Cython 0.23 发布,Python 的 C 语言扩展

来源: OSCHINA
编辑: oschina
2015-08-10 00:00:00

Cython 0.23 发布,此版本相比第二个测试版本改进很小,还有一些 bug 修复。

https://pypi.python.org/pypi/Cython/0.23

此版本是个重要的特性版本,值得关注的改进如下:

  * PEP 492 (async/await) 和 PEP 448 (star-unpacking generalizations)
  * 内联生长器表达式
  * 支持 C++ 分配和 bool() 操作符
  * 支持覆盖分析

强烈建议更新,下载:

http://cython.org/

http://cython.org/release/Cython-0.23.tar.gz

http://cython.org/release/Cython-0.23.zip


SHA1 sums:
5a49123b39d89a20fb88bdc7886210

c39cdff695  Cython-0.23.tar.gz
9bec6c6df861be49ed608a0ab688b67bba543abe  Cython-0.23.zip

0.23 (2015-08-08)
=================

新特性
--------------

* PEP 492 (async/await) was implemented.
  See https://www.python.org/dev/peps/pep-0492/

* PEP 448 (Additional Unpacking Generalizations) was implemented.
  See https://www.python.org/dev/peps/pep-0448/

* Support for coverage.py 4.0+ can be enabled by adding the plugin
  "Cython.Coverage" to the ".coveragerc" config file.

* Annotated HTML source pages can integrate (XML) coverage reports.

* Tracing is supported in ``nogil`` functions/sections and module init
  code.

* When generators are used in a Cython module and the module imports the
  modules "inspect" and/or "asyncio", Cython enables interoperability by
  patching these modules during the import to recognise Cython's internal
  generator and coroutine types. This can be disabled by C compiling the
  module with "-D CYTHON_PATCH_ASYNCIO=0" or "-D CYTHON_PATCH_INSPECT=0"

* When generators or coroutines are used in a Cython module, their types
  are registered with the ``Generator`` and ``Coroutine`` ABCs in the
  ``collections`` or ``collections.abc`` stdlib module at import time to
  enable interoperability with code that needs to detect and process Python
  generators/coroutines.  These ABCs were added in CPython 3.5 and are
  available for older Python versions through the ``backports_abc`` module
  on PyPI.  See https://bugs.python.org/issue24018

* Adding/subtracting/dividing/modulus and equality comparisons with
  constant Python floats and small integers are faster.

* Binary and/or/xor/rshift operations with small constant Python integers
  are faster.

* When called on generator expressions, the builtins ``all()``, ``any()``,
  ``dict()``, ``list()``, ``set()``, ``sorted()`` and ``unicode.join()``
  avoid the generator iteration overhead by inlining a part of their
  functionality into the for-loop.

* Keyword argument dicts are no longer copied on function entry when they
  are not being used or only passed through to other function calls (e.g.
  in wrapper functions).

* The ``PyTypeObject`` declaration in ``cpython.object`` was extended.

* The builtin ``type`` type is now declared as PyTypeObject in source,
  allowing for extern functions taking type parameters to have the correct
  C signatures.  Note that this might break code that uses ``type`` just
  for passing around Python types in typed variables.  Removing the type
  declaration provides a backwards compatible fix.

* ``wraparound()`` and ``boundscheck()`` are available as no-ops in pure
  Python mode.

* Const iterators were added to the provided C++ STL declarations.

* Smart pointers were added to the provided C++ STL declarations.
  Patch by Daniel Filonik.

* ``NULL`` is allowed as default argument when embedding signatures.
  This fixes ticket 843.

* When compiling with ``--embed``, the internal module name is changed to
  ``__main__`` to allow arbitrary program names, including those that would
  be invalid for modules.  Note that this prevents reuse of the generated
  C code as an importable module.

* External C++ classes that overload the assignment operator can be used.
  Patch by Ian Henriksen.

* Support operator bool() for C++ classes so they can be used in
  if-statements.

Bugs 修复
----------

* Calling "yield from" from Python on a Cython generator that returned a
  value triggered a crash in CPython.  This is now being worked around.
  See https://bugs.python.org/issue23996

* Language level 3 did not enable true division (a.k.a. float division)
  for integer operands.

* Functions with fused argument types that included a generic 'object'
  fallback could end up using that fallback also for other explicitly
  listed object types.

* Relative cimports could accidentally fall back to trying an absolute
  cimport on failure.

* The result of calling a C struct constructor no longer requires an
  intermediate assignment when coercing to a Python dict.

* C++ exception declarations with mapping functions could fail to compile
  when pre-declared in .pxd files.

* ``cpdef void`` methods are now permitted.

* ``abs(cint)`` could fail to compile in MSVC and used sub-optimal code
  in C++.  Patch by David Vierra, original patch by Michael Enßlin.

* Buffer index calculations using index variables with small C integer
  types could overflow for large buffer sizes.
  Original patch by David Vierra.

* C unions use a saner way to coerce from and to Python dicts.

* When compiling a module ``foo.pyx``, the directories in ``sys.path``
  are no longer searched when looking for ``foo.pxd``.
  Patch by Jeroen Demeyer.

* Memory leaks in the embedding main function were fixed.
  Original patch by Michael Enßlin.

* Some complex Python expressions could fail to compile inside of finally
  clauses.

* Unprefixed 'str' literals were not supported as C varargs arguments.

* Fixed type errors in conversion enum types to/from Python.  Note that
  this imposes stricter correctness requirements on enum declarations.

其他改进
-------------

* Changed mangling scheme in header files generated by ``cdef api``
  declarations.

* Installation under CPython 3.3+ no longer requires a pass of the
  2to3 tool.  This also makes it possible to run Cython in Python
  3.3+ from a source checkout without installing it first.
  Patch by Petr Viktorin.

* ``jedi-typer.py`` (in ``Tools/``) was extended and renamed to
  ``jedityper.py`` (to make it importable) and now works with and
  requires Jedi 0.9.  Patch by Tzer-jen Wei.

更多内容请看发行说明

准确说Cython是单独的一门语言,专门用来写在Python里面import用的扩展库。实际上Cython的语法基本上跟Python一致,而 Cython有专门的“编译器”先将 Cython代码转变成C(自动加入了一大堆的C-Python API),然后使用C编译器编译出最终的Python可调用的模块。

展开阅读全文
8 收藏
分享
0 评论
8 收藏
分享
返回顶部
顶部
返回顶部
顶部