path: root/Doc/whatsnew/3.3.rst

****************************
  What's New In Python 3.3
****************************

:Author: Raymond Hettinger
:Release: |release|
:Date: |today|

.. Rules for maintenance:

   * 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 :issue:`12345`.)

   This saves the maintainer the effort of going through the Mercurial log
   when researching a change.

This article explains the new features in Python 3.3, compared to 3.2.


.. _pep-393:

PEP 393: Flexible String Representation
=======================================

The Unicode string type is changed to support multiple internal
representations, depending on the character with the largest Unicode ordinal
(1, 2, or 4 bytes) in the represented string.  This allows a space-efficient
representation in common cases, but gives access to full UCS-4 on all
systems.  For compatibility with existing APIs, several representations may
exist in parallel; over time, this compatibility should be phased out.

On the Python side, there should be no downside to this change.

On the C API side, PEP 393 is fully backward compatible.  The legacy API
should remain available at least five years.  Applications using the legacy
API will not fully benefit of the memory reduction, or - worse - may use
a bit more memory, because Python may have to maintain two versions of each
string (in the legacy format and in the new efficient storage).

Functionality
-------------

Changes introduced by :pep:`393` are the following:

* Python now always supports the full range of Unicode codepoints, including
  non-BMP ones (i.e. from ``U+0000`` to ``U+10FFFF``).  The distinction between
  narrow and wide builds no longer exists and Python now behaves like a wide
  build, even under Windows.

* With the death of narrow builds, the problems specific to narrow builds have
  also been fixed, for example:

  * :func:`len` now always returns 1 for non-BMP characters,
    so ``len('\U0010FFFF') == 1``;

  * surrogate pairs are not recombined in string literals,
    so ``'\uDBFF\uDFFF' != '\U0010FFFF'``;

  * indexing or slicing non-BMP characters returns the expected value,
    so ``'\U0010FFFF'[0]`` now returns ``'\U0010FFFF'`` and not ``'\uDBFF'``;

  * all other functions in the standard library now correctly handle
    non-BMP codepoints.

* The value of :data:`sys.maxunicode` is now always ``1114111`` (``0x10FFFF``
  in hexadecimal).  The :c:func:`PyUnicode_GetMax` function still returns
  either ``0xFFFF`` or ``0x10FFFF`` for backward compatibility, and it should
  not be used with the new Unicode API (see :issue:`13054`).

* The :file:`./configure` flag ``--with-wide-unicode`` has been removed.

Performance and resource usage
------------------------------

The storage of Unicode strings now depends on the highest codepoint in the string:

* pure ASCII and Latin1 strings (``U+0000-U+00FF``) use 1 byte per codepoint;

* BMP strings (``U+0000-U+FFFF``) use 2 bytes per codepoint;

* non-BMP strings (``U+10000-U+10FFFF``) use 4 bytes per codepoint.

The net effect is that for most applications, memory usage of string storage
should decrease significantly - especially compared to former wide unicode
builds - as, in many cases, strings will be pure ASCII even in international
contexts (because many strings store non-human language data, such as XML
fragments, HTTP headers, JSON-encoded data, etc.).  We also hope that it
will, for the same reasons, increase CPU cache efficiency on non-trivial
applications.

.. The memory usage of Python 3.3 is two to three times smaller than Python 3.2,
   and a little bit better than Python 2.7, on a `Django benchmark
   <http://mail.python.org/pipermail/python-dev/2011-September/113714.html>`_.
   XXX The result should be moved in the PEP and a link to the PEP should
   be added here.


PEP 3151: Reworking the OS and IO exception hierarchy
=====================================================

:pep:`3151` - Reworking the OS and IO exception hierarchy
 PEP written and implemented by Antoine Pitrou.

The hierarchy of exceptions raised by operating system errors is now both
simplified and finer-grained.

You don't have to worry anymore about choosing the appropriate exception
type between :exc:`OSError`, :exc:`IOError`, :exc:`EnvironmentError`,
:exc:`WindowsError`, :exc:`mmap.error`, :exc:`socket.error` or
:exc:`select.error`.  All these exception types are now only one:
:exc:`OSError`.  The other names are kept as aliases for compatibility
reasons.

Also, it is now easier to catch a specific error condition.  Instead of
inspecting the ``errno`` attribute (or ``args[0]``) for a particular
constant from the :mod:`errno` module, you can catch the adequate
:exc:`OSError` subclass.  The available subclasses are the following:

* :exc:`BlockingIOError`
* :exc:`ChildProcessError`
* :exc:`ConnectionError`
* :exc:`FileExistsError`
* :exc:`FileNotFoundError`
* :exc:`InterruptedError`
* :exc:`IsADirectoryError`
* :exc:`NotADirectoryError`
* :exc:`PermissionError`
* :exc:`ProcessLookupError`
* :exc:`TimeoutError`

And the :exc:`ConnectionError` itself has finer-grained subclasses:

* :exc:`BrokenPipeError`
* :exc:`ConnectionAbortedError`
* :exc:`ConnectionRefusedError`
* :exc:`ConnectionResetError`

Thanks to the new exceptions, common usages of the :mod:`errno` can now be
avoided.  For example, the following code written for Python 3.2::

    from errno import ENOENT, EACCES, EPERM

    try:
        with open("document.txt") as f:
            content = f.read()
    except IOError as err:
        if err.errno == ENOENT:
            print("document.txt file is missing")
        elif err.errno in (EACCES, EPERM):
            print("You are not allowed to read document.txt")
        else:
            raise

can now be written without the :mod:`errno` import and without manual
inspection of exception attributes::

    try:
        with open("document.txt") as f:
            content = f.read()
    except FileNotFoundError:
        print("document.txt file is missing")
    except PermissionError:
        print("You are not allowed to read document.txt")


PEP 380: Syntax for Delegating to a Subgenerator
================================================

PEP 380 adds the ``yield from`` expression, allowing a generator to delegate
part of its operations to another generator. This allows a section of code
containing 'yield' to be factored out and placed in another generator.
Additionally, the subgenerator is allowed to return with a value, and the
value is made available to the delegating generator.
While designed primarily for use in delegating to a subgenerator, the ``yield
from`` expression actually allows delegation to arbitrary subiterators.

(Implementation by Greg Ewing, integrated into 3.3 by Renaud Blanch, Ryan
Kelly and Nick Coghlan, documentation by Zbigniew Jędrzejewski-Szmek and
Nick Coghlan)


PEP 3155: Qualified name for classes and functions
==================================================

:pep:`3155` - Qualified name for classes and functions
 PEP written and implemented by Antoine Pitrou.

Functions and class objects have a new ``__qualname__`` attribute representing
the "path" from the module top-level to their definition.  For global functions
and classes, this is the same as ``__name__``.  For other functions and classes,
it provides better information about where they were actually defined, and
how they might be accessible from the global scope.

Example with (non-bound) methods::

   >>> class C:
   ...     def meth(self):
   ...         pass
   >>> C.meth.__name__
   'meth'
   >>> C.meth.__qualname__
   'C.meth'

Example with nested classes::

   >>> class C:
   ...     class D:
   ...         def meth(self):
   ...             pass
   ...
   >>> C.D.__name__
   'D'
   >>> C.D.__qualname__
   'C.D'
   >>> C.D.meth.__name__
   'meth'
   >>> C.D.meth.__qualname__
   'C.D.meth'

Example with nested functions::

   >>> def outer():
   ...     def inner():
   ...         pass
   ...     return inner
   ...
   >>> outer().__name__
   'inner'
   >>> outer().__qualname__
   'outer.<locals>.inner'

The string representation of those objects is also changed to include the
new, more precise information::

   >>> str(C.D)
   "<class '__main__.C.D'>"
   >>> str(C.D.meth)
   '<function C.D.meth at 0x7f46b9fe31e0>'


Other Language Changes
======================

Some smaller changes made to the core Python language are:

* Added support for Unicode name aliases and named sequences.
  Both :func:`unicodedata.lookup()` and ``'\N{...}'`` now resolve name aliases,
  and :func:`unicodedata.lookup()` resolves named sequences too.

  (Contributed by Ezio Melotti in :issue:`12753`)

* Equality comparisons on :func:`range` objects now return a result reflecting
  the equality of the underlying sequences generated by those range objects.

  (:issue:`13021`)

* The ``count()``, ``find()``, ``rfind()``, ``index()`` and ``rindex()``
  methods of :class:`bytes` and :class:`bytearray` objects now accept an
  integer between 0 and 255 as their first argument.

  (:issue:`12170`)

* Memoryview objects are now hashable when the underlying object is hashable.

  (Contributed by Antoine Pitrou in :issue:`13411`)


New and Improved Modules
========================

array
-----

The :mod:`array` module supports the :c:type:`long long` type using ``q`` and
``Q`` type codes.

(Contributed by Oren Tirosh and Hirokazu Yamamoto in :issue:`1172711`)


codecs
------

The :mod:`~encodings.mbcs` codec has be rewritten to handle correclty
``replace`` and ``ignore`` error handlers on all Windows versions. The
:mod:`~encodings.mbcs` codec is now supporting all error handlers, instead of
only ``replace`` to encode and ``ignore`` to decode.

A new Windows-only codec has been added: ``cp65001`` (:issue:`13216`). It is
the Windows code page 65001 (Windows UTF-8, ``CP_UTF8``). For example, it is
used by ``sys.stdout`` if the console output code page is set to cp65001 (e.g.
using ``chcp 65001`` command).

Multibyte CJK decoders now resynchronize faster. They only ignore the first
byte of an invalid byte sequence. For example, ``b'\xff\n'.decode('gb2312',
'replace')`` now returns a ``\n`` after the replacement character.

(:issue:`12016`)

Don't reset incremental encoders of CJK codecs at each call to their encode()
method anymore. For example::

    $ ./python -q
    >>> import codecs
    >>> encoder = codecs.getincrementalencoder('hz')('strict')
    >>> b''.join(encoder.encode(x) for x in '\u52ff\u65bd\u65bc\u4eba\u3002 Bye.')
    b'~{NpJ)l6HK!#~} Bye.'

This example gives ``b'~{Np~}~{J)~}~{l6~}~{HK~}~{!#~} Bye.'`` with older Python
versions.

(:issue:`12100`)

The ``unicode_internal`` codec has been deprecated.

crypt
-----

Addition of salt and modular crypt format and the :func:`~crypt.mksalt`
function to the :mod:`crypt` module.

(:issue:`10924`)

curses
------

 * If the :mod:`curses` module is linked to the ncursesw library, use Unicode
   functions when Unicode strings or characters are passed (e.g.
   :c:func:`waddwstr`), and bytes functions otherwise (e.g. :c:func:`waddstr`).
 * Use the locale encoding instead of ``utf-8`` to encode Unicode strings.
 * :class:`curses.window` has a new :attr:`curses.window.encoding` attribute.
 * The :class:`curses.window` class has a new :meth:`~curses.window.get_wch`
   method to get a wide character
 * The :mod:`curses` module has a new :meth:`~curses.unget_wch` function to
   push a wide character so the next :meth:`~curses.window.get_wch` will return
   it

(Contributed by Iñigo Serna in :issue:`6755`)

abc
---

Improved support for abstract base classes containing descriptors composed with
abstract methods. The recommended approach to declaring abstract descriptors is
now to provide :attr:`__isabstractmethod__` as a dynamically updated
property. The built-in descriptors have been updated accordingly.

  * :class:`abc.abstractproperty` has been deprecated, use :class:`property`
    with :func:`abc.abstractmethod` instead.
  * :class:`abc.abstractclassmethod` has been deprecated, use
    :class:`classmethod` with :func:`abc.abstractmethod` instead.
  * :class:`abc.abstractstaticmethod` has been deprecated, use
    :class:`staticmethod` with :func:`abc.abstractmethod` instead.

(Contributed by Darren Dale in :issue:`11610`)

faulthandler
------------

New module: :mod:`faulthandler`.

 * :envvar:`PYTHONFAULTHANDLER`
 * :option:`-X` ``faulthandler``

time
----

* The :mod:`time` module has new :func:`~time.clock_getres` and
  :func:`~time.clock_gettime` functions and ``CLOCK_xxx`` constants.
  :func:`~time.clock_gettime` can be used with :data:`time.CLOCK_MONOTONIC` to
  get a monotonic clock.

  (Contributed by Victor Stinner in :issue:`10278`)


ftplib
------

The :class:`~ftplib.FTP_TLS` class now provides a new
:func:`~ftplib.FTP_TLS.ccc` function to revert control channel back to
plaintext.  This can be useful to take advantage of firewalls that know how to
handle NAT with non-secure FTP without opening fixed ports.

(Contributed by Giampaolo Rodolà in :issue:`12139`)


imaplib
-------

The :class:`~imaplib.IMAP4_SSL` constructor now accepts an SSLContext
parameter to control parameters of the secure channel.

(Contributed by Sijin Joseph in :issue:`8808`)


io
--

The :func:`~io.open` function has a new ``'x'`` mode that can be used to
exclusively create a new file, and raise a :exc:`FileExistsError` if the file
already exists. It is based on the C11 'x' mode to fopen().

(Contributed by David Townshend in :issue:`12760`)


lzma
----

The newly-added :mod:`lzma` module provides data compression and decompression
using the LZMA algorithm, including support for the ``.xz`` and ``.lzma``
file formats.

(Contributed by Nadeem Vawda and Per Øyvind Karlsen in :issue:`6715`)


math
----

The :mod:`math` module has a new function:

  * :func:`~math.log2`: return the base-2 logarithm of *x*
    (Written by Mark Dickinson in :issue:`11888`).


nntplib
-------

The :class:`nntplib.NNTP` class now supports the context manager protocol to
unconditionally consume :exc:`socket.error` exceptions and to close the NNTP
connection when done::

  >>> from nntplib import NNTP
  >>> with NNTP('news.gmane.org') as n:
  ...     n.group('gmane.comp.python.committers')
  ...
  ('211 1755 1 1755 gmane.comp.python.committers', 1755, 1, 1755, 'gmane.comp.python.committers')
  >>>

(Contributed by Giampaolo Rodolà in :issue:`9795`)


os
--

* The :mod:`os` module has a new :func:`~os.pipe2` function that makes it
  possible to create a pipe with :data:`~os.O_CLOEXEC` or
  :data:`~os.O_NONBLOCK` flags set atomically. This is especially useful to
  avoid race conditions in multi-threaded programs.

* The :mod:`os` module has a new :func:`~os.sendfile` function which provides
  an efficent "zero-copy" way for copying data from one file (or socket)
  descriptor to another. The phrase "zero-copy" refers to the fact that all of
  the copying of data between the two descriptors is done entirely by the
  kernel, with no copying of data into userspace buffers. :func:`~os.sendfile`
  can be used to efficiently copy data from a file on disk to a network socket,
  e.g. for downloading a file.

  (Patch submitted by Ross Lagerwall and Giampaolo Rodolà in :issue:`10882`.)

* The :mod:`os` module has two new functions: :func:`~os.getpriority` and
  :func:`~os.setpriority`. They can be used to get or set process
  niceness/priority in a fashion similar to :func:`os.nice` but extended to all
  processes instead of just the current one.

  (Patch submitted by Giampaolo Rodolà in :issue:`10784`.)

* "at" functions (:issue:`4761`):

  * :func:`~os.faccessat`
  * :func:`~os.fchmodat`
  * :func:`~os.fchownat`
  * :func:`~os.fstatat`
  * :func:`~os.futimesat`
  * :func:`~os.futimesat`
  * :func:`~os.linkat`
  * :func:`~os.mkdirat`
  * :func:`~os.mkfifoat`
  * :func:`~os.mknodat`
  * :func:`~os.openat`
  * :func:`~os.readlinkat`
  * :func:`~os.renameat`
  * :func:`~os.symlinkat`
  * :func:`~os.unlinkat`
  * :func:`~os.utimensat`
  * :func:`~os.utimensat`

* extended attributes (:issue:`12720`):

  * :func:`~os.fgetxattr`
  * :func:`~os.flistxattr`
  * :func:`~os.fremovexattr`
  * :func:`~os.fsetxattr`
  * :func:`~os.getxattr`
  * :func:`~os.lgetxattr`
  * :func:`~os.listxattr`
  * :func:`~os.llistxattr`
  * :func:`~os.lremovexattr`
  * :func:`~os.lsetxattr`
  * :func:`~os.removexattr`
  * :func:`~os.setxattr`

* Scheduler functions (:issue:`12655`):

  * :func:`~os.sched_get_priority_max`
  * :func:`~os.sched_get_priority_min`
  * :func:`~os.sched_getaffinity`
  * :func:`~os.sched_getparam`
  * :func:`~os.sched_getscheduler`
  * :func:`~os.sched_rr_get_interval`
  * :func:`~os.sched_setaffinity`
  * :func:`~os.sched_setparam`
  * :func:`~os.sched_setscheduler`
  * :func:`~os.sched_yield`

* Add some extra posix functions to the os module (:issue:`10812`):

  * :func:`~os.fexecve`
  * :func:`~os.futimens`
  * :func:`~os.futimens`
  * :func:`~os.futimes`
  * :func:`~os.futimes`
  * :func:`~os.lockf`
  * :func:`~os.lutimes`
  * :func:`~os.lutimes`
  * :func:`~os.posix_fadvise`
  * :func:`~os.posix_fallocate`
  * :func:`~os.pread`
  * :func:`~os.pwrite`
  * :func:`~os.readv`
  * :func:`~os.sync`
  * :func:`~os.truncate`
  * :func:`~os.waitid`
  * :func:`~os.writev`

* Other new functions:

  * :func:`~os.fdlistdir` (:issue:`10755`)
  * :func:`~os.getgrouplist` (:issue:`9344`)


packaging
---------

:mod:`distutils` has undergone additions and refactoring under a new name,
:mod:`packaging`, to allow developers to break backward compatibility.
:mod:`distutils` is still provided in the standard library, but users are
encouraged to transition to :mod:`packaging`.  For older versions of Python, a
backport compatible with 2.4+ and 3.1+ will be made available on PyPI under the
name :mod:`distutils2`.

.. TODO add examples and howto to the packaging docs and link to them


pydoc
-----

The Tk GUI and the :func:`~pydoc.serve` function have been removed from the
:mod:`pydoc` module: ``pydoc -g`` and :func:`~pydoc.serve` have been deprecated
in Python 3.2.


sys
---

* The :mod:`sys` module has a new :data:`~sys.thread_info` :term:`struct
  sequence` holding informations about the thread implementation.

  (:issue:`11223`)


signal
------

* The :mod:`signal` module has new functions:

  * :func:`~signal.pthread_sigmask`: fetch and/or change the signal mask of the
    calling thread (Contributed by Jean-Paul Calderone in :issue:`8407`) ;
  * :func:`~signal.pthread_kill`: send a signal to a thread ;
  * :func:`~signal.sigpending`: examine pending functions ;
  * :func:`~signal.sigwait`: wait a signal.
  * :func:`~signal.sigwaitinfo`: wait for a signal, returning detailed
    information about it.
  * :func:`~signal.sigtimedwait`: like :func:`~signal.sigwaitinfo` but with a
    timeout.

* The signal handler writes the signal number as a single byte instead of
  a nul byte into the wakeup file descriptor. So it is possible to wait more
  than one signal and know which signals were raised.

* :func:`signal.signal` and :func:`signal.siginterrupt` raise an OSError,
  instead of a RuntimeError: OSError has an errno attribute.

socket
------

* The :class:`~socket.socket` class now exposes additional methods to process
  ancillary data when supported by the underlying platform:

  * :func:`~socket.socket.sendmsg`
  * :func:`~socket.socket.recvmsg`
  * :func:`~socket.socket.recvmsg_into`

  (Contributed by David Watson in :issue:`6560`, based on an earlier patch by
  Heiko Wundram)

* The :class:`~socket.socket` class now supports the PF_CAN protocol family
  (http://en.wikipedia.org/wiki/Socketcan), on Linux
  (http://lwn.net/Articles/253425).

  (Contributed by Matthias Fuchs, updated by Tiago Gonçalves in :issue:`10141`)

* The :class:`~socket.socket` class now supports the PF_RDS protocol family
  (http://en.wikipedia.org/wiki/Reliable_Datagram_Sockets and
  http://oss.oracle.com/projects/rds/).

ssl
---

* The :mod:`ssl` module has two new random generation functions:

  * :func:`~ssl.RAND_bytes`: generate cryptographically strong
    pseudo-random bytes.
  * :func:`~ssl.RAND_pseudo_bytes`: generate pseudo-random bytes.

  (Contributed by Victor Stinner in :issue:`12049`)

* The :mod:`ssl` module now exposes a finer-grained exception hierarchy
  in order to make it easier to inspect the various kinds of errors.

  (Contributed by Antoine Pitrou in :issue:`11183`)

* :meth:`~ssl.SSLContext.load_cert_chain` now accepts a *password* argument
  to be used if the private key is encrypted.

  (Contributed by Adam Simpkins in :issue:`12803`)

* Diffie-Hellman key exchange, both regular and Elliptic Curve-based, is
  now supported through the :meth:`~ssl.SSLContext.load_dh_params` and
  :meth:`~ssl.SSLContext.set_ecdh_curve` methods.

  (Contributed by Antoine Pitrou in :issue:`13626` and :issue:`13627`)

* SSL sockets have a new :meth:`~ssl.SSLSocket.get_channel_binding` method
  allowing the implementation of certain authentication mechanisms such as
  SCRAM-SHA-1-PLUS.

  (Contributed by Jacek Konieczny in :issue:`12551`)

* You can query the SSL compression algorithm used by an SSL socket, thanks
  to its new :meth:`~ssl.SSLSocket.compression` method.

  (Contributed by Antoine Pitrou in :issue:`13634`)


shutil
------

* The :mod:`shutil` module has these new fuctions:

  * :func:`~shutil.disk_usage`: provides total, used and free disk space
    statistics. (Contributed by Giampaolo Rodolà in :issue:`12442`)
  * :func:`~shutil.chown`: allows one to change user and/or group of the given
    path also specifying the user/group names and not only their numeric
    ids. (Contributed by Sandro Tosi in :issue:`12191`)

smtplib
-------

The :class:`~smtplib.SMTP_SSL` constructor and the :meth:`~smtplib.SMTP.starttls`
method now accept an SSLContext parameter to control parameters of the secure
channel.

(Contributed by Kasun Herath in :issue:`8809`)

urllib
------

The :class:`~urllib.request.Request` class, now accepts a *method* argument
used by :meth:`~urllib.request.Request.get_method` to determine what HTTP method
should be used.  For example, this will send a ``'HEAD'`` request::

   >>> urlopen(Request('http://www.python.org', method='HEAD'))

(:issue:`1673007`)

sched
-----

* :meth:`~sched.scheduler.run` now accepts a *blocking* parameter which when
  set to False makes the method execute the scheduled events due to expire
  soonest (if any) and then return immediately.
  This is useful in case you want to use the :class:`~sched.scheduler` in
  non-blocking applications.  (Contributed by Giampaolo Rodolà in :issue:`13449`)

* :class:`~sched.scheduler` class can now be safely used in multi-threaded
  environments.  (Contributed by Josiah Carlson and Giampaolo Rodolà in
  :issue:`8684`)

* *timefunc* and *delayfunct* parameters of :class:`~sched.scheduler` class
  constructor are now optional and defaults to :func:`time.time` and
  :func:`time.sleep` respectively.  (Contributed by Chris Clark in
  :issue:`13245`)

* :meth:`~sched.scheduler.enter` and :meth:`~sched.scheduler.enterabs`
  *argument* parameter is now optional.  (Contributed by Chris Clark in
  :issue:`13245`)

* :meth:`~sched.scheduler.enter` and :meth:`~sched.scheduler.enterabs`
  now accept a *kwargs* parameter.  (Contributed by Chris Clark in
  :issue:`13245`)

Optimizations
=============

Major performance enhancements have been added:

* Thanks to the :pep:`393`, some operations on Unicode strings has been optimized:

  * the memory footprint is divided by 2 to 4 depending on the text
  * encode an ASCII string to UTF-8 doesn't need to encode characters anymore,
    the UTF-8 representation is shared with the ASCII representation
  * the UTF-8 encoder has been optimized
  * repeating a single ASCII letter and getting a substring of a ASCII strings
    is 4 times faster


Build and C API Changes
=======================

Changes to Python's build process and to the C API include:

* The :pep:`393` added new Unicode types, macros and functions:

  * High-level API:

    * :c:func:`PyUnicode_CopyCharacters`
    * :c:func:`PyUnicode_FindChar`
    * :c:func:`PyUnicode_GetLength`, :c:macro:`PyUnicode_GET_LENGTH`
    * :c:func:`PyUnicode_New`
    * :c:func:`PyUnicode_Substring`
    * :c:func:`PyUnicode_ReadChar`, :c:func:`PyUnicode_WriteChar`

  * Low-level API:

    * :c:type:`Py_UCS1`, :c:type:`Py_UCS2`, :c:type:`Py_UCS4` types
    * :c:type:`PyASCIIObject` and :c:type:`PyCompactUnicodeObject` structures
    * :c:macro:`PyUnicode_READY`
    * :c:func:`PyUnicode_FromKindAndData`
    * :c:func:`PyUnicode_AsUCS4`, :c:func:`PyUnicode_AsUCS4Copy`
    * :c:macro:`PyUnicode_DATA`, :c:macro:`PyUnicode_1BYTE_DATA`,
      :c:macro:`PyUnicode_2BYTE_DATA`, :c:macro:`PyUnicode_4BYTE_DATA`
    * :c:macro:`PyUnicode_KIND` with :c:type:`PyUnicode_Kind` enum:
      :c:data:`PyUnicode_WCHAR_KIND`, :c:data:`PyUnicode_1BYTE_KIND`,
      :c:data:`PyUnicode_2BYTE_KIND`, :c:data:`PyUnicode_4BYTE_KIND`
    * :c:macro:`PyUnicode_READ`, :c:macro:`PyUnicode_READ_CHAR`, :c:macro:`PyUnicode_WRITE`
    * :c:macro:`PyUnicode_MAX_CHAR_VALUE`



Deprecated
==========

Unsupported Operating Systems
-----------------------------

OS/2 and VMS are no longer supported due to the lack of a maintainer.

Windows 2000 and Windows platforms which set ``COMSPEC`` to ``command.com``
are no longer supported due to maintenance burden.


Deprecated Python modules, functions and methods
------------------------------------------------

* The :mod:`packaging` module replaces the :mod:`distutils` module
* The ``unicode_internal`` codec has been deprecated because of the
  :pep:`393`, use UTF-8, UTF-16 (``utf-16-le`` or ``utf-16-le``), or UTF-32
  (``utf-32-le`` or ``utf-32-le``)
* :meth:`ftplib.FTP.nlst` and :meth:`ftplib.FTP.dir`: use
  :meth:`ftplib.FTP.mlsd`
* :func:`platform.popen`: use the :mod:`subprocess` module. Check especially
  the :ref:`subprocess-replacements` section.
* :issue:`13374`: The Windows bytes API has been deprecated in the :mod:`os`
  module. Use Unicode filenames, instead of bytes filenames, to not depend on
  the ANSI code page anymore and to support any filename.


Deprecated functions and types of the C API
-------------------------------------------

The :c:type:`Py_UNICODE` has been deprecated by the :pep:`393` and will be
removed in Python 4. All functions using this type are deprecated:

Unicode functions and methods using :c:type:`Py_UNICODE` and
:c:type:`Py_UNICODE*` types:

 * :c:macro:`PyUnicode_FromUnicode`: use :c:func:`PyUnicode_FromWideChar` or
   :c:func:`PyUnicode_FromKindAndData`
 * :c:macro:`PyUnicode_AS_UNICODE`, :c:func:`PyUnicode_AsUnicode`,
   :c:func:`PyUnicode_AsUnicodeAndSize`: use :c:func:`PyUnicode_AsWideCharString`
 * :c:macro:`PyUnicode_AS_DATA`: use :c:macro:`PyUnicode_DATA` with
   :c:macro:`PyUnicode_READ` and :c:macro:`PyUnicode_WRITE`
 * :c:macro:`PyUnicode_GET_SIZE`, :c:func:`PyUnicode_GetSize`: use
   :c:macro:`PyUnicode_GET_LENGTH` or :c:func:`PyUnicode_GetLength`
 * :c:macro:`PyUnicode_GET_DATA_SIZE`: use
   ``PyUnicode_GET_LENGTH(str) * PyUnicode_KIND(str)`` (only work on ready
   strings)
 * :c:func:`PyUnicode_AsUnicodeCopy`: use :c:func:`PyUnicode_AsUCS4Copy` or
   :c:func:`PyUnicode_AsWideCharString`
 * :c:func:`PyUnicode_GetMax`


Functions and macros manipulating Py_UNICODE* strings:

 * :c:macro:`Py_UNICODE_strlen`: use :c:func:`PyUnicode_GetLength` or
   :c:macro:`PyUnicode_GET_LENGTH`
 * :c:macro:`Py_UNICODE_strcat`: use :c:func:`PyUnicode_CopyCharacters` or
   :c:func:`PyUnicode_FromFormat`
 * :c:macro:`Py_UNICODE_strcpy`, :c:macro:`Py_UNICODE_strncpy`,
   :c:macro:`Py_UNICODE_COPY`: use :c:func:`PyUnicode_CopyCharacters` or
   :c:func:`PyUnicode_Substring`
 * :c:macro:`Py_UNICODE_strcmp`: use :c:func:`PyUnicode_Compare`
 * :c:macro:`Py_UNICODE_strncmp`: use :c:func:`PyUnicode_Tailmatch`
 * :c:macro:`Py_UNICODE_strchr`, :c:macro:`Py_UNICODE_strrchr`: use
   :c:func:`PyUnicode_FindChar`
 * :c:macro:`Py_UNICODE_FILL`: use :c:func:`PyUnicode_Fill`
 * :c:macro:`Py_UNICODE_MATCH`

Encoders:

 * :c:func:`PyUnicode_Encode`: use :c:func:`PyUnicode_AsEncodedObject`
 * :c:func:`PyUnicode_EncodeUTF7`
 * :c:func:`PyUnicode_EncodeUTF8`: use :c:func:`PyUnicode_AsUTF8` or
   :c:func:`PyUnicode_AsUTF8String`
 * :c:func:`PyUnicode_EncodeUTF32`
 * :c:func:`PyUnicode_EncodeUTF16`
 * :c:func:`PyUnicode_EncodeUnicodeEscape:` use
   :c:func:`PyUnicode_AsUnicodeEscapeString`
 * :c:func:`PyUnicode_EncodeRawUnicodeEscape:` use
   :c:func:`PyUnicode_AsRawUnicodeEscapeString`
 * :c:func:`PyUnicode_EncodeLatin1`: use :c:func:`PyUnicode_AsLatin1String`
 * :c:func:`PyUnicode_EncodeASCII`: use :c:func:`PyUnicode_AsASCIIString`
 * :c:func:`PyUnicode_EncodeCharmap`
 * :c:func:`PyUnicode_TranslateCharmap`
 * :c:func:`PyUnicode_EncodeMBCS`: use :c:func:`PyUnicode_AsMBCSString` or
   :c:func:`PyUnicode_EncodeCodePage` (with ``CP_ACP`` code_page)
 * :c:func:`PyUnicode_EncodeDecimal`,
   :c:func:`PyUnicode_TransformDecimalToASCII`


Porting to Python 3.3
=====================

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

Porting Python code
-------------------

* :issue:`12326`: On Linux, sys.platform doesn't contain the major version
  anymore. It is now always 'linux', instead of 'linux2' or 'linux3' depending
  on the Linux version used to build Python. Replace sys.platform == 'linux2'
  with sys.platform.startswith('linux'), or directly sys.platform == 'linux' if
  you don't need to support older Python versions.

Porting C code
--------------

* Due to :ref:`PEP 393 <pep-393>`, the :c:type:`Py_UNICODE` type and all
  functions using this type are deprecated (but will stay available for
  at least five years).  If you were using low-level Unicode APIs to
  construct and access unicode objects and you want to benefit of the
  memory footprint reduction provided by the PEP 393, you have to convert
  your code to the new :doc:`Unicode API <../c-api/unicode>`.

  However, if you only have been using high-level functions such as
  :c:func:`PyUnicode_Concat()`, :c:func:`PyUnicode_Join` or
  :c:func:`PyUnicode_FromFormat()`, your code will automatically take
  advantage of the new unicode representations.

Other issues
------------

.. Issue #11591: When :program:`python` was started with :option:`-S`,
   ``import site`` will not add site-specific paths to the module search
   paths.  In previous versions, it did.  See changeset for doc changes in
   various files.  Contributed by Carl Meyer with editions by Éric Araujo.

.. Issue #10998: the -Q command-line flag and related artifacts have been
   removed.  Code checking sys.flags.division_warning will need updating.
   Contributed by Éric Araujo.