Python 3.3 有什么新变化

本文介绍了 Python 3.3 相比 3.2 的新增特性。 Python 3.3 于 2012 年 9 月 29 日 发布。 有关完整详细信息，请参见 changelog。

参见

PEP 398 - Python 3.3 发布计划

摘要 — 发布重点

新的语法特性：

• 新增 yield from 表达式用于 生成器委托。

• u'unicode' 语法重新被接受用于 str 对象。

新的库模块：

• faulthandler (帮助调试低层级的崩溃)

• ipaddress (代表 IP 地址和掩码的高层级对象)

• lzma (使用 XZ / LZMA 算法压缩数据)

• unittest.mock (使用模拟对象替换你的受测试系统中的某些部分)

• venv (Python 虚拟环境，类似于流行的 virtualenv 包)

新的内置特性：

• 重写 I/O 异常的层次结构.

实现的改进：

• 基于 importlib 重写 import machinery

• 更紧凑的 Unicode 字符串。

• 更紧凑的 属性字典。

显著改进的库模块：

• 针对 decimal 模块的 C 加速器。

• email 模块中更好的 Unicode 处理 (暂定)。

安全改进：

• 哈希随机化被默认启用。

请继续阅读有关面向用户的改变的详细清单。

PEP 405: 虚拟环境

虚拟环境有助于创建 独立的 Python 设置，同时共享全系统的基础安装，便于维护。 虚拟环境有自己的私有站点包（即本地安装的库），并可选择与系统范围的站点包分离。 虚拟环境的概念和实现 受到流行的``virtualenv`` 第三方 包 的启发，但受益于与解释器 核心更紧密的集成。

本 PEP 添加了 venv 模块用于编程访问，以及 pyvenv 脚本用于命令在线访问和管理。 Python 解释器会检查 pyvenv.cfg，文件的存在标志着虚拟环境目录树的基础。

参见

PEP 405 - Python虚拟环境

PEP 由 Carl Meyer 撰写 ; 由 Carl Meyer 和 Vinay Sajip 实现。

PEP 420: 隐式命名空间包

原生支持不要求 __init__.py 标记文件和可以自动跨越多个路径节的包目录（灵感来自多个命名空间包的第三方方案，如 PEP 420 中所述）

参见

PEP 420 - 隐式命名空间包

PEP 由 Eric V. Smith 撰写，由 Eric V. Smith 和 Barry Warsaw 实现

PEP 3118: 新的内存视图实现和缓冲协议文档

PEP 3118 的实现已获得大幅改进。

新的 memoryview 实现全面修复了 Py_buffer 结构体中曾导致多起崩溃报告的动态分配字段的所有权和生命周期问题。 此外，还修复了多个函数在非连续或多维输入时崩溃或返回不正确结果的问题。

现在 memoryview 对象具有符合 PEP-3118 标准的 getbufferproc()，可以检查使用者的请求类型。 新增了许多新的特性，其中的大部分已适用于非连续数组和带有子偏移量的数组。

文档已进行更新，清楚地列出了导出方和使用方的责任。 缓冲区请求旗标志被划分为基本旗标和复合旗标。 对非连续和多维的 NumPy 风格数组的内存布局进行了说明。

相关特性

• 现在 struct 模块语法中所有原生单字符格式指示符（可以选择添加 ‘@’ 前缀）均受到支持。

• 在某些限制条件下，cast() 方法允许改变 C 连续数组的格式和形状。

• 任何数组类型都支持多维列表的表示形式。

• 任何数组类型都支持多维比较操作。

• 格式为 B、b 或 c 的可哈希（只读）类型的一维 memoryview 现在将是可哈希的。 （由 Antoine Pitrou 在 bpo-13411 中贡献。）

• 支持对 1 维数据类型的任意切片。 例如，现在可以通过使用负步长值以 O(1) 复杂度对 memoryview 进行翻转。

API 的变化

• 官方的最大维度数量限制已更改为 64。

• 空形状、区间和子偏移量的表示形式现在是空元组而不是 None。

• 现在对格式为 ‘B’ (无符号字节型) 的 memoryview 元素的访问将返回一个整数（遵循结构体模块语法）。 要返回字节串对象则必须先将视图强制转换为 ‘c’。

• 现在 memoryview 比较将使用操作数的逻辑结构并会按值来比较所有数组元素。 结构体模块语法中的所有格式化字符串均受到支持。 带有不可识别的格式化字符串的视图仍然被允许，但无论视图内容如何比较结果总是不相等。

• 更多改变请参阅 Build and C API Changes 和 Porting C code。

（由 Stefan Krah 在 bpo-10181 中贡献。）

参见

PEP 3118 - 修改缓冲区协议

PEP 393: 灵活的字符串表示

Unicode字符串类型已改为支持多种内部表示法，具体取决于所表示的字符串中具有最大 Unicode 序号（1、2 或 4 字节）的字符 。 这样，在常见情况下可以节省空间，但在所有系统上都能使用完整的 UCS-4。 对于使用现有应用程序接口的兼容性 来说，可能会并行存在几种表示法；随着时间的推移，这种兼容性 应逐步淘汰。

在 Python 一方，此项改变应当没有任何缺点。

在 C API 方面，PEP 393 完全向下兼容。 旧的 API 至少还能使用五年。 使用传统 API 的应用程序不会完全受益于内存的减少，或者更糟的是，可能会使用更多的内存，因为 Python 可能需要维护每个字符串的两个版本（传统格式和新的高效存储）。

功能

由 PEP 393 引入的改变如下:

• Python 现在始终支持全部 Unicode 码位，包括非 BMP 码位 (即从``U+0000`` 到 U+10FFFF)。 窄编译版本和宽编译版本之间的区别已不复存在，Python 现在的行为就像宽编译版本，甚至在 Windows 下也是如此。

• 随着窄编译版本的消亡，窄编译版本特有的问题也得到了解决，例如：

  • 现在 len() 对于非 BMP 字符总是返回 1，因此 len('\U0010FFFF') == 1；

  • 替换对不会在字符串字面值中重新合并，因此 '\uDBFF\uDFFF' != '\U0010FFFF'；

  • 索引或切分非 BMP 字符会返回预期的值，因此 '\U0010FFFF'[0] 现在会返回 '\U0010FFFF' 而不是 '\uDBFF'；

  • 标准库中的所有其他函数现在都能正确处理非 BMP 代码点。

• sys.maxunicode 的值现在总是 1114111 (十六进制为 0x10FFFF)。 PyUnicode_GetMax() 函数仍返回 0xFFFF 或 0x10FFFF 以便向下兼容，但不应与新的 Unicode API 一起使用 (参见 bpo-13054)。

• ./configure 标志 --with-wide-unicode 已被移除。

性能和资源使用情况

现在，Unicode 字符串的存储取决于字符串中的最高码位：

• 纯 ASCII 和 Latin1 字符串 (U+0000-U+00FF) 每个码位使用 1 个字节；

• BMP 字符串 (U+0000-U+FFFF) 每个码位使用 2 个字节；

• 非 BMP 字符串 (U+10000-U+10FFFF) 每个码位使用 4 个字节。

这样做的效果是，对于大多数应用而言，字符串存储的内存使用量应该会大幅减少 —— 尤其是与以前的宽 unicode 版本相比 —— 因为在许多情况下，即使在国际环境中，字符串也将是纯 ASCII 格式（因为许多字符串存储的是非人类语言数据，如 XML 片段、HTTP 标头、JSON 编码数据等）。 出于同样的原因，我们还希望它能提高非小应用程序的 CPU 缓存效率。 在 Django 基准测试中，Python 3.3 的内存使用量比 Python 3.2 少两到三倍，比 Python 2.7 略好一些（详情请参见 PEP）。

参见

PEP 393 - 灵活的字符串表示

PEP 由 Martin von Löwis 撰写 ; 由 Torsten Becker 和 Martin von Löwis 实现。

PEP 397: 适用于Windows的Python启动器

Python 3.3 的 Windows 安装程序现在包含一个 py 启动程序，可用于以版本无关的方式启动 Python 应用程序。

双击 *.py 文件时会隐式调用该启动器。 如果系统中只安装了一个 Python 版本，则将使用该版本运行文件。 如果安装了多个版本，则默认使用最新版本，但也可以通过在 Python 脚本中加入 Unix 风格的“shebang 行”来覆盖该版本。

启动器也可以作为 py 应用程序在命令行中显式使用。运行 py 遵循与隐式启动脚本相同的版本选择规则，但可以通过传递适当的参数来选择更具体的版本（例如，当 Python 2 也已安装时，使用 -3 来请求 Python 3；当安装了较新的 Python 版本时，使用 -2.6 来特别请求较早的 Python 版本）。

除了启动器之外，Windows 安装程序现在还包含一个选项，可将新安装的 Python 添加到系统 PATH 中。 （由 Brian Curtin 在 bpo-3561 中贡献）。

参见

PEP 397 - 适用于Windows的Python启动器

PEP 由 Mark Hammond 和 Martin v. Löwis 撰写 ; 由 Vinay Sajip实现。

启动器文档: 适用于Windows的Python启动器

安装器 PATH 修改: 查找Python可执行文件

PEP 3151: 重写 OS 和 IO 异常的层次结构

现在，由操作系统错误引发的异常层次结构既得到了简化，又更加精细。

您不必再为在 OSError、IOError、EnvironmentError、WindowsError、mmap.error、socket.error 或 select.error 之间选择合适的异常类型而烦恼。 所有这些异常类型现在都只有一个: OSError。 出于兼容性考虑，其他名称将作为别名保留。

此外，现在捕捉特定错误条件也更容易了。无需从 errno 模块中检查 errno 属性（或 args[0] ）中的特定常量，您可以捕捉适当的 OSError 子类。可用的子类如下：

• BlockingIOError

• ChildProcessError

• ConnectionError

• FileExistsError

• FileNotFoundError

• InterruptedError

• IsADirectoryError

• NotADirectoryError

• PermissionError

• ProcessLookupError

• TimeoutError

并且 ConnectionError 本身具有细粒度的子类:

• BrokenPipeError

• ConnectionAbortedError

• ConnectionRefusedError

• ConnectionResetError

有了新的异常，现在就可以避免 errno 的常见用法了。 例如，下面是为 Python 3.2 编写的代码：

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

现在无需导入 errno，也无需手动检查异常属性：

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")

参见

PEP 3151 - 重写 OS 和 IO 异常的层次结构

PEP 由 Antoine Pitrou 撰写并实现

PEP 380: 委托给子生成器的语法

PEP 380 增加了 yield from 表达式，允许 generator 将其部分操作委托给另一个生成器。 这样，包含 yield 的代码部分就可以被分解出来，放在另一个生成器中。 此外，还允许子生成器返回一个值，并将该值提供给委托生成器。

虽然 yield from 表达式主要用于委托给子生成器，但它实际上允许委托给任意子生成器。

对于简单的迭代器而言，yield from iterable 本质上只是 for item in iterable: yield item 的简写形式：

>>> def g(x):
...     yield from range(x, 0, -1)
...     yield from range(x)
...
>>> list(g(5))
[5, 4, 3, 2, 1, 0, 1, 2, 3, 4]

However, unlike an ordinary loop, yield from allows subgenerators to receive sent and thrown values directly from the calling scope, and return a final value to the outer generator:

>>> def accumulate():
...     tally = 0
...     while 1:
...         next = yield
...         if next is None:
...             return tally
...         tally += next
...
>>> def gather_tallies(tallies):
...     while 1:
...         tally = yield from accumulate()
...         tallies.append(tally)
...
>>> tallies = []
>>> acc = gather_tallies(tallies)
>>> next(acc)  # Ensure the accumulator is ready to accept values
>>> for i in range(4):
...     acc.send(i)
...
>>> acc.send(None)  # Finish the first tally
>>> for i in range(5):
...     acc.send(i)
...
>>> acc.send(None)  # Finish the second tally
>>> tallies
[6, 10]

The main principle driving this change is to allow even generators that are designed to be used with the send and throw methods to be split into multiple subgenerators as easily as a single large function can be split into multiple subfunctions.

参见

PEP 380 - 委托给子生成器的语法

PEP 由 Greg Ewing 撰写，由 Greg Ewing 实现。由 Renaud Blanch，Ryan Kelly 和 Nick Coghlan 集成到3.3，由 Zbigniew Jędrzejewski-Szmek 和 Nick Coghlan 编写文档

PEP 409: 清除异常上下文

PEP 409 introduces new syntax that allows the display of the chained exception context to be disabled. This allows cleaner error messages in applications that convert between exception types:

>>> class D:
...     def __init__(self, extra):
...         self._extra_attributes = extra
...     def __getattr__(self, attr):
...         try:
...             return self._extra_attributes[attr]
...         except KeyError:
...             raise AttributeError(attr) from None
...
>>> D({}).x
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "<stdin>", line 8, in __getattr__
AttributeError: x

Without the from None suffix to suppress the cause, the original exception would be displayed by default:

>>> class C:
...     def __init__(self, extra):
...         self._extra_attributes = extra
...     def __getattr__(self, attr):
...         try:
...             return self._extra_attributes[attr]
...         except KeyError:
...             raise AttributeError(attr)
...
>>> C({}).x
Traceback (most recent call last):
  File "<stdin>", line 6, in __getattr__
KeyError: 'x'
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "<stdin>", line 8, in __getattr__
AttributeError: x

No debugging capability is lost, as the original exception context remains available if needed (for example, if an intervening library has incorrectly suppressed valuable underlying details):

>>> try:
...     D({}).x
... except AttributeError as exc:
...     print(repr(exc.__context__))
...
KeyError('x',)

参见

PEP 409 - 清除异常上下文

PEP 由 Ethan Furman 撰写 ，由 Ethan Furman 和 Nick Coghlan 实现。

PEP 414: 显式的Unicode文本

To ease the transition from Python 2 for Unicode aware Python applications that make heavy use of Unicode literals, Python 3.3 once again supports the “u“ prefix for string literals. This prefix has no semantic significance in Python 3, it is provided solely to reduce the number of purely mechanical changes in migrating to Python 3, making it easier for developers to focus on the more significant semantic changes (such as the stricter default separation of binary and text data).

参见

PEP 414 - 显式的Unicode文本

PEP 由 Armin Ronacher 撰写

PEP 3155: 类和函数的限定名称

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>'

参见

PEP 3155 - 类和函数的限定名称

PEP 由 Antoine Pitrou 撰写并实现

PEP 412: Key-Sharing Dictionary

Dictionaries used for the storage of objects’ attributes are now able to share part of their internal storage between each other (namely, the part which stores the keys and their respective hashes). This reduces the memory consumption of programs creating many instances of non-builtin types.

参见

PEP 412 - Key-Sharing Dictionary

PEP 由 Mark Shannon 撰写并实现。

PEP 362: 函数签名对象

A new function inspect.signature() makes introspection of python callables easy and straightforward. A broad range of callables is supported: python functions, decorated or not, classes, and functools.partial() objects. New classes inspect.Signature, inspect.Parameter and inspect.BoundArguments hold information about the call signatures, such as, annotations, default values, parameters kinds, and bound arguments, which considerably simplifies writing decorators and any code that validates or amends calling signatures or arguments.

参见

PEP 362: - 函数签名对象

PEP 由 Brett Cannon，Yury Selivanov，Larry Hastings，Jiwon Seo 撰写，由 Yury Selivanov 实现

PEP 421: 添加 sys.implementation

A new attribute on the sys module exposes details specific to the implementation of the currently running interpreter. The initial set of attributes on sys.implementation are name, version, hexversion, and cache_tag.

The intention of sys.implementation is to consolidate into one namespace the implementation-specific data used by the standard library. This allows different Python implementations to share a single standard library code base much more easily. In its initial state, sys.implementation holds only a small portion of the implementation-specific data. Over time that ratio will shift in order to make the standard library more portable.

One example of improved standard library portability is cache_tag. As of Python 3.3, sys.implementation.cache_tag is used by importlib to support PEP 3147 compliance. Any Python implementation that uses importlib for its built-in import system may use cache_tag to control the caching behavior for modules.

SimpleNamespace

The implementation of sys.implementation also introduces a new type to Python: types.SimpleNamespace. In contrast to a mapping-based namespace, like dict, SimpleNamespace is attribute-based, like object. However, unlike object, SimpleNamespace instances are writable. This means that you can add, remove, and modify the namespace through normal attribute access.

参见

PEP 421 - 添加 sys.implementation

PEP 由 Eric Snow 撰写并实现

使用 importlib 作为导入的实现

bpo-2377 - Replace __import__ w/ importlib.__import__ bpo-13959 - Re-implement parts of imp in pure Python bpo-14605 - Make import machinery explicit bpo-14646 - Require loaders set __loader__ and __package__

The __import__() function is now powered by importlib.__import__(). This work leads to the completion of “phase 2” of PEP 302. There are multiple benefits to this change. First, it has allowed for more of the machinery powering import to be exposed instead of being implicit and hidden within the C code. It also provides a single implementation for all Python VMs supporting Python 3.3 to use, helping to end any VM-specific deviations in import semantics. And finally it eases the maintenance of import, allowing for future growth to occur.

For the common user, there should be no visible change in semantics. For those whose code currently manipulates import or calls import programmatically, the code changes that might possibly be required are covered in the Porting Python code section of this document.

新的API

One of the large benefits of this work is the exposure of what goes into making the import statement work. That means the various importers that were once implicit are now fully exposed as part of the importlib package.

The abstract base classes defined in importlib.abc have been expanded to properly delineate between meta path finders and path entry finders by introducing importlib.abc.MetaPathFinder and importlib.abc.PathEntryFinder, respectively. The old ABC of importlib.abc.Finder is now only provided for backwards-compatibility and does not enforce any method requirements.

In terms of finders, importlib.machinery.FileFinder exposes the mechanism used to search for source and bytecode files of a module. Previously this class was an implicit member of sys.path_hooks.

For loaders, the new abstract base class importlib.abc.FileLoader helps write a loader that uses the file system as the storage mechanism for a module’s code. The loader for source files (importlib.machinery.SourceFileLoader), sourceless bytecode files (importlib.machinery.SourcelessFileLoader), and extension modules (importlib.machinery.ExtensionFileLoader) are now available for direct use.

ImportError now has name and path attributes which are set when there is relevant data to provide. The message for failed imports will also provide the full name of the module now instead of just the tail end of the module’s name.

The importlib.invalidate_caches() function will now call the method with the same name on all finders cached in sys.path_importer_cache to help clean up any stored state as necessary.

可见的改变

For potential required changes to code, see the Porting Python code section.

Beyond the expanse of what importlib now exposes, there are other visible changes to import. The biggest is that sys.meta_path and sys.path_hooks now store all of the meta path finders and path entry hooks used by import. Previously the finders were implicit and hidden within the C code of import instead of being directly exposed. This means that one can now easily remove or change the order of the various finders to fit one’s needs.

Another change is that all modules have a __loader__ attribute, storing the loader used to create the module. PEP 302 has been updated to make this attribute mandatory for loaders to implement, so in the future once 3rd-party loaders have been updated people will be able to rely on the existence of the attribute. Until such time, though, import is setting the module post-load.

Loaders are also now expected to set the __package__ attribute from PEP 366. Once again, import itself is already setting this on all loaders from importlib and import itself is setting the attribute post-load.

None is now inserted into sys.path_importer_cache when no finder can be found on sys.path_hooks. Since imp.NullImporter is not directly exposed on sys.path_hooks it could no longer be relied upon to always be available to use as a value representing no finder found.

All other changes relate to semantic changes which should be taken into consideration when updating code for Python 3.3, and thus should be read about in the Porting Python code section of this document.

（由 Brett Cannon 编写的实现）

其他语言特性修改

对Python 语言核心进行的小改动：

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

  （由 Ezio Melotti 在 bpo-12753 中贡献。）

• Unicode 数据库更新至 UCD 版本 6.1.0

• Equality comparisons on range() objects now return a result reflecting the equality of the underlying sequences generated by those range objects. (bpo-13201)

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

  （由 Petri Lehtinen 在 bpo-12170 中贡献。）

• The rjust(), ljust(), and center() methods of bytes and bytearray now accept a bytearray for the fill argument. (Contributed by Petri Lehtinen in bpo-12380.)

• New methods have been added to list and bytearray: copy() and clear() (bpo-10516). Consequently, MutableSequence now also defines a clear() method (bpo-11388).

• Raw bytes literals can now be written rb"..." as well as br"...".

  （由 Antoine Pitrou 在 bpo-13748 中贡献。）

• dict.setdefault() now does only one lookup for the given key, making it atomic when used with built-in types.

  （由 Filip Gruszczyński 在 bpo-13521 中贡献。）

• The error messages produced when a function call does not match the function signature have been significantly improved.

  （由 Benjamin Peterson 贡献。）

A Finer-Grained Import Lock

Previous versions of CPython have always relied on a global import lock. This led to unexpected annoyances, such as deadlocks when importing a module would trigger code execution in a different thread as a side-effect. Clumsy workarounds were sometimes employed, such as the PyImport_ImportModuleNoBlock() C API function.

In Python 3.3, importing a module takes a per-module lock. This correctly serializes importation of a given module from multiple threads (preventing the exposure of incompletely initialized modules), while eliminating the aforementioned annoyances.

（由 Antoine Pitrou 在 bpo-9260 中贡献。）

Builtin functions and types

• open() gets a new opener parameter: the underlying file descriptor for the file object is then obtained by calling opener with (file, flags). It can be used to use custom flags like os.O_CLOEXEC for example. The 'x' mode was added: open for exclusive creation, failing if the file already exists.

• print(): added the flush keyword argument. If the flush keyword argument is true, the stream is forcibly flushed.

• hash(): hash randomization is enabled by default, see object.__hash__() and PYTHONHASHSEED.

• The str type gets a new casefold() method: return a casefolded copy of the string, casefolded strings may be used for caseless matching. For example, 'ß'.casefold() returns 'ss'.

• The sequence documentation has been substantially rewritten to better explain the binary/text sequence distinction and to provide specific documentation sections for the individual builtin sequence types (bpo-4966).

新增模块

faulthandler

This new debug module faulthandler contains functions to dump Python tracebacks explicitly, on a fault (a crash like a segmentation fault), after a timeout, or on a user signal. Call faulthandler.enable() to install fault handlers for the SIGSEGV, SIGFPE, SIGABRT, SIGBUS, and SIGILL signals. You can also enable them at startup by setting the PYTHONFAULTHANDLER environment variable or by using -X faulthandler command line option.

Linux 上的段错误示例:

$ python -q -X faulthandler
>>> import ctypes
>>> ctypes.string_at(0)
Fatal Python error: Segmentation fault
Current thread 0x00007fb899f39700:
  File "/home/python/cpython/Lib/ctypes/__init__.py", line 486 in string_at
  File "<stdin>", line 1 in <module>
Segmentation fault

ipaddress

The new ipaddress module provides tools for creating and manipulating objects representing IPv4 and IPv6 addresses, networks and interfaces (i.e. an IP address associated with a specific IP subnet).

（由 Google 和 Peter Moody 在 bpo-3144 中贡献。）

lzma

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

（由 Nadeem Vawda 和 Per Øyvind Karlsen 在 bpo-6715 中贡献。）

改进的模块

abc

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

• abc.abstractproperty has been deprecated, use property with abc.abstractmethod() instead.

• abc.abstractclassmethod has been deprecated, use classmethod with abc.abstractmethod() instead.

• abc.abstractstaticmethod has been deprecated, use staticmethod with abc.abstractmethod() instead.

（由 Pablo Galindo 在 bpo-11610 中贡献。）

abc.ABCMeta.register() now returns the registered subclass, which means it can now be used as a class decorator (bpo-10868).

array

The array module supports the long long type using q and Q type codes.

（由 Oren Tirosh 和 Hirokazu Yamamoto 在 bpo-1172711 中贡献。）

base64

ASCII-only Unicode strings are now accepted by the decoding functions of the base64 modern interface. For example, base64.b64decode('YWJj') returns b'abc'. (Contributed by Catalin Iacob in bpo-13641.)

binascii

In addition to the binary objects they normally accept, the a2b_ functions now all also accept ASCII-only strings as input. (Contributed by Antoine Pitrou in bpo-13637.)

bz2

bz2 模块已被重新编写。 在此过程中，添加了一些新的特征:

• 新的 bz2.open() 函数：以二进制或文本模式打开 bzip2 压缩文件。

• bz2.BZ2File 现在可以读写任意文件型对象，具体方式是通过其构造器的 fileobj 参数。

  （由 Nadeem Vawda 在 bpo-5863 中贡献。）

• bz2.BZ2File and bz2.decompress() can now decompress multi-stream inputs (such as those produced by the pbzip2 tool). bz2.BZ2File can now also be used to create this type of file, using the 'a' (append) mode.

  （由 Nir Aides 在 bpo-1625 中贡献。）

• bz2.BZ2File now implements all of the io.BufferedIOBase API, except for the detach() and truncate() methods.

编码器

The mbcs codec has been rewritten to handle correctly replace and ignore error handlers on all Windows versions. The mbcs codec now supports all error handlers, instead of only replace to encode and ignore to decode.

A new Windows-only codec has been added: cp65001 (bpo-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.

(bpo-12016)

Incremental CJK codec encoders are no longer reset at each call to their encode() methods. For example:

>>> 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.

(bpo-12100)

unicode_internal 编解码器已被弃用。

collections

Addition of a new ChainMap class to allow treating a number of mappings as a single unit. (Written by Raymond Hettinger for bpo-11089, made public in bpo-11297.)

The abstract base classes have been moved in a new collections.abc module, to better differentiate between the abstract and the concrete collections classes. Aliases for ABCs are still present in the collections module to preserve existing imports. (bpo-11085)

The Counter class now supports the unary + and - operators, as well as the in-place operators +=, -=, |=, and &=. (Contributed by Raymond Hettinger in bpo-13121.)

contextlib

ExitStack now provides a solid foundation for programmatic manipulation of context managers and similar cleanup functionality. Unlike the previous contextlib.nested API (which was deprecated and removed), the new API is designed to work correctly regardless of whether context managers acquire their resources in their __init__ method (for example, file objects) or in their __enter__ method (for example, synchronisation objects from the threading module).

(bpo-13585)

crypt

Addition of salt and modular crypt format (hashing method) and the mksalt() function to the crypt module.

(bpo-10924)

curses

• If the curses module is linked to the ncursesw library, use Unicode functions when Unicode strings or characters are passed (e.g. waddwstr()), and bytes functions otherwise (e.g. waddstr()).

• Use the locale encoding instead of utf-8 to encode Unicode strings.

• curses.window 添加了新的 curses.window.encoding 属性。

• curses.window 类有一个新的 get_wch() 方法用来获取一个宽字符。

• curses 模块有一个新的 unget_wch() 函数用来推入一个宽字符以便下一个 get_wch() 将返回它。

（由 Iñigo Serna 在 bpo-6755 中贡献。）

datetime

• Equality comparisons between naive and aware datetime instances now return False instead of raising TypeError (bpo-15006).

• New datetime.datetime.timestamp() method: Return POSIX timestamp corresponding to the datetime instance.

• The datetime.datetime.strftime() method supports formatting years older than 1000.

• The datetime.datetime.astimezone() method can now be called without arguments to convert datetime instance to the system timezone.

decimal

bpo-7652 - integrate fast native decimal arithmetic.

C-module and libmpdec written by Stefan Krah.

The new C version of the decimal module integrates the high speed libmpdec library for arbitrary precision correctly rounded decimal floating-point arithmetic. libmpdec conforms to IBM’s General Decimal Arithmetic Specification.

Performance gains range from 10x for database applications to 100x for numerically intensive applications. These numbers are expected gains for standard precisions used in decimal floating-point arithmetic. Since the precision is user configurable, the exact figures may vary. For example, in integer bignum arithmetic the differences can be significantly higher.

The following table is meant as an illustration. Benchmarks are available at https://www.bytereef.org/mpdecimal/quickstart.html.

	decimal.py	_decimal	加速
pi	42.02秒	0.345秒	120倍
telco	172.19秒	5.68秒	30倍
psycopg	3.57秒	0.29秒	12倍

相关特性

• The FloatOperation signal optionally enables stricter semantics for mixing floats and Decimals.

• If Python is compiled without threads, the C version automatically disables the expensive thread local context machinery. In this case, the variable HAVE_THREADS is set to False.

API 的变化

• C模块上下文限制（如下表），具体取决于计算机体系结构：

  	32位	64位
  MAX_PREC	425000000	999999999999999999
  MAX_EMAX	425000000	999999999999999999
  MIN_EMIN	-425000000	-999999999999999999

• In the context templates (DefaultContext, BasicContext and ExtendedContext) the magnitude of Emax and Emin has changed to 999999.

• The Decimal constructor in decimal.py does not observe the context limits and converts values with arbitrary exponents or precision exactly. Since the C version has internal limits, the following scheme is used: If possible, values are converted exactly, otherwise InvalidOperation is raised and the result is NaN. In the latter case it is always possible to use create_decimal() in order to obtain a rounded or inexact value.

• The power function in decimal.py is always correctly rounded. In the C version, it is defined in terms of the correctly rounded exp() and ln() functions, but the final result is only “almost always correctly rounded”.

• In the C version, the context dictionary containing the signals is a MutableMapping. For speed reasons, flags and traps always refer to the same MutableMapping that the context was initialized with. If a new signal dictionary is assigned, flags and traps are updated with the new values, but they do not reference the RHS dictionary.

• Pickling a Context produces a different output in order to have a common interchange format for the Python and C versions.

• The order of arguments in the Context constructor has been changed to match the order displayed by repr().

• The watchexp parameter in the quantize() method is deprecated.

email

Policy Framework

The email package now has a policy framework. A Policy is an object with several methods and properties that control how the email package behaves. The primary policy for Python 3.3 is the Compat32 policy, which provides backward compatibility with the email package in Python 3.2. A policy can be specified when an email message is parsed by a parser, or when a Message object is created, or when an email is serialized using a generator. Unless overridden, a policy passed to a parser is inherited by all the Message object and sub-objects created by the parser. By default a generator will use the policy of the Message object it is serializing. The default policy is compat32.

The minimum set of controls implemented by all policy objects are:

A new policy instance, with new settings, is created using the clone() method of policy objects. clone takes any of the above controls as keyword arguments. Any control not specified in the call retains its default value. Thus you can create a policy that uses \r\n linesep characters like this:

mypolicy = compat32.clone(linesep='\r\n')

Policies can be used to make the generation of messages in the format needed by your application simpler. Instead of having to remember to specify linesep='\r\n' in all the places you call a generator, you can specify it once, when you set the policy used by the parser or the Message, whichever your program uses to create Message objects. On the other hand, if you need to generate messages in multiple forms, you can still specify the parameters in the appropriate generator call. Or you can have custom policy instances for your different cases, and pass those in when you create the generator.

Provisional Policy with New Header API

While the policy framework is worthwhile all by itself, the main motivation for introducing it is to allow the creation of new policies that implement new features for the email package in a way that maintains backward compatibility for those who do not use the new policies. Because the new policies introduce a new API, we are releasing them in Python 3.3 as a provisional policy. Backwards incompatible changes (up to and including removal of the code) may occur if deemed necessary by the core developers.

The new policies are instances of EmailPolicy, and add the following additional controls:

The header_factory is the key to the new features provided by the new policies. When one of the new policies is used, any header retrieved from a Message object is an object produced by the header_factory, and any time you set a header on a Message it becomes an object produced by header_factory. All such header objects have a name attribute equal to the header name. Address and Date headers have additional attributes that give you access to the parsed data of the header. This means you can now do things like this:

>>> m = Message(policy=SMTP)
>>> m['To'] = 'Éric <foo@example.com>'
>>> m['to']
'Éric <foo@example.com>'
>>> m['to'].addresses
(Address(display_name='Éric', username='foo', domain='example.com'),)
>>> m['to'].addresses[0].username
'foo'
>>> m['to'].addresses[0].display_name
'Éric'
>>> m['Date'] = email.utils.localtime()
>>> m['Date'].datetime
datetime.datetime(2012, 5, 25, 21, 39, 24, 465484, tzinfo=datetime.timezone(datetime.timedelta(-1, 72000), 'EDT'))
>>> m['Date']
'Fri, 25 May 2012 21:44:27 -0400'
>>> print(m)
To: =?utf-8?q?=C3=89ric?= <foo@example.com>
Date: Fri, 25 May 2012 21:44:27 -0400

You will note that the unicode display name is automatically encoded as utf-8 when the message is serialized, but that when the header is accessed directly, you get the unicode version. This eliminates any need to deal with the email.header decode_header() or make_header() functions.

You can also create addresses from parts:

>>> m['cc'] = [Group('pals', [Address('Bob', 'bob', 'example.com'),
...                           Address('Sally', 'sally', 'example.com')]),
...            Address('Bonzo', addr_spec='bonz@laugh.com')]
>>> print(m)
To: =?utf-8?q?=C3=89ric?= <foo@example.com>
Date: Fri, 25 May 2012 21:44:27 -0400
cc: pals: Bob <bob@example.com>, Sally <sally@example.com>;, Bonzo <bonz@laugh.com>

Decoding to unicode is done automatically:

>>> m2 = message_from_string(str(m))
>>> m2['to']
'Éric <foo@example.com>'

When you parse a message, you can use the addresses and groups attributes of the header objects to access the groups and individual addresses:

>>> m2['cc'].addresses
(Address(display_name='Bob', username='bob', domain='example.com'), Address(display_name='Sally', username='sally', domain='example.com'), Address(display_name='Bonzo', username='bonz', domain='laugh.com'))
>>> m2['cc'].groups
(Group(display_name='pals', addresses=(Address(display_name='Bob', username='bob', domain='example.com'), Address(display_name='Sally', username='sally', domain='example.com')), Group(display_name=None, addresses=(Address(display_name='Bonzo', username='bonz', domain='laugh.com'),))

In summary, if you use one of the new policies, header manipulation works the way it ought to: your application works with unicode strings, and the email package transparently encodes and decodes the unicode to and from the RFC standard Content Transfer Encodings.

Other API Changes

New BytesHeaderParser, added to the parser module to complement HeaderParser and complete the Bytes API.

New utility functions:

• format_datetime(): given a datetime, produce a string formatted for use in an email header.

• parsedate_to_datetime(): given a date string from an email header, convert it into an aware datetime, or a naive datetime if the offset is -0000.

• localtime(): With no argument, returns the current local time as an aware datetime using the local timezone. Given an aware datetime, converts it into an aware datetime using the local timezone.

ftplib

• ftplib.FTP now accepts a source_address keyword argument to specify the (host, port) to use as the source address in the bind call when creating the outgoing socket. (Contributed by Giampaolo Rodolà in bpo-8594.)

• The FTP_TLS class now provides a new 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 bpo-12139.)

• Added ftplib.FTP.mlsd() method which provides a parsable directory listing format and deprecates ftplib.FTP.nlst() and ftplib.FTP.dir(). (Contributed by Giampaolo Rodolà in bpo-11072.)

functools

The functools.lru_cache() decorator now accepts a typed keyword argument (that defaults to False to ensure that it caches values of different types that compare equal in separate cache slots. (Contributed by Raymond Hettinger in bpo-13227.)

gc

It is now possible to register callbacks invoked by the garbage collector before and after collection using the new callbacks list.

hmac

A new compare_digest() function has been added to prevent side channel attacks on digests through timing analysis. (Contributed by Nick Coghlan and Christian Heimes in bpo-15061.)

http

http.server.BaseHTTPRequestHandler now buffers the headers and writes them all at once when end_headers() is called. A new method flush_headers() can be used to directly manage when the accumulated headers are sent. (Contributed by Andrew Schaaf in bpo-3709.)

http.server now produces valid HTML 4.01 strict output. (Contributed by Ezio Melotti in bpo-13295.)

http.client.HTTPResponse now has a readinto() method, which means it can be used as an io.RawIOBase class. (Contributed by John Kuhn in bpo-13464.)

html

html.parser.HTMLParser is now able to parse broken markup without raising errors, therefore the strict argument of the constructor and the HTMLParseError exception are now deprecated. The ability to parse broken markup is the result of a number of bug fixes that are also available on the latest bug fix releases of Python 2.7/3.2. (Contributed by Ezio Melotti in bpo-15114, and bpo-14538, bpo-13993, bpo-13960, bpo-13358, bpo-1745761, bpo-755670, bpo-13357, bpo-12629, bpo-1200313, bpo-670664, bpo-13273, bpo-12888, bpo-7311.)

A new html5 dictionary that maps HTML5 named character references to the equivalent Unicode character(s) (e.g. html5['gt;'] == '>') has been added to the html.entities module. The dictionary is now also used by HTMLParser. (Contributed by Ezio Melotti in bpo-11113 and bpo-15156.)

imaplib

The IMAP4_SSL constructor now accepts an SSLContext parameter to control parameters of the secure channel.

（由 Sijin Joseph 在 bpo-8808 中贡献。）

inspect

A new getclosurevars() function has been added. This function reports the current binding of all names referenced from the function body and where those names were resolved, making it easier to verify correct internal state when testing code that relies on stateful closures.

（由 Meador Inge 和 Nick Coghlan 在 bpo-13062 中贡献。）

A new getgeneratorlocals() function has been added. This function reports the current binding of local variables in the generator’s stack frame, making it easier to verify correct internal state when testing generators.

（由 Meador Inge 在 bpo-15153 中贡献。）

io

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

（由 David Townshend 在 bpo-12760 中贡献。）

The constructor of the TextIOWrapper class has a new write_through optional argument. If write_through is True, calls to write() are guaranteed not to be buffered: any data written on the TextIOWrapper object is immediately handled to its underlying binary buffer.

itertools

accumulate() now takes an optional func argument for providing a user-supplied binary function.

logging

The basicConfig() function now supports an optional handlers argument taking an iterable of handlers to be added to the root logger.

A class level attribute append_nul has been added to SysLogHandler to allow control of the appending of the NUL (\000) byte to syslog records, since for some daemons it is required while for others it is passed through to the log.

math

The math module has a new function, log2(), which returns the base-2 logarithm of x.

(Written by Mark Dickinson in bpo-11888.)

mmap

The read() method is now more compatible with other file-like objects: if the argument is omitted or specified as None, it returns the bytes from the current file position to the end of the mapping. (Contributed by Petri Lehtinen in bpo-12021.)

multiprocessing

The new multiprocessing.connection.wait() function allows polling multiple objects (such as connections, sockets and pipes) with a timeout. (Contributed by Richard Oudkerk in bpo-12328.)

multiprocessing.Connection objects can now be transferred over multiprocessing connections. (Contributed by Richard Oudkerk in bpo-4892.)

multiprocessing.Process now accepts a daemon keyword argument to override the default behavior of inheriting the daemon flag from the parent process (bpo-6064).

New attribute multiprocessing.Process.sentinel allows a program to wait on multiple Process objects at one time using the appropriate OS primitives (for example, select on posix systems).

New methods multiprocessing.pool.Pool.starmap() and starmap_async() provide itertools.starmap() equivalents to the existing multiprocessing.pool.Pool.map() and map_async() functions. (Contributed by Hynek Schlawack in bpo-12708.)

nntplib

The nntplib.NNTP class now supports the context management protocol to unconditionally consume 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')
>>>

（由 Giampaolo Rodolà 在 bpo-9795 中贡献。）

os

• The os module has a new pipe2() function that makes it possible to create a pipe with O_CLOEXEC or O_NONBLOCK flags set atomically. This is especially useful to avoid race conditions in multi-threaded programs.

• The os module has a new sendfile() function which provides an efficient “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. 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 bpo-10882.)

• To avoid race conditions like symlink attacks and issues with temporary files and directories, it is more reliable (and also faster) to manipulate file descriptors instead of file names. Python 3.3 enhances existing functions and introduces new functions to work on file descriptors (bpo-4761, bpo-10755 and bpo-14626).

  • The os module has a new fwalk() function similar to walk() except that it also yields file descriptors referring to the directories visited. This is especially useful to avoid symlink races.

  • The following functions get new optional dir_fd (paths relative to directory descriptors) and/or follow_symlinks (not following symlinks): access(), chflags(), chmod(), chown(), link(), lstat(), mkdir(), mkfifo(), mknod(), open(), readlink(), remove(), rename(), replace(), rmdir(), stat(), symlink(), unlink(), utime(). Platform support for using these parameters can be checked via the sets os.supports_dir_fd and os.supports_follows_symlinks.

  • The following functions now support a file descriptor for their path argument: chdir(), chmod(), chown(), execve(), listdir(), pathconf(), exists(), stat(), statvfs(), utime(). Platform support for this can be checked via the os.supports_fd set.

• access() accepts an effective_ids keyword argument to turn on using the effective uid/gid rather than the real uid/gid in the access check. Platform support for this can be checked via the supports_effective_ids set.

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

  (Patch submitted by Giampaolo Rodolà in bpo-10784.)

• The new os.replace() function allows cross-platform renaming of a file with overwriting the destination. With os.rename(), an existing destination file is overwritten under POSIX, but raises an error under Windows. (Contributed by Antoine Pitrou in bpo-8828.)

• The stat family of functions (stat(), fstat(), and lstat()) now support reading a file’s timestamps with nanosecond precision. Symmetrically, utime() can now write file timestamps with nanosecond precision. (Contributed by Larry Hastings in bpo-14127.)

• The new os.get_terminal_size() function queries the size of the terminal attached to a file descriptor. See also shutil.get_terminal_size(). (Contributed by Zbigniew Jędrzejewski-Szmek in bpo-13609.)

• New functions to support Linux extended attributes (bpo-12720): getxattr(), listxattr(), removexattr(), setxattr().

• New interface to the scheduler. These functions control how a process is allocated CPU time by the operating system. New functions: sched_get_priority_max(), sched_get_priority_min(), sched_getaffinity(), sched_getparam(), sched_getscheduler(), sched_rr_get_interval(), sched_setaffinity(), sched_setparam(), sched_setscheduler(), sched_yield(),

• New functions to control the file system:

  • posix_fadvise(): Announces an intention to access data in a specific pattern thus allowing the kernel to make optimizations.

  • posix_fallocate(): Ensures that enough disk space is allocated for a file.

  • sync(): Force write of everything to disk.

• Additional new posix functions:

  • lockf(): Apply, test or remove a POSIX lock on an open file descriptor.

  • pread(): Read from a file descriptor at an offset, the file offset remains unchanged.

  • pwrite(): Write to a file descriptor from an offset, leaving the file offset unchanged.

  • readv(): Read from a file descriptor into a number of writable buffers.

  • truncate(): Truncate the file corresponding to path, so that it is at most length bytes in size.

  • waitid(): Wait for the completion of one or more child processes.

  • writev(): Write the contents of buffers to a file descriptor, where buffers is an arbitrary sequence of buffers.

  • getgrouplist() (bpo-9344): Return list of group ids that specified user belongs to.

• times() and uname(): Return type changed from a tuple to a tuple-like object with named attributes.

• Some platforms now support additional constants for the lseek() function, such as os.SEEK_HOLE and os.SEEK_DATA.

• New constants RTLD_LAZY, RTLD_NOW, RTLD_GLOBAL, RTLD_LOCAL, RTLD_NODELETE, RTLD_NOLOAD, and RTLD_DEEPBIND are available on platforms that support them. These are for use with the sys.setdlopenflags() function, and supersede the similar constants defined in ctypes and DLFCN. (Contributed by Victor Stinner in bpo-13226.)

• os.symlink() now accepts (and ignores) the target_is_directory keyword argument on non-Windows platforms, to ease cross-platform support.

pdb

Tab-completion is now available not only for command names, but also their arguments. For example, for the break command, function and file names are completed.

（由 Georg Brandl 在 bpo-14210 中贡献）

pickle

pickle.Pickler objects now have an optional dispatch_table attribute allowing per-pickler reduction functions to be set.

（由 Richard Oudkerk 在 bpo-14166 中贡献。）

pydoc

The Tk GUI and the serve() function have been removed from the pydoc module: pydoc -g and serve() have been deprecated in Python 3.2.

re

str regular expressions now support \u and \U escapes.

（由 Serhiy Storchaka 在 bpo-3665 中贡献。）

sched

• 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 scheduler in non-blocking applications. (Contributed by Giampaolo Rodolà in bpo-13449.)

• scheduler class can now be safely used in multi-threaded environments. (Contributed by Josiah Carlson and Giampaolo Rodolà in bpo-8684.)

• timefunc and delayfunct parameters of scheduler class constructor are now optional and defaults to time.time() and time.sleep() respectively. (Contributed by Chris Clark in bpo-13245.)

• enter() and enterabs() argument parameter is now optional. (Contributed by Chris Clark in bpo-13245.)

• enter() and enterabs() now accept a kwargs parameter. (Contributed by Chris Clark in bpo-13245.)

select

Solaris and derivative platforms have a new class select.devpoll for high performance asynchronous sockets via /dev/poll. (Contributed by Jesús Cea Avión in bpo-6397.)

shlex

The previously undocumented helper function quote from the pipes modules has been moved to the shlex module and documented. quote() properly escapes all characters in a string that might be otherwise given special meaning by the shell.

shutil

• 新的函数：

  • disk_usage(): provides total, used and free disk space statistics. (Contributed by Giampaolo Rodolà in bpo-12442.)

  • 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 bpo-12191.)

  • shutil.get_terminal_size(): returns the size of the terminal window to which the interpreter is attached. (Contributed by Zbigniew Jędrzejewski-Szmek in bpo-13609.)

• copy2() and copystat() now preserve file timestamps with nanosecond precision on platforms that support it. They also preserve file “extended attributes” on Linux. (Contributed by Larry Hastings in bpo-14127 and bpo-15238.)

• Several functions now take an optional symlinks argument: when that parameter is true, symlinks aren’t dereferenced and the operation instead acts on the symlink itself (or creates one, if relevant). (Contributed by Hynek Schlawack in bpo-12715.)

• When copying files to a different file system, move() now handles symlinks the way the posix mv command does, recreating the symlink rather than copying the target file contents. (Contributed by Jonathan Niehof in bpo-9993.) move() now also returns the dst argument as its result.

• rmtree() is now resistant to symlink attacks on platforms which support the new dir_fd parameter in os.open() and os.unlink(). (Contributed by Martin von Löwis and Hynek Schlawack in bpo-4489.)

signal

• signal 模块新增的函数:

  • pthread_sigmask(): 获取和/或改变调用方线程的信号掩码（由 Jean-Paul Calderone 在 bpo-8407 中贡献）；

  • pthread_kill(): 向指定线程发送信号；

  • sigpending(): 检查挂起的函数；

  • sigwait(): 等待一个信号;

  • sigwaitinfo(): 等待信号，返回相关的详细信息；

  • sigtimedwait(): like 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.

• signal.signal() and signal.siginterrupt() raise an OSError, instead of a RuntimeError: OSError has an errno attribute.

smtpd

The smtpd module now supports RFC 5321 (extended SMTP) and RFC 1870 (size extension). Per the standard, these extensions are enabled if and only if the client initiates the session with an EHLO command.

(Initial ELHO support by Alberto Trevino. Size extension by Juhana Jauhiainen. Substantial additional work on the patch contributed by Michele Orrù and Dan Boswell. bpo-8739)

smtplib

The SMTP, SMTP_SSL, and LMTP classes now accept a source_address keyword argument to specify the (host, port) to use as the source address in the bind call when creating the outgoing socket. (Contributed by Paulo Scardine in bpo-11281.)

SMTP now supports the context management protocol, allowing an SMTP instance to be used in a with statement. (Contributed by Giampaolo Rodolà in bpo-11289.)

The SMTP_SSL constructor and the starttls() method now accept an SSLContext parameter to control parameters of the secure channel. (Contributed by Kasun Herath in bpo-8809.)

socket

• The socket class now exposes additional methods to process ancillary data when supported by the underlying platform:

  • sendmsg()

  • recvmsg()

  • recvmsg_into()

  (Contributed by David Watson in bpo-6560, based on an earlier patch by Heiko Wundram)

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

  (Contributed by Matthias Fuchs, updated by Tiago Gonçalves in bpo-10141.)

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

• The socket class now supports the PF_SYSTEM protocol family on OS X. (Contributed by Michael Goderbauer in bpo-13777.)

• New function sethostname() allows the hostname to be set on Unix systems if the calling process has sufficient privileges. (Contributed by Ross Lagerwall in bpo-10866.)

socketserver

BaseServer now has an overridable method service_actions() that is called by the serve_forever() method in the service loop. ForkingMixIn now uses this to clean up zombie child processes. (Contributed by Justin Warkentin in bpo-11109.)

sqlite3

New sqlite3.Connection method set_trace_callback() can be used to capture a trace of all sql commands processed by sqlite. (Contributed by Torsten Landschoff in bpo-11688.)

ssl

• The ssl module has two new random generation functions:

  • RAND_bytes(): generate cryptographically strong pseudo-random bytes.

  • RAND_pseudo_bytes(): 生成伪随机字节。

  （由 Victor Stinner 在 bpo-12049 中贡献。）

• The 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 bpo-11183.)

• load_cert_chain() now accepts a password argument to be used if the private key is encrypted. (Contributed by Adam Simpkins in bpo-12803.)

• Diffie-Hellman key exchange, both regular and Elliptic Curve-based, is now supported through the load_dh_params() and set_ecdh_curve() methods. (Contributed by Antoine Pitrou in bpo-13626 and bpo-13627.)

• SSL sockets have a new get_channel_binding() method allowing the implementation of certain authentication mechanisms such as SCRAM-SHA-1-PLUS. (Contributed by Jacek Konieczny in bpo-12551.)

• You can query the SSL compression algorithm used by an SSL socket, thanks to its new compression() method. The new attribute OP_NO_COMPRESSION can be used to disable compression. (Contributed by Antoine Pitrou in bpo-13634.)

• Support has been added for the Next Protocol Negotiation extension using the ssl.SSLContext.set_npn_protocols() method. (Contributed by Colin Marc in bpo-14204.)

• SSL errors can now be introspected more easily thanks to library and reason attributes. (Contributed by Antoine Pitrou in bpo-14837.)

• The get_server_certificate() function now supports IPv6. (Contributed by Charles-François Natali in bpo-11811.)

• New attribute OP_CIPHER_SERVER_PREFERENCE allows setting SSLv3 server sockets to use the server’s cipher ordering preference rather than the client’s (bpo-13635).

stat

The undocumented tarfile.filemode function has been moved to stat.filemode(). It can be used to convert a file’s mode to a string of the form ‘-rwxrwxrwx’.

（由 Giampaolo Rodolà 在 bpo-14807 中贡献。）

struct

The struct module now supports ssize_t and size_t via the new codes n and N, respectively. (Contributed by Antoine Pitrou in bpo-3163.)

subprocess

Command strings can now be bytes objects on posix platforms. (Contributed by Victor Stinner in bpo-8513.)

A new constant DEVNULL allows suppressing output in a platform-independent fashion. (Contributed by Ross Lagerwall in bpo-5870.)

sys

The sys module has a new thread_info named tuple holding information about the thread implementation (bpo-11223).

tarfile

tarfile now supports lzma encoding via the lzma module. (Contributed by Lars Gustäbel in bpo-5689.)

tempfile

tempfile.SpooledTemporaryFile‘s truncate() method now accepts a size parameter. (Contributed by Ryan Kelly in bpo-9957.)

textwrap

The textwrap module has a new indent() that makes it straightforward to add a common prefix to selected lines in a block of text (bpo-13857).

threading

threading.Condition, threading.Semaphore, threading.BoundedSemaphore, threading.Event, and threading.Timer, all of which used to be factory functions returning a class instance, are now classes and may be subclassed. (Contributed by Éric Araujo in bpo-10968.)

The threading.Thread constructor now accepts a daemon keyword argument to override the default behavior of inheriting the daemon flag value from the parent thread (bpo-6064).

The formerly private function _thread.get_ident is now available as the public function threading.get_ident(). This eliminates several cases of direct access to the _thread module in the stdlib. Third party code that used _thread.get_ident should likewise be changed to use the new public interface.

time

The PEP 418 added new functions to the time module:

• get_clock_info(): Get information on a clock.

• monotonic(): Monotonic clock (cannot go backward), not affected by system clock updates.

• perf_counter(): Performance counter with the highest available resolution to measure a short duration.

• process_time(): Sum of the system and user CPU time of the current process.

Other new functions:

• clock_getres(), clock_gettime() and clock_settime() functions with CLOCK__xxx_ constants. (Contributed by Victor Stinner in bpo-10278.)

To improve cross platform consistency, sleep() now raises a ValueError when passed a negative sleep value. Previously this was an error on posix, but produced an infinite sleep on Windows.

types

Add a new types.MappingProxyType class: Read-only proxy of a mapping. (bpo-14386)

The new functions types.new_class() and types.prepare_class() provide support for PEP 3115 compliant dynamic type creation. (bpo-14588)

unittest

assertRaises(), assertRaisesRegex(), assertWarns(), and assertWarnsRegex() now accept a keyword argument msg when used as context managers. (Contributed by Ezio Melotti and Winston Ewert in bpo-10775.)

unittest.TestCase.run() now returns the TestResult object.

urllib

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

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

(bpo-1673007)

webbrowser

The webbrowser module supports more “browsers”: Google Chrome (named chrome, chromium, chrome-browser or chromium-browser depending on the version and operating system), and the generic launchers xdg-open, from the FreeDesktop.org project, and gvfs-open, which is the default URI handler for GNOME 3. (The former contributed by Arnaud Calmettes in bpo-13620, the latter by Matthias Klose in bpo-14493.)

xml.etree.ElementTree

The xml.etree.ElementTree module now imports its C accelerator by default; there is no longer a need to explicitly import xml.etree.cElementTree (this module stays for backwards compatibility, but is now deprecated). In addition, the iter family of methods of Element has been optimized (rewritten in C). The module’s documentation has also been greatly improved with added examples and a more detailed reference.

zlib

New attribute zlib.Decompress.eof makes it possible to distinguish between a properly formed compressed stream and an incomplete or truncated one. (Contributed by Nadeem Vawda in bpo-12646.)

New attribute zlib.ZLIB_RUNTIME_VERSION reports the version string of the underlying zlib library that is loaded at runtime. (Contributed by Torsten Landschoff in bpo-12306.)

性能优化

已增加的主要性能改善:

• 得益于:pep:393 ，Unicode 字符串的某些操作已得到优化：

  • the memory footprint is divided by 2 to 4 depending on the text

  • 将 ASCII 字符串编码为 UTF-8 不再需要对字符进行编码，UTF-8 的表示法与 ASCII 的表示法是共享的

  • the UTF-8 encoder has been optimized

  • repeating a single ASCII letter and getting a substring of an ASCII string is 4 times faster

• UTF-8 编码现在快 2 到 4 倍。 UTF-16 编码的速度现在提高了 10 倍。

  （由 Serhiy Storchaka 在 bpo-14624, bpo-14738 和 bpo-15026 中贡献）。

构建和 C API 的改变

针对 Python 构建过程和 C API 的改变包括:

• 新的 PEP 3118 相关功能：

  • PyMemoryView_FromMemory()

• PEP 393 添加了新的 Unicode 类型，宏和函数

  • 高阶 API

    • PyUnicode_CopyCharacters()

    • PyUnicode_FindChar()

    • PyUnicode_GetLength(), PyUnicode_GET_LENGTH

    • PyUnicode_New()

    • PyUnicode_Substring()

    • PyUnicode_ReadChar(), PyUnicode_WriteChar()

  • 低阶 API:

    • Py_UCS1, Py_UCS2, Py_UCS4 类型

    • PyASCIIObject 和 PyCompactUnicodeObject 结构

    • PyUnicode_READY

    • PyUnicode_FromKindAndData()

    • PyUnicode_AsUCS4(), PyUnicode_AsUCS4Copy()

    • PyUnicode_DATA, PyUnicode_1BYTE_DATA, PyUnicode_2BYTE_DATA, PyUnicode_4BYTE_DATA

    • PyUnicode_KIND with PyUnicode_Kind enum: PyUnicode_WCHAR_KIND, PyUnicode_1BYTE_KIND, PyUnicode_2BYTE_KIND, PyUnicode_4BYTE_KIND

    • PyUnicode_READ, PyUnicode_READ_CHAR, PyUnicode_WRITE

    • PyUnicode_MAX_CHAR_VALUE

• PyArg_ParseTuple 现在接受 c 格式的 bytearray (bpo-12380)。

弃用

不支持的操作系统

由于缺少维护人员，不再支持 OS/2 和 VMS 系统 。

由于维护负担，将 COMSPEC 设置为 command.com 的 Windows平台（含Windows 2000）不再受支持。

OSF支持在3.2中被弃用，现在已经被完全删除。

已弃用的 Python 模块、函数和方法

• 向 object.__format__() 传递非空字符串的做法已被弃用，在 Python 3.4 中会产生一个 TypeError (bpo-9856)。

• 由于:pep:393，unicode_internal 编解码器已被弃用 。请使用 UTF-8、UTF-16 (utf-16-le 或``utf-16-be``) 或 UTF-32 (utf-32-le 或``utf-32-be``)

• ftplib.FTP.nlst() and ftplib.FTP.dir(): use ftplib.FTP.mlsd()

• platform.popen(): use the subprocess module. Check especially the 使用 subprocess 模块替换旧函数 section (bpo-11377).

• bpo-13374: The Windows bytes API has been deprecated in the os module. Use Unicode filenames, instead of bytes filenames, to not depend on the ANSI code page anymore and to support any filename.

• bpo-13988: The xml.etree.cElementTree module is deprecated. The accelerator is used automatically whenever available.

• The behaviour of time.clock() depends on the platform: use the new time.perf_counter() or time.process_time() function instead, depending on your requirements, to have a well defined behaviour.

• The os.stat_float_times() function is deprecated.

• abc module:

  • abc.abstractproperty has been deprecated, use property with abc.abstractmethod() instead.

  • abc.abstractclassmethod has been deprecated, use classmethod with abc.abstractmethod() instead.

  • abc.abstractstaticmethod has been deprecated, use staticmethod with abc.abstractmethod() instead.

• importlib package:

  • importlib.abc.SourceLoader.path_mtime() is now deprecated in favour of importlib.abc.SourceLoader.path_stats() as bytecode files now store both the modification time and size of the source file the bytecode file was compiled from.

已弃用的 C API 函数和类型

Py_UNICODE 已经在 PEP 393 弃用，并将于 Python 4 中移除。所有使用此类型的函数都已弃用：

Unicode functions and methods using Py_UNICODE and Py_UNICODE* types:

• PyUnicode_FromUnicode: use PyUnicode_FromWideChar() or PyUnicode_FromKindAndData()

• PyUnicode_AS_UNICODE, PyUnicode_AsUnicode(), PyUnicode_AsUnicodeAndSize(): use PyUnicode_AsWideCharString()

• PyUnicode_AS_DATA: use PyUnicode_DATA with PyUnicode_READ and PyUnicode_WRITE

• PyUnicode_GET_SIZE, PyUnicode_GetSize(): use PyUnicode_GET_LENGTH or PyUnicode_GetLength()

• PyUnicode_GET_DATA_SIZE: use PyUnicode_GET_LENGTH(str) * PyUnicode_KIND(str) (only work on ready strings)

• PyUnicode_AsUnicodeCopy(): use PyUnicode_AsUCS4Copy() or PyUnicode_AsWideCharString()

• PyUnicode_GetMax()

Functions and macros manipulating Py_UNICODE* strings:

• Py_UNICODE_strlen(): use PyUnicode_GetLength() or PyUnicode_GET_LENGTH

• Py_UNICODE_strcat(): use PyUnicode_CopyCharacters() or PyUnicode_FromFormat()

• Py_UNICODE_strcpy(), Py_UNICODE_strncpy(), Py_UNICODE_COPY(): use PyUnicode_CopyCharacters() or PyUnicode_Substring()

• Py_UNICODE_strcmp(): use PyUnicode_Compare()

• Py_UNICODE_strncmp(): use PyUnicode_Tailmatch()

• Py_UNICODE_strchr(), Py_UNICODE_strrchr(): use PyUnicode_FindChar()

• Py_UNICODE_FILL(): use PyUnicode_Fill()

• Py_UNICODE_MATCH

编码器:

• PyUnicode_Encode(): use PyUnicode_AsEncodedObject()

• PyUnicode_EncodeUTF7()

• PyUnicode_EncodeUTF8(): use PyUnicode_AsUTF8() or PyUnicode_AsUTF8String()

• PyUnicode_EncodeUTF32()

• PyUnicode_EncodeUTF16()

• PyUnicode_EncodeUnicodeEscape() use PyUnicode_AsUnicodeEscapeString()

• PyUnicode_EncodeRawUnicodeEscape() use PyUnicode_AsRawUnicodeEscapeString()

• PyUnicode_EncodeLatin1(): use PyUnicode_AsLatin1String()

• PyUnicode_EncodeASCII(): use PyUnicode_AsASCIIString()

• PyUnicode_EncodeCharmap()

• PyUnicode_TranslateCharmap()

• PyUnicode_EncodeMBCS(): use PyUnicode_AsMBCSString() or PyUnicode_EncodeCodePage() (with CP_ACP code_page)

• PyUnicode_EncodeDecimal(), PyUnicode_TransformDecimalToASCII()

弃用的功能

array 模块的``‘u’`` 格式代码现已弃用，将在 Python 4 中与 (Py_UNICODE) API 的其他部分一起删除。

移植到 Python 3.3

本节列出了先前描述的更改以及可能需要更改代码的其他错误修正.

移植 Python 代码

• 默认启用哈希随机化。 将 PYTHONHASHSEED 环境变量设为 0 可禁用哈希随机化。 另请参阅 object.__hash__() 方法。

• bpo-12326: 在 Linux 上，sys.platform 不再包含主要版本。现在它始终是 “linux”，而不是 “linux2” 或 “linux3”，这取决于用于构建 Python 的 Linux 版本。请用 sys.platform.startswith(‘linux’) 替换 sys.platform == ‘linux2’，如果不需要支持较旧的 Python 版本，则可直接替换成 sys.platform == ‘linux’。

• bpo-13847, bpo-14180: time 和 datetime: 现在如果时间戳超出范围将会引发 OverflowError 而不是 ValueError。 现在如果 C 函数 gmtime() 或 localtime() 失败 将会引发 OSError。

• The default finders used by import now utilize a cache of what is contained within a specific directory. If you create a Python source file or sourceless bytecode file, make sure to call importlib.invalidate_caches() to clear out the cache for the finders to notice the new file.

• ImportError now uses the full name of the module that was attempted to be imported. Doctests that check ImportErrors’ message will need to be updated to use the full name of the module instead of just the tail of the name.

• The index argument to __import__() now defaults to 0 instead of -1 and no longer support negative values. It was an oversight when PEP 328 was implemented that the default value remained -1. If you need to continue to perform a relative import followed by an absolute import, then perform the relative import using an index of 1, followed by another import using an index of 0. It is preferred, though, that you use importlib.import_module() rather than call __import__() directly.

• __import__() no longer allows one to use an index value other than 0 for top-level modules. E.g. __import__('sys', level=1) is now an error.

• Because sys.meta_path and sys.path_hooks now have finders on them by default, you will most likely want to use list.insert() instead of list.append() to add to those lists.

• Because None is now inserted into sys.path_importer_cache, if you are clearing out entries in the dictionary of paths that do not have a finder, you will need to remove keys paired with values of None and imp.NullImporter to be backwards-compatible. This will lead to extra overhead on older versions of Python that re-insert None into sys.path_importer_cache where it represents the use of implicit finders, but semantically it should not change anything.

• importlib.abc.Finder no longer specifies a find_module() abstract method that must be implemented. If you were relying on subclasses to implement that method, make sure to check for the method’s existence first. You will probably want to check for find_loader() first, though, in the case of working with path entry finders.

• pkgutil has been converted to use importlib internally. This eliminates many edge cases where the old behaviour of the PEP 302 import emulation failed to match the behaviour of the real import system. The import emulation itself is still present, but is now deprecated. The pkgutil.iter_importers() and pkgutil.walk_packages() functions special case the standard import hooks so they are still supported even though they do not provide the non-standard iter_modules() method.

• A longstanding RFC-compliance bug (bpo-1079) in the parsing done by email.header.decode_header() has been fixed. Code that uses the standard idiom to convert encoded headers into unicode (str(make_header(decode_header(h))) will see no change, but code that looks at the individual tuples returned by decode_header will see that whitespace that precedes or follows ASCII sections is now included in the ASCII section. Code that builds headers using make_header should also continue to work without change, since make_header continues to add whitespace between ASCII and non-ASCII sections if it is not already present in the input strings.

• email.utils.formataddr() now does the correct content transfer encoding when passed non-ASCII display names. Any code that depended on the previous buggy behavior that preserved the non-ASCII unicode in the formatted output string will need to be changed (bpo-1690608).

• poplib.POP3.quit() may now raise protocol errors like all other poplib methods. Code that assumes quit does not raise poplib.error_proto errors may need to be changed if errors on quit are encountered by a particular application (bpo-11291).

• The strict argument to email.parser.Parser, deprecated since Python 2.4, has finally been removed.

• The deprecated method unittest.TestCase.assertSameElements has been removed.

• The deprecated variable time.accept2dyear has been removed.

• 被弃用的 Context._clamp 属性已从 decimal 模块中移除。 在此之前它已被公有属性 clamp 取代。 （参见 bpo-8540。）

• The undocumented internal helper class SSLFakeFile has been removed from smtplib, since its functionality has long been provided directly by socket.socket.makefile().

• Passing a negative value to time.sleep() on Windows now raises an error instead of sleeping forever. It has always raised an error on posix.

• The ast.__version__ constant has been removed. If you need to make decisions affected by the AST version, use sys.version_info to make the decision.

• Code that used to work around the fact that the threading module used factory functions by subclassing the private classes will need to change to subclass the now-public classes.

• The undocumented debugging machinery in the threading module has been removed, simplifying the code. This should have no effect on production code, but is mentioned here in case any application debug frameworks were interacting with it (bpo-13550).

移植 C 代码

• In the course of changes to the buffer API the undocumented smalltable member of the Py_buffer structure has been removed and the layout of the PyMemoryViewObject has changed.

  All extensions relying on the relevant parts in memoryobject.h or object.h must be rebuilt.

• Due to PEP 393, the 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 PEP 393, you have to convert your code to the new Unicode API.

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

• PyImport_GetMagicNumber() now returns -1 upon failure.

• As a negative value for the level argument to __import__() is no longer valid, the same now holds for PyImport_ImportModuleLevel(). This also means that the value of level used by PyImport_ImportModuleEx() is now 0 instead of -1.

Building C extensions

• The range of possible file names for C extensions has been narrowed. Very rarely used spellings have been suppressed: under POSIX, files named xxxmodule.so, xxxmodule.abi3.so and xxxmodule.cpython-*.so are no longer recognized as implementing the xxx module. If you had been generating such files, you have to switch to the other spellings (i.e., remove the module string from the file names).

  （在 bpo-14040 中实现。）

命令行开关的变化

• 删除了 -Q命令-line旗标 和相关工具。 检查 sys.flags.division_warning 的代码需要更新。

  （bpo-10998，由 Éric Araujo 贡献。）

• 当 python 附带 -S 启动时，import site 将不再向模块搜索路径添加 site 专属路径。 在之前版本中则会这样做。

  （由 Carl Meyer 在 bpo-11591 中贡献并由 Éric Araujo 进行修改。）