.formatargvalues`{.interpreted-text role="func"} functions were inadvertently marked as deprecated with the release of Python 3.5.0.

Use of `re.LOCALE`{.interpreted-text role="const"} flag with str patterns or `re.ASCII`{.interpreted-text role="const"} is now deprecated. (Contributed by Serhiy Storchaka in `22407`{.interpreted-text role="issue"}.)

Use of unrecognized special sequences consisting of `'\'` and an ASCII letter in regular expression patterns and replacement patterns now raises a deprecation warning and will be forbidden in Python 3.6. (Contributed by Serhiy Storchaka in `23622`{.interpreted-text role="issue"}.)

The undocumented and unofficial *use_load_tests* default argument of the `unittest.TestLoader.loadTestsFromModule`{.interpreted-text role="meth"} method now is deprecated and ignored. (Contributed by Robert Collins and Barry A. Warsaw in `16662`{.interpreted-text role="issue"}.)

## Removed

### API and Feature Removals

The following obsolete and previously deprecated APIs and features have been removed:

- The `__version__` attribute has been dropped from the email package. The email code hasn\'t been shipped separately from the stdlib for a long time, and the `__version__` string was not updated in the last few releases.
- The internal `Netrc` class in the `ftplib`{.interpreted-text role="mod"} module was deprecated in 3.4, and has now been removed. (Contributed by Matt Chaput in `6623`{.interpreted-text role="issue"}.)
- The concept of `.pyo` files has been removed.
- The JoinableQueue class in the provisional `asyncio`{.interpreted-text role="mod"} module was deprecated in 3.4.4 and is now removed. (Contributed by A. Jesse Jiryu Davis in `23464`{.interpreted-text role="issue"}.)

## Porting to Python 3.5

This section lists previously described changes and other bugfixes that may require changes to your code.

### Changes in Python behavior

- Due to an oversight, earlier Python versions erroneously accepted the following syntax:

      f(1 for x in [1], *args)
      f(1 for x in [1], **kwargs)

  Python 3.5 now correctly raises a `SyntaxError`{.interpreted-text role="exc"}, as generator expressions must be put in parentheses if not a sole argument to a function.

### Changes in the Python API

- `475`{.interpreted-text role="pep"}: System calls are now retried when interrupted by a signal instead of raising `InterruptedError`{.interpreted-text role="exc"} if the Python signal handler does not raise an exception.
- Before Python 3.5, a `datetime.time`{.interpreted-text role="class"} object was considered to be false if it represented midnight in UTC. This behavior was considered obscure and error-prone and has been removed in Python 3.5. See `13936`{.interpreted-text role="issue"} for full details.
- The `!ssl.SSLSocket.send`{.interpreted-text role="meth"} method now raises either `ssl.SSLWantReadError`{.interpreted-text role="exc"} or `ssl.SSLWantWriteError`{.interpreted-text role="exc"} on a non-blocking socket if the operation would block. Previously, it would return `0`. (Contributed by Nikolaus Rath in `20951`{.interpreted-text role="issue"}.)
- The `__name__` attribute of generators is now set from the function name, instead of being set from the code name. Use `gen.gi_code.co_name` to retrieve the code name. Generators also have a new `__qualname__` attribute, the qualified name, which is now used for the representation of a generator (`repr(gen)`). (Contributed by Victor Stinner in `21205`{.interpreted-text role="issue"}.)
- The deprecated \"strict\" mode and argument of `~html.parser.HTMLParser`{.interpreted-text role="class"}, `!HTMLParser.error`{.interpreted-text role="meth"}, and the `!HTMLParserError`{.interpreted-text role="exc"} exception have been removed. (Contributed by Ezio Melotti in `15114`{.interpreted-text role="issue"}.) The *convert_charrefs* argument of `~html.parser.HTMLParser`{.interpreted-text role="class"} is now `True` by default. (Contributed by Berker Peksag in `21047`{.interpreted-text role="issue"}.)
- Although it is not formally part of the API, it is worth noting for porting purposes (ie: fixing tests) that error messages that were previously of the form \"\'sometype\' does not support the buffer protocol\" are now of the form \"a `bytes-like object`{.interpreted-text role="term"} is required, not \'sometype\'\". (Contributed by Ezio Melotti in `16518`{.interpreted-text role="issue"}.)
- If the current directory is set to a directory that no longer exists then `FileNotFoundError`{.interpreted-text role="exc"} will no longer be raised and instead `~importlib.machinery.FileFinder.find_spec`{.interpreted-text role="meth"} will return `None` **without** caching `None` in `sys.path_importer_cache`{.interpreted-text role="data"}, which is different than the typical case (`22834`{.interpreted-text role="issue"}).
- HTTP status code and messages from `http.client`{.interpreted-text role="mod"} and `http.server`{.interpreted-text role="mod"} were refactored into a common `~http.HTTPStatus`{.interpreted-text role="class"} enum. The values in `http.client`{.interpreted-text role="mod"} and `http.server`{.interpreted-text role="mod"} remain available for backwards compatibility. (Contributed by Demian Brecht in `21793`{.interpreted-text role="issue"}.)
- When an import loader defines `~importlib.abc.Loader.exec_module`{.interpreted-text role="meth"} it is now expected to also define `~importlib.abc.Loader.create_module`{.interpreted-text role="meth"} (raises a `DeprecationWarning`{.interpreted-text role="exc"} now, will be an error in Python 3.6). If the loader inherits from `importlib.abc.Loader`{.interpreted-text role="class"} then there is nothing to do, else simply define `~importlib.abc.Loader.create_module`{.interpreted-text role="meth"} to return `None`. (Contributed by Brett Cannon in `23014`{.interpreted-text role="issue"}.)
- The `re.split`{.interpreted-text role="func"} function always ignored empty pattern matches, so the `"x*"` pattern worked the same as `"x+"`, and the `"\b"` pattern never worked. Now `re.split`{.interpreted-text role="func"} raises a warning if the pattern could match an empty string. For compatibility, use patterns that never match an empty string (e.g. `"x+"` instead of `"x*"`). Patterns that could only match an empty string (such as `"\b"`) now raise an error. (Contributed by Serhiy Storchaka in `22818`{.interpreted-text role="issue"}.)
- The `http.cookies.Morsel`{.interpreted-text role="class"} dict-like interface has been made self consistent: morsel comparison now takes the `~http.cookies.Morsel.key`{.interpreted-text role="attr"} and `~http.cookies.Morsel.value`{.interpreted-text role="attr"} into account, `~http.cookies.Morsel.copy`{.interpreted-text role="meth"} now results in a `~http.cookies.Morsel`{.interpreted-text role="class"} instance rather than a `dict`{.interpreted-text role="class"}, and `~http.cookies.Morsel.update`{.interpreted-text role="meth"} will now raise an exception if any of the keys in the update dictionary are invalid. In addition, the undocumented *LegalChars* parameter of `~http.cookies.Morsel.set`{.interpreted-text role="func"} is deprecated and is now ignored. (Contributed by Demian Brecht in `2211`{.interpreted-text role="issue"}.)
- `488`{.interpreted-text role="pep"} has removed `.pyo` files from Python and introduced the optional `opt-` tag in `.pyc` file names. The `importlib.util.cache_from_source`{.interpreted-text role="func"} has gained an *optimization* parameter to help control the `opt-` tag. Because of this, the *debug_override* parameter of the function is now deprecated. `.pyo` files are also no longer supported as a file argument to the Python interpreter and thus serve no purpose when distributed on their own (i.e. sourceless code distribution). Due to the fact that the magic number for bytecode has changed in Python 3.5, all old `.pyo` files from previous versions of Python are invalid regardless of this PEP.
- The `socket`{.interpreted-text role="mod"} module now exports the `~socket.CAN_RAW_FD_FRAMES`{.interpreted-text role="const"} constant on linux 3.6 and greater.
- The `ssl.cert_time_to_seconds`{.interpreted-text role="func"} function now interprets the input time as UTC and not as local time, per `5280`{.interpreted-text role="rfc"}. Additionally, the return value is always an `int`{.interpreted-text role="class"}. (Contributed by Akira Li in `19940`{.interpreted-text role="issue"}.)
- The `pygettext.py` Tool now uses the standard +NNNN format for timezones in the POT-Creation-Date header.
- The `smtplib`{.interpreted-text role="mod"} module now uses `sys.stderr`{.interpreted-text role="data"} instead of the previous module-level `!stderr`{.interpreted-text role="data"} variable for debug output. If your (test) program depends on patching the module-level variable to capture the debug output, you will need to update it to capture sys.stderr instead.
- The `str.startswith`{.interpreted-text role="meth"} and `str.endswith`{.interpreted-text role="meth"} methods no longer return `True` when finding the empty string and the indexes are completely out of range. (Contributed by Serhiy Storchaka in `24284`{.interpreted-text role="issue"}.)
- The `inspect.getdoc`{.interpreted-text role="func"} function now returns documentation strings inherited from base classes. Documentation strings no longer need to be duplicated if the inherited documentation is appropriate. To suppress an inherited string, an empty string must be specified (or the documentation may be filled in). This change affects the output of the `pydoc`{.interpreted-text role="mod"} module and the `help`{.interpreted-text role="func"} function. (Contributed by Serhiy Storchaka in `15582`{.interpreted-text role="issue"}.)
- Nested `functools.partial`{.interpreted-text role="func"} calls are now flattened. If you were relying on the previous behavior, you can now either add an attribute to a `functools.partial`{.interpreted-text role="func"} object or you can create a subclass of `functools.partial`{.interpreted-text role="func"}. (Contributed by Alexander Belopolsky in `7830`{.interpreted-text role="issue"}.)

### Changes in the C API

- The undocumented `!format`{.interpreted-text role="c:member"} member of the (non-public) `!PyMemoryViewObject`{.interpreted-text role="c:type"} structure has been removed. All extensions relying on the relevant parts in `memoryobject.h` must be rebuilt.
- The `!PyMemAllocator`{.interpreted-text role="c:type"} structure was renamed to `PyMemAllocatorEx`{.interpreted-text role="c:type"} and a new `calloc` field was added.
- Removed non-documented macro `!PyObject_REPR()`{.interpreted-text role="c:macro"} which leaked references. Use format character `%R` in `PyUnicode_FromFormat`{.interpreted-text role="c:func"}-like functions to format the `repr`{.interpreted-text role="func"} of the object. (Contributed by Serhiy Storchaka in `22453`{.interpreted-text role="issue"}.)
- Because the lack of the `~type.__module__`{.interpreted-text role="attr"} attribute breaks pickling and introspection, a deprecation warning is now raised for builtin types without the `~type.__module__`{.interpreted-text role="attr"} attribute. This will be an `AttributeError`{.interpreted-text role="exc"} in the future. (Contributed by Serhiy Storchaka in `20204`{.interpreted-text role="issue"}.)
- As part of the `492`{.interpreted-text role="pep"} implementation, the `tp_reserved` slot of `PyTypeObject`{.interpreted-text role="c:type"} was replaced with a `~PyTypeObject.tp_as_async`{.interpreted-text role="c:member"} slot. Refer to `coro-objects`{.interpreted-text role="ref"} for new types, structures and functions.

## Notable changes in Python 3.5.4

### New `make regen-all` build target

To simplify cross-compilation, and to ensure that CPython can reliably be compiled without requiring an existing version of Python to already be available, the autotools-based build system no longer attempts to implicitly recompile generated files based on file modification times.

Instead, a new `make regen-all` command has been added to force regeneration of these files when desired (e.g. after an initial version of Python has already been built based on the pregenerated versions).

More selective regeneration targets are also defined - see `Makefile.pre.in`{.interpreted-text role="source"} for details.

(Contributed by Victor Stinner in `23404`{.interpreted-text role="issue"}.)

::: versionadded
3.5.4
:::

### Removal of `make touch` build target

The `make touch` build target previously used to request implicit regeneration of generated files by updating their modification times has been removed.

It has been replaced by the new `make regen-all` target.

(Contributed by Victor Stinner in `23404`{.interpreted-text role="issue"}.)

::: versionchanged
3.5.4
:::