22

✏️ Editing: 2.7.html

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
 "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">

<html xmlns="http://www.w3.org/1999/xhtml">
 <head>
 <meta http-equiv="X-UA-Compatible" content="IE=Edge" />
 <meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
 <title>What’s New in Python 2.7 &#8212; Python 3.6.7 documentation</title>
 <link rel="stylesheet" href="../_static/pydoctheme.css" type="text/css" />
 <link rel="stylesheet" href="../_static/pygments.css" type="text/css" />
 <script type="text/javascript" id="documentation_options" data-url_root="../" src="../_static/documentation_options.js"></script>
 <script type="text/javascript" src="../_static/jquery.js"></script>
 <script type="text/javascript" src="../_static/underscore.js"></script>
 <script type="text/javascript" src="../_static/doctools.js"></script>
 <script type="text/javascript" src="../_static/sidebar.js"></script>
 <link rel="search" type="application/opensearchdescription+xml"
 title="Search within Python 3.6.7 documentation"
 href="../_static/opensearch.xml"/>
 <link rel="author" title="About these documents" href="../about.html" />
 <link rel="index" title="Index" href="../genindex.html" />
 <link rel="search" title="Search" href="../search.html" />
 <link rel="copyright" title="Copyright" href="../copyright.html" />
 <link rel="next" title="What’s New in Python 2.6" href="2.6.html" />
 <link rel="prev" title="What’s New In Python 3.0" href="3.0.html" />
 <link rel="shortcut icon" type="image/png" href="../_static/py.png" />
 <link rel="canonical" href="https://docs.python.org/3/whatsnew/2.7.html" />
 
 <script type="text/javascript" src="../_static/copybutton.js"></script>

</head><body> 
 <div class="related" role="navigation" aria-label="related navigation">
 <h3>Navigation</h3>
 <ul>
 <li class="right" style="margin-right: 10px">
 <a href="../genindex.html" title="General Index"
 accesskey="I">index</a></li>
 <li class="right" >
 <a href="../py-modindex.html" title="Python Module Index"
 >modules</a> |</li>
 <li class="right" >
 <a href="2.6.html" title="What’s New in Python 2.6"
 accesskey="N">next</a> |</li>
 <li class="right" >
 <a href="3.0.html" title="What’s New In Python 3.0"
 accesskey="P">previous</a> |</li>
 <li><img src="../_static/py.png" alt=""
 style="vertical-align: middle; margin-top: -1px"/></li>
 <li><a href="https://www.python.org/">Python</a> &#187;</li>
 <li>
 <a href="../index.html">3.6.7 Documentation</a> &#187;
 </li>

<li class="nav-item nav-item-1"><a href="index.html" accesskey="U">What’s New in Python</a> &#187;</li>
 <li class="right">

<div class="inline-search" style="display: none" role="search">
 <form class="inline-search" action="../search.html" method="get">
 <input placeholder="Quick search" type="text" name="q" />
 <input type="submit" value="Go" />
 <input type="hidden" name="check_keywords" value="yes" />
 <input type="hidden" name="area" value="default" />
 </form>
 </div>
 <script type="text/javascript">$('.inline-search').show(0);</script>
 |
 </li>

</ul>
 </div>

<div class="document">
 <div class="documentwrapper">
 <div class="bodywrapper">
 <div class="body" role="main">
 
 <div class="section" id="what-s-new-in-python-2-7">
<h1>What’s New in Python 2.7<a class="headerlink" href="#what-s-new-in-python-2-7" title="Permalink to this headline">¶</a></h1>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Author:</th><td class="field-body">A.M. Kuchling (amk at amk.ca)</td>
</tr>
</tbody>
</table>
This article explains the new features in Python 2.7. Python 2.7 was released
on July 3, 2010.
Numeric handling has been improved in many ways, for both
floating-point numbers and for the <a class="reference internal" href="../library/decimal.html#decimal.Decimal" title="decimal.Decimal"><code class="xref py py-class docutils literal notranslate">Decimal</code></a> class.
There are some useful additions to the standard library, such as a
greatly enhanced <a class="reference internal" href="../library/unittest.html#module-unittest" title="unittest: Unit testing framework for Python."><code class="xref py py-mod docutils literal notranslate">unittest</code></a> module, the <a class="reference internal" href="../library/argparse.html#module-argparse" title="argparse: Command-line option and argument parsing library."><code class="xref py py-mod docutils literal notranslate">argparse</code></a> module
for parsing command-line options, convenient <a class="reference internal" href="../library/collections.html#collections.OrderedDict" title="collections.OrderedDict"><code class="xref py py-class docutils literal notranslate">OrderedDict</code></a>
and <a class="reference internal" href="../library/collections.html#collections.Counter" title="collections.Counter"><code class="xref py py-class docutils literal notranslate">Counter</code></a> classes in the <a class="reference internal" href="../library/collections.html#module-collections" title="collections: Container datatypes"><code class="xref py py-mod docutils literal notranslate">collections</code></a> module,
and many other improvements.
Python 2.7 is planned to be the last of the 2.x releases, so we worked
on making it a good release for the long term. To help with porting
to Python 3, several new features from the Python 3.x series have been
included in 2.7.
This article doesn’t attempt to provide a complete specification of
the new features, but instead provides a convenient overview. For
full details, you should refer to the documentation for Python 2.7 at
<a class="reference external" href="https://docs.python.org">https://docs.python.org</a>. If you want to understand the rationale for
the design and implementation, refer to the PEP for a particular new
feature or the issue on <a class="reference external" href="https://bugs.python.org">https://bugs.python.org</a> in which a change was
discussed. Whenever possible, “What’s New in Python” links to the
bug/patch item for each change.
<div class="section" id="the-future-for-python-2-x">
<h2>The Future for Python 2.x<a class="headerlink" href="#the-future-for-python-2-x" title="Permalink to this headline">¶</a></h2>
Python 2.7 is the last major release in the 2.x series, as the Python
maintainers have shifted the focus of their new feature development efforts
to the Python 3.x series. This means that while Python 2 continues to
receive bug fixes, and to be updated to build correctly on new hardware and
versions of supported operated systems, there will be no new full feature
releases for the language or standard library.
However, while there is a large common subset between Python 2.7 and Python
3, and many of the changes involved in migrating to that common subset, or
directly to Python 3, can be safely automated, some other changes (notably
those associated with Unicode handling) may require careful consideration,
and preferably robust automated regression test suites, to migrate
effectively.
This means that Python 2.7 will remain in place for a long time, providing a
stable and supported base platform for production systems that have not yet
been ported to Python 3. The full expected lifecycle of the Python 2.7
series is detailed in <a class="pep reference external" href="https://www.python.org/dev/peps/pep-0373">PEP 373</a>.
Some key consequences of the long-term significance of 2.7 are:
<ul class="simple">
<li>As noted above, the 2.7 release has a much longer period of maintenance
when compared to earlier 2.x versions. Python 2.7 is currently expected to
remain supported by the core development team (receiving security updates
and other bug fixes) until at least 2020 (10 years after its initial
release, compared to the more typical support period of 18–24 months).</li>
<li>As the Python 2.7 standard library ages, making effective use of the
Python Package Index (either directly or via a redistributor) becomes
more important for Python 2 users. In addition to a wide variety of third
party packages for various tasks, the available packages include backports
of new modules and features from the Python 3 standard library that are
compatible with Python 2, as well as various tools and libraries that can
make it easier to migrate to Python 3. The <a class="reference external" href="https://packaging.python.org">Python Packaging User Guide</a> provides guidance on downloading and
installing software from the Python Package Index.</li>
<li>While the preferred approach to enhancing Python 2 is now the publication
of new packages on the Python Package Index, this approach doesn’t
necessarily work in all cases, especially those related to network
security. In exceptional cases that cannot be handled adequately by
publishing new or updated packages on PyPI, the Python Enhancement
Proposal process may be used to make the case for adding new features
directly to the Python 2 standard library. Any such additions, and the
maintenance releases where they were added, will be noted in the
<a class="reference internal" href="#py27-maintenance-enhancements">New Features Added to Python 2.7 Maintenance Releases</a> section below.</li>
</ul>
For projects wishing to migrate from Python 2 to Python 3, or for library
and framework developers wishing to support users on both Python 2 and
Python 3, there are a variety of tools and guides available to help decide
on a suitable approach and manage some of the technical details involved.
The recommended starting point is the <a class="reference internal" href="../howto/pyporting.html#pyporting-howto">Porting Python 2 Code to Python 3</a> HOWTO guide.
</div>
<div class="section" id="changes-to-the-handling-of-deprecation-warnings">
<h2>Changes to the Handling of Deprecation Warnings<a class="headerlink" href="#changes-to-the-handling-of-deprecation-warnings" title="Permalink to this headline">¶</a></h2>
For Python 2.7, a policy decision was made to silence warnings only of
interest to developers by default. <a class="reference internal" href="../library/exceptions.html#DeprecationWarning" title="DeprecationWarning"><code class="xref py py-exc docutils literal notranslate">DeprecationWarning</code></a> and its
descendants are now ignored unless otherwise requested, preventing
users from seeing warnings triggered by an application. This change
was also made in the branch that became Python 3.2. (Discussed
on stdlib-sig and carried out in <a class="reference external" href="https://bugs.python.org/issue7319">bpo-7319</a>.)
In previous releases, <a class="reference internal" href="../library/exceptions.html#DeprecationWarning" title="DeprecationWarning"><code class="xref py py-exc docutils literal notranslate">DeprecationWarning</code></a> messages were
enabled by default, providing Python developers with a clear
indication of where their code may break in a future major version
of Python.
However, there are increasingly many users of Python-based
applications who are not directly involved in the development of
those applications. <a class="reference internal" href="../library/exceptions.html#DeprecationWarning" title="DeprecationWarning"><code class="xref py py-exc docutils literal notranslate">DeprecationWarning</code></a> messages are
irrelevant to such users, making them worry about an application
that’s actually working correctly and burdening application developers
with responding to these concerns.
You can re-enable display of <a class="reference internal" href="../library/exceptions.html#DeprecationWarning" title="DeprecationWarning"><code class="xref py py-exc docutils literal notranslate">DeprecationWarning</code></a> messages by
running Python with the <a class="reference internal" href="../using/cmdline.html#cmdoption-w"><code class="xref std std-option docutils literal notranslate">-Wdefault</code></a> (short form:
<a class="reference internal" href="../using/cmdline.html#cmdoption-w"><code class="xref std std-option docutils literal notranslate">-Wd</code></a>) switch, or by setting the <a class="reference internal" href="../using/cmdline.html#envvar-PYTHONWARNINGS"><code class="xref std std-envvar docutils literal notranslate">PYTHONWARNINGS</code></a>
environment variable to <code class="docutils literal notranslate">&quot;default&quot;</code> (or <code class="docutils literal notranslate">&quot;d&quot;</code>) before running
Python. Python code can also re-enable them
by calling <code class="docutils literal notranslate">warnings.simplefilter('default')</code>.
The <code class="docutils literal notranslate">unittest</code> module also automatically reenables deprecation warnings
when running tests.
</div>
<div class="section" id="python-3-1-features">
<h2>Python 3.1 Features<a class="headerlink" href="#python-3-1-features" title="Permalink to this headline">¶</a></h2>
Much as Python 2.6 incorporated features from Python 3.0,
version 2.7 incorporates some of the new features
in Python 3.1. The 2.x series continues to provide tools
for migrating to the 3.x series.
A partial list of 3.1 features that were backported to 2.7:
<ul class="simple">
<li>The syntax for set literals (<code class="docutils literal notranslate">{1,2,3}</code> is a mutable set).</li>
<li>Dictionary and set comprehensions (<code class="docutils literal notranslate">{i: i*2 for i in range(3)}</code>).</li>
<li>Multiple context managers in a single <a class="reference internal" href="../reference/compound_stmts.html#with"><code class="xref std std-keyword docutils literal notranslate">with</code></a> statement.</li>
<li>A new version of the <a class="reference internal" href="../library/io.html#module-io" title="io: Core tools for working with streams."><code class="xref py py-mod docutils literal notranslate">io</code></a> library, rewritten in C for performance.</li>
<li>The ordered-dictionary type described in <a class="reference internal" href="#pep-0372">PEP 372: Adding an Ordered Dictionary to collections</a>.</li>
<li>The new <code class="docutils literal notranslate">&quot;,&quot;</code> format specifier described in <a class="reference internal" href="#pep-0378">PEP 378: Format Specifier for Thousands Separator</a>.</li>
<li>The <a class="reference internal" href="../library/stdtypes.html#memoryview" title="memoryview"><code class="xref py py-class docutils literal notranslate">memoryview</code></a> object.</li>
<li>A small subset of the <a class="reference internal" href="../library/importlib.html#module-importlib" title="importlib: The implementation of the import machinery."><code class="xref py py-mod docutils literal notranslate">importlib</code></a> module,
<a class="reference external" href="#importlib-section">described below</a>.</li>
<li>The <a class="reference internal" href="../library/functions.html#repr" title="repr"><code class="xref py py-func docutils literal notranslate">repr()</code></a> of a float <code class="docutils literal notranslate">x</code> is shorter in many cases: it’s now
based on the shortest decimal string that’s guaranteed to round back
to <code class="docutils literal notranslate">x</code>. As in previous versions of Python, it’s guaranteed that
<code class="docutils literal notranslate">float(repr(x))</code> recovers <code class="docutils literal notranslate">x</code>.</li>
<li>Float-to-string and string-to-float conversions are correctly rounded.
The <a class="reference internal" href="../library/functions.html#round" title="round"><code class="xref py py-func docutils literal notranslate">round()</code></a> function is also now correctly rounded.</li>
<li>The <a class="reference internal" href="../c-api/capsule.html#c.PyCapsule" title="PyCapsule"><code class="xref c c-type docutils literal notranslate">PyCapsule</code></a> type, used to provide a C API for extension modules.</li>
<li>The <a class="reference internal" href="../c-api/long.html#c.PyLong_AsLongAndOverflow" title="PyLong_AsLongAndOverflow"><code class="xref c c-func docutils literal notranslate">PyLong_AsLongAndOverflow()</code></a> C API function.</li>
</ul>
Other new Python3-mode warnings include:
<ul class="simple">
<li><code class="xref py py-func docutils literal notranslate">operator.isCallable()</code> and <code class="xref py py-func docutils literal notranslate">operator.sequenceIncludes()</code>,
which are not supported in 3.x, now trigger warnings.</li>
<li>The <code class="xref std std-option docutils literal notranslate">-3</code> switch now automatically
enables the <code class="xref std std-option docutils literal notranslate">-Qwarn</code> switch that causes warnings
about using classic division with integers and long integers.</li>
</ul>
</div>
<div class="section" id="pep-372-adding-an-ordered-dictionary-to-collections">
<h2>PEP 372: Adding an Ordered Dictionary to collections<a class="headerlink" href="#pep-372-adding-an-ordered-dictionary-to-collections" title="Permalink to this headline">¶</a></h2>
Regular Python dictionaries iterate over key/value pairs in arbitrary order.
Over the years, a number of authors have written alternative implementations
that remember the order that the keys were originally inserted. Based on
the experiences from those implementations, 2.7 introduces a new
<a class="reference internal" href="../library/collections.html#collections.OrderedDict" title="collections.OrderedDict"><code class="xref py py-class docutils literal notranslate">OrderedDict</code></a> class in the <a class="reference internal" href="../library/collections.html#module-collections" title="collections: Container datatypes"><code class="xref py py-mod docutils literal notranslate">collections</code></a> module.
The <a class="reference internal" href="../library/collections.html#collections.OrderedDict" title="collections.OrderedDict"><code class="xref py py-class docutils literal notranslate">OrderedDict</code></a> API provides the same interface as regular
dictionaries but iterates over keys and values in a guaranteed order
depending on when a key was first inserted:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; from collections import OrderedDict
&gt;&gt;&gt; d = OrderedDict([(&#39;first&#39;, 1),
... (&#39;second&#39;, 2),
... (&#39;third&#39;, 3)])
&gt;&gt;&gt; d.items()
[(&#39;first&#39;, 1), (&#39;second&#39;, 2), (&#39;third&#39;, 3)]
</pre></div>
</div>
If a new entry overwrites an existing entry, the original insertion
position is left unchanged:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; d[&#39;second&#39;] = 4
&gt;&gt;&gt; d.items()
[(&#39;first&#39;, 1), (&#39;second&#39;, 4), (&#39;third&#39;, 3)]
</pre></div>
</div>
Deleting an entry and reinserting it will move it to the end:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; del d[&#39;second&#39;]
&gt;&gt;&gt; d[&#39;second&#39;] = 5
&gt;&gt;&gt; d.items()
[(&#39;first&#39;, 1), (&#39;third&#39;, 3), (&#39;second&#39;, 5)]
</pre></div>
</div>
The <a class="reference internal" href="../library/collections.html#collections.OrderedDict.popitem" title="collections.OrderedDict.popitem"><code class="xref py py-meth docutils literal notranslate">popitem()</code></a> method has an optional last
argument that defaults to <code class="docutils literal notranslate">True</code>. If last is true, the most recently
added key is returned and removed; if it’s false, the
oldest key is selected:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; od = OrderedDict([(x,0) for x in range(20)])
&gt;&gt;&gt; od.popitem()
(19, 0)
&gt;&gt;&gt; od.popitem()
(18, 0)
&gt;&gt;&gt; od.popitem(last=False)
(0, 0)
&gt;&gt;&gt; od.popitem(last=False)
(1, 0)
</pre></div>
</div>
Comparing two ordered dictionaries checks both the keys and values,
and requires that the insertion order was the same:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; od1 = OrderedDict([(&#39;first&#39;, 1),
... (&#39;second&#39;, 2),
... (&#39;third&#39;, 3)])
&gt;&gt;&gt; od2 = OrderedDict([(&#39;third&#39;, 3),
... (&#39;first&#39;, 1),
... (&#39;second&#39;, 2)])
&gt;&gt;&gt; od1 == od2
False
&gt;&gt;&gt; # Move &#39;third&#39; key to the end
&gt;&gt;&gt; del od2[&#39;third&#39;]; od2[&#39;third&#39;] = 3
&gt;&gt;&gt; od1 == od2
True
</pre></div>
</div>
Comparing an <a class="reference internal" href="../library/collections.html#collections.OrderedDict" title="collections.OrderedDict"><code class="xref py py-class docutils literal notranslate">OrderedDict</code></a> with a regular dictionary
ignores the insertion order and just compares the keys and values.
How does the <a class="reference internal" href="../library/collections.html#collections.OrderedDict" title="collections.OrderedDict"><code class="xref py py-class docutils literal notranslate">OrderedDict</code></a> work? It maintains a
doubly-linked list of keys, appending new keys to the list as they’re inserted.
A secondary dictionary maps keys to their corresponding list node, so
deletion doesn’t have to traverse the entire linked list and therefore
remains O(1).
The standard library now supports use of ordered dictionaries in several
modules.
<ul class="simple">
<li>The <code class="xref py py-mod docutils literal notranslate">ConfigParser</code> module uses them by default, meaning that
configuration files can now be read, modified, and then written back
in their original order.</li>
<li>The <a class="reference internal" href="../library/collections.html#collections.somenamedtuple._asdict" title="collections.somenamedtuple._asdict"><code class="xref py py-meth docutils literal notranslate">_asdict()</code></a> method for
<a class="reference internal" href="../library/collections.html#collections.namedtuple" title="collections.namedtuple"><code class="xref py py-func docutils literal notranslate">collections.namedtuple()</code></a> now returns an ordered dictionary with the
values appearing in the same order as the underlying tuple indices.</li>
<li>The <a class="reference internal" href="../library/json.html#module-json" title="json: Encode and decode the JSON format."><code class="xref py py-mod docutils literal notranslate">json</code></a> module’s <a class="reference internal" href="../library/json.html#json.JSONDecoder" title="json.JSONDecoder"><code class="xref py py-class docutils literal notranslate">JSONDecoder</code></a> class
constructor was extended with an object_pairs_hook parameter to
allow <code class="xref py py-class docutils literal notranslate">OrderedDict</code> instances to be built by the decoder.
Support was also added for third-party tools like
<a class="reference external" href="http://pyyaml.org/">PyYAML</a>.</li>
</ul>
<div class="admonition seealso">
See also
<dl class="last docutils">
<dt><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0372">PEP 372</a> - Adding an ordered dictionary to collections</dt>
<dd>PEP written by Armin Ronacher and Raymond Hettinger;
implemented by Raymond Hettinger.</dd>
</dl>
</div>
</div>
<div class="section" id="pep-378-format-specifier-for-thousands-separator">
<h2>PEP 378: Format Specifier for Thousands Separator<a class="headerlink" href="#pep-378-format-specifier-for-thousands-separator" title="Permalink to this headline">¶</a></h2>
To make program output more readable, it can be useful to add
separators to large numbers, rendering them as
18,446,744,073,709,551,616 instead of 18446744073709551616.
The fully general solution for doing this is the <a class="reference internal" href="../library/locale.html#module-locale" title="locale: Internationalization services."><code class="xref py py-mod docutils literal notranslate">locale</code></a> module,
which can use different separators (“,” in North America, “.” in
Europe) and different grouping sizes, but <a class="reference internal" href="../library/locale.html#module-locale" title="locale: Internationalization services."><code class="xref py py-mod docutils literal notranslate">locale</code></a> is complicated
to use and unsuitable for multi-threaded applications where different
threads are producing output for different locales.
Therefore, a simple comma-grouping mechanism has been added to the
mini-language used by the <a class="reference internal" href="../library/stdtypes.html#str.format" title="str.format"><code class="xref py py-meth docutils literal notranslate">str.format()</code></a> method. When
formatting a floating-point number, simply include a comma between the
width and the precision:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; &#39;{:20,.2f}&#39;.format(18446744073709551616.0)
&#39;18,446,744,073,709,551,616.00&#39;
</pre></div>
</div>
When formatting an integer, include the comma after the width:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; &#39;{:20,d}&#39;.format(18446744073709551616)
&#39;18,446,744,073,709,551,616&#39;
</pre></div>
</div>
This mechanism is not adaptable at all; commas are always used as the
separator and the grouping is always into three-digit groups. The
comma-formatting mechanism isn’t as general as the <a class="reference internal" href="../library/locale.html#module-locale" title="locale: Internationalization services."><code class="xref py py-mod docutils literal notranslate">locale</code></a>
module, but it’s easier to use.
<div class="admonition seealso">
See also
<dl class="last docutils">
<dt><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0378">PEP 378</a> - Format Specifier for Thousands Separator</dt>
<dd>PEP written by Raymond Hettinger; implemented by Eric Smith.</dd>
</dl>
</div>
</div>
<div class="section" id="pep-389-the-argparse-module-for-parsing-command-lines">
<h2>PEP 389: The argparse Module for Parsing Command Lines<a class="headerlink" href="#pep-389-the-argparse-module-for-parsing-command-lines" title="Permalink to this headline">¶</a></h2>
The <a class="reference internal" href="../library/argparse.html#module-argparse" title="argparse: Command-line option and argument parsing library."><code class="xref py py-mod docutils literal notranslate">argparse</code></a> module for parsing command-line arguments was
added as a more powerful replacement for the
<a class="reference internal" href="../library/optparse.html#module-optparse" title="optparse: Command-line option parsing library. (deprecated)"><code class="xref py py-mod docutils literal notranslate">optparse</code></a> module.
This means Python now supports three different modules for parsing
command-line arguments: <a class="reference internal" href="../library/getopt.html#module-getopt" title="getopt: Portable parser for command line options; support both short and long option names."><code class="xref py py-mod docutils literal notranslate">getopt</code></a>, <a class="reference internal" href="../library/optparse.html#module-optparse" title="optparse: Command-line option parsing library. (deprecated)"><code class="xref py py-mod docutils literal notranslate">optparse</code></a>, and
<a class="reference internal" href="../library/argparse.html#module-argparse" title="argparse: Command-line option and argument parsing library."><code class="xref py py-mod docutils literal notranslate">argparse</code></a>. The <a class="reference internal" href="../library/getopt.html#module-getopt" title="getopt: Portable parser for command line options; support both short and long option names."><code class="xref py py-mod docutils literal notranslate">getopt</code></a> module closely resembles the C
library’s <code class="xref c c-func docutils literal notranslate">getopt()</code> function, so it remains useful if you’re writing a
Python prototype that will eventually be rewritten in C.
<a class="reference internal" href="../library/optparse.html#module-optparse" title="optparse: Command-line option parsing library. (deprecated)"><code class="xref py py-mod docutils literal notranslate">optparse</code></a> becomes redundant, but there are no plans to remove it
because there are many scripts still using it, and there’s no
automated way to update these scripts. (Making the <a class="reference internal" href="../library/argparse.html#module-argparse" title="argparse: Command-line option and argument parsing library."><code class="xref py py-mod docutils literal notranslate">argparse</code></a>
API consistent with <a class="reference internal" href="../library/optparse.html#module-optparse" title="optparse: Command-line option parsing library. (deprecated)"><code class="xref py py-mod docutils literal notranslate">optparse</code></a>’s interface was discussed but
rejected as too messy and difficult.)
In short, if you’re writing a new script and don’t need to worry
about compatibility with earlier versions of Python, use
<a class="reference internal" href="../library/argparse.html#module-argparse" title="argparse: Command-line option and argument parsing library."><code class="xref py py-mod docutils literal notranslate">argparse</code></a> instead of <a class="reference internal" href="../library/optparse.html#module-optparse" title="optparse: Command-line option parsing library. (deprecated)"><code class="xref py py-mod docutils literal notranslate">optparse</code></a>.
Here’s an example:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>import argparse

parser = argparse.ArgumentParser(description=&#39;Command-line example.&#39;)

# Add optional switches
parser.add_argument(&#39;-v&#39;, action=&#39;store_true&#39;, dest=&#39;is_verbose&#39;,
 help=&#39;produce verbose output&#39;)
parser.add_argument(&#39;-o&#39;, action=&#39;store&#39;, dest=&#39;output&#39;,
 metavar=&#39;FILE&#39;,
 help=&#39;direct output to FILE instead of stdout&#39;)
parser.add_argument(&#39;-C&#39;, action=&#39;store&#39;, type=int, dest=&#39;context&#39;,
 metavar=&#39;NUM&#39;, default=0,
 help=&#39;display NUM lines of added context&#39;)

# Allow any number of additional arguments.
parser.add_argument(nargs=&#39;*&#39;, action=&#39;store&#39;, dest=&#39;inputs&#39;,
 help=&#39;input filenames (default is stdin)&#39;)

args = parser.parse_args()
print args.__dict__
</pre></div>
</div>
Unless you override it, <code class="xref std std-option docutils literal notranslate">-h</code> and <code class="xref std std-option docutils literal notranslate">--help</code> switches
are automatically added, and produce neatly formatted output:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>-&gt; ./python.exe argparse-example.py --help
usage: argparse-example.py [-h] [-v] [-o FILE] [-C NUM] [inputs [inputs ...]]

Command-line example.

positional arguments:
 inputs input filenames (default is stdin)

optional arguments:
 -h, --help show this help message and exit
 -v produce verbose output
 -o FILE direct output to FILE instead of stdout
 -C NUM display NUM lines of added context
</pre></div>
</div>
As with <a class="reference internal" href="../library/optparse.html#module-optparse" title="optparse: Command-line option parsing library. (deprecated)"><code class="xref py py-mod docutils literal notranslate">optparse</code></a>, the command-line switches and arguments
are returned as an object with attributes named by the dest parameters:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>-&gt; ./python.exe argparse-example.py -v
{&#39;output&#39;: None,
 &#39;is_verbose&#39;: True,
 &#39;context&#39;: 0,
 &#39;inputs&#39;: []}

-&gt; ./python.exe argparse-example.py -v -o /tmp/output -C 4 file1 file2
{&#39;output&#39;: &#39;/tmp/output&#39;,
 &#39;is_verbose&#39;: True,
 &#39;context&#39;: 4,
 &#39;inputs&#39;: [&#39;file1&#39;, &#39;file2&#39;]}
</pre></div>
</div>
<a class="reference internal" href="../library/argparse.html#module-argparse" title="argparse: Command-line option and argument parsing library."><code class="xref py py-mod docutils literal notranslate">argparse</code></a> has much fancier validation than <a class="reference internal" href="../library/optparse.html#module-optparse" title="optparse: Command-line option parsing library. (deprecated)"><code class="xref py py-mod docutils literal notranslate">optparse</code></a>; you
can specify an exact number of arguments as an integer, 0 or more
arguments by passing <code class="docutils literal notranslate">'*'</code>, 1 or more by passing <code class="docutils literal notranslate">'+'</code>, or an
optional argument with <code class="docutils literal notranslate">'?'</code>. A top-level parser can contain
sub-parsers to define subcommands that have different sets of
switches, as in <code class="docutils literal notranslate">svn commit</code>, <code class="docutils literal notranslate">svn checkout</code>, etc. You can
specify an argument’s type as <a class="reference internal" href="../library/argparse.html#argparse.FileType" title="argparse.FileType"><code class="xref py py-class docutils literal notranslate">FileType</code></a>, which will
automatically open files for you and understands that <code class="docutils literal notranslate">'-'</code> means
standard input or output.
<div class="admonition seealso">
See also
<dl class="last docutils">
<dt><a class="reference internal" href="../library/argparse.html#module-argparse" title="argparse: Command-line option and argument parsing library."><code class="xref py py-mod docutils literal notranslate">argparse</code></a> documentation</dt>
<dd>The documentation page of the argparse module.</dd>
<dt><a class="reference internal" href="../library/argparse.html#upgrading-optparse-code">Upgrading optparse code</a></dt>
<dd>Part of the Python documentation, describing how to convert
code that uses <a class="reference internal" href="../library/optparse.html#module-optparse" title="optparse: Command-line option parsing library. (deprecated)"><code class="xref py py-mod docutils literal notranslate">optparse</code></a>.</dd>
<dt><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0389">PEP 389</a> - argparse - New Command Line Parsing Module</dt>
<dd>PEP written and implemented by Steven Bethard.</dd>
</dl>
</div>
</div>
<div class="section" id="pep-391-dictionary-based-configuration-for-logging">
<h2>PEP 391: Dictionary-Based Configuration For Logging<a class="headerlink" href="#pep-391-dictionary-based-configuration-for-logging" title="Permalink to this headline">¶</a></h2>
The <a class="reference internal" href="../library/logging.html#module-logging" title="logging: Flexible event logging system for applications."><code class="xref py py-mod docutils literal notranslate">logging</code></a> module is very flexible; applications can define
a tree of logging subsystems, and each logger in this tree can filter
out certain messages, format them differently, and direct messages to
a varying number of handlers.
All this flexibility can require a lot of configuration. You can
write Python statements to create objects and set their properties,
but a complex set-up requires verbose but boring code.
<a class="reference internal" href="../library/logging.html#module-logging" title="logging: Flexible event logging system for applications."><code class="xref py py-mod docutils literal notranslate">logging</code></a> also supports a <code class="xref py py-func docutils literal notranslate">fileConfig()</code>
function that parses a file, but the file format doesn’t support
configuring filters, and it’s messier to generate programmatically.
Python 2.7 adds a <code class="xref py py-func docutils literal notranslate">dictConfig()</code> function that
uses a dictionary to configure logging. There are many ways to
produce a dictionary from different sources: construct one with code;
parse a file containing JSON; or use a YAML parsing library if one is
installed. For more information see <a class="reference internal" href="../library/logging.config.html#logging-config-api">Configuration functions</a>.
The following example configures two loggers, the root logger and a
logger named “network”. Messages sent to the root logger will be
sent to the system log using the syslog protocol, and messages
to the “network” logger will be written to a <code class="file docutils literal notranslate">network.log</code> file
that will be rotated once the log reaches 1MB.
<div class="highlight-python3 notranslate"><div class="highlight"><pre>import logging
import logging.config

configdict = {
 &#39;version&#39;: 1, # Configuration schema in use; must be 1 for now
 &#39;formatters&#39;: {
 &#39;standard&#39;: {
 &#39;format&#39;: (&#39;%(asctime)s %(name)-15s &#39;
 &#39;%(levelname)-8s %(message)s&#39;)}},

&#39;handlers&#39;: {&#39;netlog&#39;: {&#39;backupCount&#39;: 10,
 &#39;class&#39;: &#39;logging.handlers.RotatingFileHandler&#39;,
 &#39;filename&#39;: &#39;/logs/network.log&#39;,
 &#39;formatter&#39;: &#39;standard&#39;,
 &#39;level&#39;: &#39;INFO&#39;,
 &#39;maxBytes&#39;: 1000000},
 &#39;syslog&#39;: {&#39;class&#39;: &#39;logging.handlers.SysLogHandler&#39;,
 &#39;formatter&#39;: &#39;standard&#39;,
 &#39;level&#39;: &#39;ERROR&#39;}},

# Specify all the subordinate loggers
 &#39;loggers&#39;: {
 &#39;network&#39;: {
 &#39;handlers&#39;: [&#39;netlog&#39;]
 }
 },
 # Specify properties of the root logger
 &#39;root&#39;: {
 &#39;handlers&#39;: [&#39;syslog&#39;]
 },
}

# Set up configuration
logging.config.dictConfig(configdict)

# As an example, log two error messages
logger = logging.getLogger(&#39;/&#39;)
logger.error(&#39;Database not found&#39;)

netlogger = logging.getLogger(&#39;network&#39;)
netlogger.error(&#39;Connection failed&#39;)
</pre></div>
</div>
Three smaller enhancements to the <a class="reference internal" href="../library/logging.html#module-logging" title="logging: Flexible event logging system for applications."><code class="xref py py-mod docutils literal notranslate">logging</code></a> module, all
implemented by Vinay Sajip, are:
<ul class="simple">
<li>The <a class="reference internal" href="../library/logging.handlers.html#logging.handlers.SysLogHandler" title="logging.handlers.SysLogHandler"><code class="xref py py-class docutils literal notranslate">SysLogHandler</code></a> class now supports
syslogging over TCP. The constructor has a socktype parameter
giving the type of socket to use, either <a class="reference internal" href="../library/socket.html#socket.SOCK_DGRAM" title="socket.SOCK_DGRAM"><code class="xref py py-const docutils literal notranslate">socket.SOCK_DGRAM</code></a>
for UDP or <a class="reference internal" href="../library/socket.html#socket.SOCK_STREAM" title="socket.SOCK_STREAM"><code class="xref py py-const docutils literal notranslate">socket.SOCK_STREAM</code></a> for TCP. The default
protocol remains UDP.</li>
<li><a class="reference internal" href="../library/logging.html#logging.Logger" title="logging.Logger"><code class="xref py py-class docutils literal notranslate">Logger</code></a> instances gained a <a class="reference internal" href="../library/logging.html#logging.Logger.getChild" title="logging.Logger.getChild"><code class="xref py py-meth docutils literal notranslate">getChild()</code></a>
method that retrieves a descendant logger using a relative path.
For example, once you retrieve a logger by doing <code class="docutils literal notranslate">log = getLogger('app')</code>,
calling <code class="docutils literal notranslate">log.getChild('network.listen')</code> is equivalent to
<code class="docutils literal notranslate">getLogger('app.network.listen')</code>.</li>
<li>The <a class="reference internal" href="../library/logging.html#logging.LoggerAdapter" title="logging.LoggerAdapter"><code class="xref py py-class docutils literal notranslate">LoggerAdapter</code></a> class gained an
<code class="xref py py-meth docutils literal notranslate">isEnabledFor()</code> method that takes a
level and returns whether the underlying logger would
process a message of that level of importance.</li>
</ul>
<div class="admonition seealso">
See also
<dl class="last docutils">
<dt><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0391">PEP 391</a> - Dictionary-Based Configuration For Logging</dt>
<dd>PEP written and implemented by Vinay Sajip.</dd>
</dl>
</div>
</div>
<div class="section" id="pep-3106-dictionary-views">
<h2>PEP 3106: Dictionary Views<a class="headerlink" href="#pep-3106-dictionary-views" title="Permalink to this headline">¶</a></h2>
The dictionary methods <a class="reference internal" href="../library/stdtypes.html#dict.keys" title="dict.keys"><code class="xref py py-meth docutils literal notranslate">keys()</code></a>, <a class="reference internal" href="../library/stdtypes.html#dict.values" title="dict.values"><code class="xref py py-meth docutils literal notranslate">values()</code></a>, and
<a class="reference internal" href="../library/stdtypes.html#dict.items" title="dict.items"><code class="xref py py-meth docutils literal notranslate">items()</code></a> are different in Python 3.x. They return an object
called a view instead of a fully materialized list.
It’s not possible to change the return values of <a class="reference internal" href="../library/stdtypes.html#dict.keys" title="dict.keys"><code class="xref py py-meth docutils literal notranslate">keys()</code></a>,
<a class="reference internal" href="../library/stdtypes.html#dict.values" title="dict.values"><code class="xref py py-meth docutils literal notranslate">values()</code></a>, and <a class="reference internal" href="../library/stdtypes.html#dict.items" title="dict.items"><code class="xref py py-meth docutils literal notranslate">items()</code></a> in Python 2.7 because
too much code would break. Instead the 3.x versions were added
under the new names <code class="xref py py-meth docutils literal notranslate">viewkeys()</code>, <code class="xref py py-meth docutils literal notranslate">viewvalues()</code>,
and <code class="xref py py-meth docutils literal notranslate">viewitems()</code>.
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; d = dict((i*10, chr(65+i)) for i in range(26))
&gt;&gt;&gt; d
{0: &#39;A&#39;, 130: &#39;N&#39;, 10: &#39;B&#39;, 140: &#39;O&#39;, 20: ..., 250: &#39;Z&#39;}
&gt;&gt;&gt; d.viewkeys()
dict_keys([0, 130, 10, 140, 20, 150, 30, ..., 250])
</pre></div>
</div>
Views can be iterated over, but the key and item views also behave
like sets. The <code class="docutils literal notranslate">&amp;</code> operator performs intersection, and <code class="docutils literal notranslate">|</code>
performs a union:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; d1 = dict((i*10, chr(65+i)) for i in range(26))
&gt;&gt;&gt; d2 = dict((i**.5, i) for i in range(1000))
&gt;&gt;&gt; d1.viewkeys() &amp; d2.viewkeys()
set([0.0, 10.0, 20.0, 30.0])
&gt;&gt;&gt; d1.viewkeys() | range(0, 30)
set([0, 1, 130, 3, 4, 5, 6, ..., 120, 250])
</pre></div>
</div>
The view keeps track of the dictionary and its contents change as the
dictionary is modified:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; vk = d.viewkeys()
&gt;&gt;&gt; vk
dict_keys([0, 130, 10, ..., 250])
&gt;&gt;&gt; d[260] = &#39;&amp;&#39;
&gt;&gt;&gt; vk
dict_keys([0, 130, 260, 10, ..., 250])
</pre></div>
</div>
However, note that you can’t add or remove keys while you’re iterating
over the view:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; for k in vk:
... d[k*2] = k
...
Traceback (most recent call last):
 File &quot;&lt;stdin&gt;&quot;, line 1, in &lt;module&gt;
RuntimeError: dictionary changed size during iteration
</pre></div>
</div>
You can use the view methods in Python 2.x code, and the 2to3
converter will change them to the standard <a class="reference internal" href="../library/stdtypes.html#dict.keys" title="dict.keys"><code class="xref py py-meth docutils literal notranslate">keys()</code></a>,
<a class="reference internal" href="../library/stdtypes.html#dict.values" title="dict.values"><code class="xref py py-meth docutils literal notranslate">values()</code></a>, and <a class="reference internal" href="../library/stdtypes.html#dict.items" title="dict.items"><code class="xref py py-meth docutils literal notranslate">items()</code></a> methods.
<div class="admonition seealso">
See also
<dl class="last docutils">
<dt><a class="pep reference external" href="https://www.python.org/dev/peps/pep-3106">PEP 3106</a> - Revamping dict.keys(), .values() and .items()</dt>
<dd>PEP written by Guido van Rossum.
Backported to 2.7 by Alexandre Vassalotti; <a class="reference external" href="https://bugs.python.org/issue1967">bpo-1967</a>.</dd>
</dl>
</div>
</div>
<div class="section" id="pep-3137-the-memoryview-object">
<h2>PEP 3137: The memoryview Object<a class="headerlink" href="#pep-3137-the-memoryview-object" title="Permalink to this headline">¶</a></h2>
The <a class="reference internal" href="../library/stdtypes.html#memoryview" title="memoryview"><code class="xref py py-class docutils literal notranslate">memoryview</code></a> object provides a view of another object’s
memory content that matches the <a class="reference internal" href="../library/stdtypes.html#bytes" title="bytes"><code class="xref py py-class docutils literal notranslate">bytes</code></a> type’s interface.
<div class="highlight-pycon notranslate"><div class="highlight"><pre>&gt;&gt;&gt; import string
&gt;&gt;&gt; m = memoryview(string.letters)
&gt;&gt;&gt; m
&lt;memory at 0x37f850&gt;
&gt;&gt;&gt; len(m) # Returns length of underlying object
52
&gt;&gt;&gt; m[0], m[25], m[26] # Indexing returns one byte
(&#39;a&#39;, &#39;z&#39;, &#39;A&#39;)
&gt;&gt;&gt; m2 = m[0:26] # Slicing returns another memoryview
&gt;&gt;&gt; m2
&lt;memory at 0x37f080&gt;
</pre></div>
</div>
The content of the view can be converted to a string of bytes or
a list of integers:
<div class="highlight-pycon notranslate"><div class="highlight"><pre>&gt;&gt;&gt; m2.tobytes()
&#39;abcdefghijklmnopqrstuvwxyz&#39;
&gt;&gt;&gt; m2.tolist()
[97, 98, 99, 100, 101, 102, 103, ... 121, 122]
&gt;&gt;&gt;
</pre></div>
</div>
<a class="reference internal" href="../library/stdtypes.html#memoryview" title="memoryview"><code class="xref py py-class docutils literal notranslate">memoryview</code></a> objects allow modifying the underlying object if
it’s a mutable object.
<div class="highlight-pycon notranslate"><div class="highlight"><pre>&gt;&gt;&gt; m2[0] = 75
Traceback (most recent call last):
 File &quot;&lt;stdin&gt;&quot;, line 1, in &lt;module&gt;
TypeError: cannot modify read-only memory
&gt;&gt;&gt; b = bytearray(string.letters) # Creating a mutable object
&gt;&gt;&gt; b
bytearray(b&#39;abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ&#39;)
&gt;&gt;&gt; mb = memoryview(b)
&gt;&gt;&gt; mb[0] = &#39;*&#39; # Assign to view, changing the bytearray.
&gt;&gt;&gt; b[0:5] # The bytearray has been changed.
bytearray(b&#39;*bcde&#39;)
&gt;&gt;&gt;
</pre></div>
</div>
<div class="admonition seealso">
See also
<dl class="last docutils">
<dt><a class="pep reference external" href="https://www.python.org/dev/peps/pep-3137">PEP 3137</a> - Immutable Bytes and Mutable Buffer</dt>
<dd>PEP written by Guido van Rossum.
Implemented by Travis Oliphant, Antoine Pitrou and others.
Backported to 2.7 by Antoine Pitrou; <a class="reference external" href="https://bugs.python.org/issue2396">bpo-2396</a>.</dd>
</dl>
</div>
</div>
<div class="section" id="other-language-changes">
<h2>Other Language Changes<a class="headerlink" href="#other-language-changes" title="Permalink to this headline">¶</a></h2>
Some smaller changes made to the core Python language are:
<ul>
<li>The syntax for set literals has been backported from Python 3.x.
Curly brackets are used to surround the contents of the resulting
mutable set; set literals are
distinguished from dictionaries by not containing colons and values.
<code class="docutils literal notranslate">{}</code> continues to represent an empty dictionary; use
<code class="docutils literal notranslate">set()</code> for an empty set.
<div class="highlight-pycon notranslate"><div class="highlight"><pre>&gt;&gt;&gt; {1, 2, 3, 4, 5}
set([1, 2, 3, 4, 5])
&gt;&gt;&gt; set() # empty set
set([])
&gt;&gt;&gt; {} # empty dict
{}
</pre></div>
</div>
Backported by Alexandre Vassalotti; <a class="reference external" href="https://bugs.python.org/issue2335">bpo-2335</a>.
</li>
<li>Dictionary and set comprehensions are another feature backported from
3.x, generalizing list/generator comprehensions to use
the literal syntax for sets and dictionaries.
<div class="highlight-pycon notranslate"><div class="highlight"><pre>&gt;&gt;&gt; {x: x*x for x in range(6)}
{0: 0, 1: 1, 2: 4, 3: 9, 4: 16, 5: 25}
&gt;&gt;&gt; {(&#39;a&#39;*x) for x in range(6)}
set([&#39;&#39;, &#39;a&#39;, &#39;aa&#39;, &#39;aaa&#39;, &#39;aaaa&#39;, &#39;aaaaa&#39;])
</pre></div>
</div>
Backported by Alexandre Vassalotti; <a class="reference external" href="https://bugs.python.org/issue2333">bpo-2333</a>.
</li>
<li>The <a class="reference internal" href="../reference/compound_stmts.html#with"><code class="xref std std-keyword docutils literal notranslate">with</code></a> statement can now use multiple context managers
in one statement. Context managers are processed from left to right
and each one is treated as beginning a new <a class="reference internal" href="../reference/compound_stmts.html#with"><code class="xref std std-keyword docutils literal notranslate">with</code></a> statement.
This means that:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>with A() as a, B() as b:
 ... suite of statements ...
</pre></div>
</div>
is equivalent to:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>with A() as a:
 with B() as b:
 ... suite of statements ...
</pre></div>
</div>
The <code class="xref py py-func docutils literal notranslate">contextlib.nested()</code> function provides a very similar
function, so it’s no longer necessary and has been deprecated.
(Proposed in <a class="reference external" href="https://codereview.appspot.com/53094">https://codereview.appspot.com/53094</a>; implemented by
Georg Brandl.)
</li>
<li>Conversions between floating-point numbers and strings are
now correctly rounded on most platforms. These conversions occur
in many different places: <a class="reference internal" href="../library/stdtypes.html#str" title="str"><code class="xref py py-func docutils literal notranslate">str()</code></a> on
floats and complex numbers; the <a class="reference internal" href="../library/functions.html#float" title="float"><code class="xref py py-class docutils literal notranslate">float</code></a> and <a class="reference internal" href="../library/functions.html#complex" title="complex"><code class="xref py py-class docutils literal notranslate">complex</code></a>
constructors;
numeric formatting; serializing and
deserializing floats and complex numbers using the
<a class="reference internal" href="../library/marshal.html#module-marshal" title="marshal: Convert Python objects to streams of bytes and back (with different constraints)."><code class="xref py py-mod docutils literal notranslate">marshal</code></a>, <a class="reference internal" href="../library/pickle.html#module-pickle" title="pickle: Convert Python objects to streams of bytes and back."><code class="xref py py-mod docutils literal notranslate">pickle</code></a>
and <a class="reference internal" href="../library/json.html#module-json" title="json: Encode and decode the JSON format."><code class="xref py py-mod docutils literal notranslate">json</code></a> modules;
parsing of float and imaginary literals in Python code;
and <a class="reference internal" href="../library/decimal.html#decimal.Decimal" title="decimal.Decimal"><code class="xref py py-class docutils literal notranslate">Decimal</code></a>-to-float conversion.
Related to this, the <a class="reference internal" href="../library/functions.html#repr" title="repr"><code class="xref py py-func docutils literal notranslate">repr()</code></a> of a floating-point number x
now returns a result based on the shortest decimal string that’s
guaranteed to round back to x under correct rounding (with
round-half-to-even rounding mode). Previously it gave a string
based on rounding x to 17 decimal digits.
The rounding library responsible for this improvement works on
Windows and on Unix platforms using the gcc, icc, or suncc
compilers. There may be a small number of platforms where correct
operation of this code cannot be guaranteed, so the code is not
used on such systems. You can find out which code is being used
by checking <a class="reference internal" href="../library/sys.html#sys.float_repr_style" title="sys.float_repr_style"><code class="xref py py-data docutils literal notranslate">sys.float_repr_style</code></a>, which will be <code class="docutils literal notranslate">short</code>
if the new code is in use and <code class="docutils literal notranslate">legacy</code> if it isn’t.
Implemented by Eric Smith and Mark Dickinson, using David Gay’s
<code class="file docutils literal notranslate">dtoa.c</code> library; <a class="reference external" href="https://bugs.python.org/issue7117">bpo-7117</a>.
</li>
<li>Conversions from long integers and regular integers to floating
point now round differently, returning the floating-point number
closest to the number. This doesn’t matter for small integers that
can be converted exactly, but for large numbers that will
unavoidably lose precision, Python 2.7 now approximates more
closely. For example, Python 2.6 computed the following:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; n = 295147905179352891391
&gt;&gt;&gt; float(n)
2.9514790517935283e+20
&gt;&gt;&gt; n - long(float(n))
65535L
</pre></div>
</div>
Python 2.7’s floating-point result is larger, but much closer to the
true value:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; n = 295147905179352891391
&gt;&gt;&gt; float(n)
2.9514790517935289e+20
&gt;&gt;&gt; n - long(float(n))
-1L
</pre></div>
</div>
(Implemented by Mark Dickinson; <a class="reference external" href="https://bugs.python.org/issue3166">bpo-3166</a>.)
Integer division is also more accurate in its rounding behaviours. (Also
implemented by Mark Dickinson; <a class="reference external" href="https://bugs.python.org/issue1811">bpo-1811</a>.)
</li>
<li>Implicit coercion for complex numbers has been removed; the interpreter
will no longer ever attempt to call a <code class="xref py py-meth docutils literal notranslate">__coerce__()</code> method on complex
objects. (Removed by Meador Inge and Mark Dickinson; <a class="reference external" href="https://bugs.python.org/issue5211">bpo-5211</a>.)
</li>
<li>The <a class="reference internal" href="../library/stdtypes.html#str.format" title="str.format"><code class="xref py py-meth docutils literal notranslate">str.format()</code></a> method now supports automatic numbering of the replacement
fields. This makes using <a class="reference internal" href="../library/stdtypes.html#str.format" title="str.format"><code class="xref py py-meth docutils literal notranslate">str.format()</code></a> more closely resemble using
<code class="docutils literal notranslate">%s</code> formatting:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; &#39;{}:{}:{}&#39;.format(2009, 04, &#39;Sunday&#39;)
&#39;2009:4:Sunday&#39;
&gt;&gt;&gt; &#39;{}:{}:{day}&#39;.format(2009, 4, day=&#39;Sunday&#39;)
&#39;2009:4:Sunday&#39;
</pre></div>
</div>
The auto-numbering takes the fields from left to right, so the first <code class="docutils literal notranslate">{...}</code>
specifier will use the first argument to <a class="reference internal" href="../library/stdtypes.html#str.format" title="str.format"><code class="xref py py-meth docutils literal notranslate">str.format()</code></a>, the next
specifier will use the next argument, and so on. You can’t mix auto-numbering
and explicit numbering – either number all of your specifier fields or none
of them – but you can mix auto-numbering and named fields, as in the second
example above. (Contributed by Eric Smith; <a class="reference external" href="https://bugs.python.org/issue5237">bpo-5237</a>.)
Complex numbers now correctly support usage with <a class="reference internal" href="../library/functions.html#format" title="format"><code class="xref py py-func docutils literal notranslate">format()</code></a>,
and default to being right-aligned.
Specifying a precision or comma-separation applies to both the real
and imaginary parts of the number, but a specified field width and
alignment is applied to the whole of the resulting <code class="docutils literal notranslate">1.5+3j</code>
output. (Contributed by Eric Smith; <a class="reference external" href="https://bugs.python.org/issue1588">bpo-1588</a> and <a class="reference external" href="https://bugs.python.org/issue7988">bpo-7988</a>.)
The ‘F’ format code now always formats its output using uppercase characters,
so it will now produce ‘INF’ and ‘NAN’.
(Contributed by Eric Smith; <a class="reference external" href="https://bugs.python.org/issue3382">bpo-3382</a>.)
A low-level change: the <a class="reference internal" href="../reference/datamodel.html#object.__format__" title="object.__format__"><code class="xref py py-meth docutils literal notranslate">object.__format__()</code></a> method now triggers
a <a class="reference internal" href="../library/exceptions.html#PendingDeprecationWarning" title="PendingDeprecationWarning"><code class="xref py py-exc docutils literal notranslate">PendingDeprecationWarning</code></a> if it’s passed a format string,
because the <a class="reference internal" href="../reference/datamodel.html#object.__format__" title="object.__format__"><code class="xref py py-meth docutils literal notranslate">__format__()</code></a> method for <a class="reference internal" href="../library/functions.html#object" title="object"><code class="xref py py-class docutils literal notranslate">object</code></a> converts
the object to a string representation and formats that. Previously
the method silently applied the format string to the string
representation, but that could hide mistakes in Python code. If
you’re supplying formatting information such as an alignment or
precision, presumably you’re expecting the formatting to be applied
in some object-specific way. (Fixed by Eric Smith; <a class="reference external" href="https://bugs.python.org/issue7994">bpo-7994</a>.)
</li>
<li>The <a class="reference internal" href="../library/functions.html#int" title="int"><code class="xref py py-func docutils literal notranslate">int()</code></a> and <code class="xref py py-func docutils literal notranslate">long()</code> types gained a <code class="docutils literal notranslate">bit_length</code>
method that returns the number of bits necessary to represent
its argument in binary:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; n = 37
&gt;&gt;&gt; bin(n)
&#39;0b100101&#39;
&gt;&gt;&gt; n.bit_length()
6
&gt;&gt;&gt; n = 2**123-1
&gt;&gt;&gt; n.bit_length()
123
&gt;&gt;&gt; (n+1).bit_length()
124
</pre></div>
</div>
(Contributed by Fredrik Johansson and Victor Stinner; <a class="reference external" href="https://bugs.python.org/issue3439">bpo-3439</a>.)
</li>
<li>The <a class="reference internal" href="../reference/simple_stmts.html#import"><code class="xref std std-keyword docutils literal notranslate">import</code></a> statement will no longer try an absolute import
if a relative import (e.g. <code class="docutils literal notranslate">from .os import sep</code>) fails. This
fixes a bug, but could possibly break certain <a class="reference internal" href="../reference/simple_stmts.html#import"><code class="xref std std-keyword docutils literal notranslate">import</code></a>
statements that were only working by accident. (Fixed by Meador Inge;
<a class="reference external" href="https://bugs.python.org/issue7902">bpo-7902</a>.)
</li>
<li>It’s now possible for a subclass of the built-in <code class="xref py py-class docutils literal notranslate">unicode</code> type
to override the <code class="xref py py-meth docutils literal notranslate">__unicode__()</code> method. (Implemented by
Victor Stinner; <a class="reference external" href="https://bugs.python.org/issue1583863">bpo-1583863</a>.)
</li>
<li>The <a class="reference internal" href="../library/stdtypes.html#bytearray" title="bytearray"><code class="xref py py-class docutils literal notranslate">bytearray</code></a> type’s <a class="reference internal" href="../library/stdtypes.html#bytearray.translate" title="bytearray.translate"><code class="xref py py-meth docutils literal notranslate">translate()</code></a> method now accepts
<code class="docutils literal notranslate">None</code> as its first argument. (Fixed by Georg Brandl;
<a class="reference external" href="https://bugs.python.org/issue4759">bpo-4759</a>.)
</li>
<li>When using <code class="docutils literal notranslate">&#64;classmethod</code> and <code class="docutils literal notranslate">&#64;staticmethod</code> to wrap
methods as class or static methods, the wrapper object now
exposes the wrapped function as their <code class="xref py py-attr docutils literal notranslate">__func__</code> attribute.
(Contributed by Amaury Forgeot d’Arc, after a suggestion by
George Sakkis; <a class="reference external" href="https://bugs.python.org/issue5982">bpo-5982</a>.)
</li>
<li>When a restricted set of attributes were set using <code class="docutils literal notranslate">__slots__</code>,
deleting an unset attribute would not raise <a class="reference internal" href="../library/exceptions.html#AttributeError" title="AttributeError"><code class="xref py py-exc docutils literal notranslate">AttributeError</code></a>
as you would expect. Fixed by Benjamin Peterson; <a class="reference external" href="https://bugs.python.org/issue7604">bpo-7604</a>.)
</li>
<li>Two new encodings are now supported: “cp720”, used primarily for
Arabic text; and “cp858”, a variant of CP 850 that adds the euro
symbol. (CP720 contributed by Alexander Belchenko and Amaury
Forgeot d’Arc in <a class="reference external" href="https://bugs.python.org/issue1616979">bpo-1616979</a>; CP858 contributed by Tim Hatch in
<a class="reference external" href="https://bugs.python.org/issue8016">bpo-8016</a>.)
</li>
<li>The <code class="xref py py-class docutils literal notranslate">file</code> object will now set the <code class="xref py py-attr docutils literal notranslate">filename</code> attribute
on the <a class="reference internal" href="../library/exceptions.html#IOError" title="IOError"><code class="xref py py-exc docutils literal notranslate">IOError</code></a> exception when trying to open a directory
on POSIX platforms (noted by Jan Kaliszewski; <a class="reference external" href="https://bugs.python.org/issue4764">bpo-4764</a>), and
now explicitly checks for and forbids writing to read-only file objects
instead of trusting the C library to catch and report the error
(fixed by Stefan Krah; <a class="reference external" href="https://bugs.python.org/issue5677">bpo-5677</a>).
</li>
<li>The Python tokenizer now translates line endings itself, so the
<a class="reference internal" href="../library/functions.html#compile" title="compile"><code class="xref py py-func docutils literal notranslate">compile()</code></a> built-in function now accepts code using any
line-ending convention. Additionally, it no longer requires that the
code end in a newline.
</li>
<li>Extra parentheses in function definitions are illegal in Python 3.x,
meaning that you get a syntax error from <code class="docutils literal notranslate">def f((x)): pass</code>. In
Python3-warning mode, Python 2.7 will now warn about this odd usage.
(Noted by James Lingard; <a class="reference external" href="https://bugs.python.org/issue7362">bpo-7362</a>.)
</li>
<li>It’s now possible to create weak references to old-style class
objects. New-style classes were always weak-referenceable. (Fixed
by Antoine Pitrou; <a class="reference external" href="https://bugs.python.org/issue8268">bpo-8268</a>.)
</li>
<li>When a module object is garbage-collected, the module’s dictionary is
now only cleared if no one else is holding a reference to the
dictionary (<a class="reference external" href="https://bugs.python.org/issue7140">bpo-7140</a>).
</li>
</ul>
<div class="section" id="interpreter-changes">
<h3>Interpreter Changes<a class="headerlink" href="#interpreter-changes" title="Permalink to this headline">¶</a></h3>
A new environment variable, <a class="reference internal" href="../using/cmdline.html#envvar-PYTHONWARNINGS"><code class="xref std std-envvar docutils literal notranslate">PYTHONWARNINGS</code></a>,
allows controlling warnings. It should be set to a string
containing warning settings, equivalent to those
used with the <a class="reference internal" href="../using/cmdline.html#cmdoption-w"><code class="xref std std-option docutils literal notranslate">-W</code></a> switch, separated by commas.
(Contributed by Brian Curtin; <a class="reference external" href="https://bugs.python.org/issue7301">bpo-7301</a>.)
For example, the following setting will print warnings every time
they occur, but turn warnings from the <code class="xref py py-mod docutils literal notranslate">Cookie</code> module into an
error. (The exact syntax for setting an environment variable varies
across operating systems and shells.)
<div class="highlight-python3 notranslate"><div class="highlight"><pre>export PYTHONWARNINGS=all,error:::Cookie:0
</pre></div>
</div>
</div>
<div class="section" id="optimizations">
<h3>Optimizations<a class="headerlink" href="#optimizations" title="Permalink to this headline">¶</a></h3>
Several performance enhancements have been added:
<ul>
<li>A new opcode was added to perform the initial setup for
<a class="reference internal" href="../reference/compound_stmts.html#with"><code class="xref std std-keyword docutils literal notranslate">with</code></a> statements, looking up the <a class="reference internal" href="../reference/datamodel.html#object.__enter__" title="object.__enter__"><code class="xref py py-meth docutils literal notranslate">__enter__()</code></a> and
<a class="reference internal" href="../reference/datamodel.html#object.__exit__" title="object.__exit__"><code class="xref py py-meth docutils literal notranslate">__exit__()</code></a> methods. (Contributed by Benjamin Peterson.)
</li>
<li>The garbage collector now performs better for one common usage
pattern: when many objects are being allocated without deallocating
any of them. This would previously take quadratic
time for garbage collection, but now the number of full garbage collections
is reduced as the number of objects on the heap grows.
The new logic only performs a full garbage collection pass when
the middle generation has been collected 10 times and when the
number of survivor objects from the middle generation exceeds 10% of
the number of objects in the oldest generation. (Suggested by Martin
von Löwis and implemented by Antoine Pitrou; <a class="reference external" href="https://bugs.python.org/issue4074">bpo-4074</a>.)
</li>
<li>The garbage collector tries to avoid tracking simple containers
which can’t be part of a cycle. In Python 2.7, this is now true for
tuples and dicts containing atomic types (such as ints, strings,
etc.). Transitively, a dict containing tuples of atomic types won’t
be tracked either. This helps reduce the cost of each
garbage collection by decreasing the number of objects to be
considered and traversed by the collector.
(Contributed by Antoine Pitrou; <a class="reference external" href="https://bugs.python.org/issue4688">bpo-4688</a>.)
</li>
<li>Long integers are now stored internally either in base 2**15 or in base
2**30, the base being determined at build time. Previously, they
were always stored in base 2**15. Using base 2**30 gives
significant performance improvements on 64-bit machines, but
benchmark results on 32-bit machines have been mixed. Therefore,
the default is to use base 2**30 on 64-bit machines and base 2**15
on 32-bit machines; on Unix, there’s a new configure option
<code class="xref std std-option docutils literal notranslate">--enable-big-digits</code> that can be used to override this default.
Apart from the performance improvements this change should be
invisible to end users, with one exception: for testing and
debugging purposes there’s a new structseq <code class="xref py py-data docutils literal notranslate">sys.long_info</code> that
provides information about the internal format, giving the number of
bits per digit and the size in bytes of the C type used to store
each digit:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; import sys
&gt;&gt;&gt; sys.long_info
sys.long_info(bits_per_digit=30, sizeof_digit=4)
</pre></div>
</div>
(Contributed by Mark Dickinson; <a class="reference external" href="https://bugs.python.org/issue4258">bpo-4258</a>.)
Another set of changes made long objects a few bytes smaller: 2 bytes
smaller on 32-bit systems and 6 bytes on 64-bit.
(Contributed by Mark Dickinson; <a class="reference external" href="https://bugs.python.org/issue5260">bpo-5260</a>.)
</li>
<li>The division algorithm for long integers has been made faster
by tightening the inner loop, doing shifts instead of multiplications,
and fixing an unnecessary extra iteration.
Various benchmarks show speedups of between 50% and 150% for long
integer divisions and modulo operations.
(Contributed by Mark Dickinson; <a class="reference external" href="https://bugs.python.org/issue5512">bpo-5512</a>.)
Bitwise operations are also significantly faster (initial patch by
Gregory Smith; <a class="reference external" href="https://bugs.python.org/issue1087418">bpo-1087418</a>).
</li>
<li>The implementation of <code class="docutils literal notranslate">%</code> checks for the left-side operand being
a Python string and special-cases it; this results in a 1–3%
performance increase for applications that frequently use <code class="docutils literal notranslate">%</code>
with strings, such as templating libraries.
(Implemented by Collin Winter; <a class="reference external" href="https://bugs.python.org/issue5176">bpo-5176</a>.)
</li>
<li>List comprehensions with an <code class="docutils literal notranslate">if</code> condition are compiled into
faster bytecode. (Patch by Antoine Pitrou, back-ported to 2.7
by Jeffrey Yasskin; <a class="reference external" href="https://bugs.python.org/issue4715">bpo-4715</a>.)
</li>
<li>Converting an integer or long integer to a decimal string was made
faster by special-casing base 10 instead of using a generalized
conversion function that supports arbitrary bases.
(Patch by Gawain Bolton; <a class="reference external" href="https://bugs.python.org/issue6713">bpo-6713</a>.)
</li>
<li>The <code class="xref py py-meth docutils literal notranslate">split()</code>, <code class="xref py py-meth docutils literal notranslate">replace()</code>, <code class="xref py py-meth docutils literal notranslate">rindex()</code>,
<code class="xref py py-meth docutils literal notranslate">rpartition()</code>, and <code class="xref py py-meth docutils literal notranslate">rsplit()</code> methods of string-like types
(strings, Unicode strings, and <a class="reference internal" href="../library/stdtypes.html#bytearray" title="bytearray"><code class="xref py py-class docutils literal notranslate">bytearray</code></a> objects) now use a
fast reverse-search algorithm instead of a character-by-character
scan. This is sometimes faster by a factor of 10. (Added by
Florent Xicluna; <a class="reference external" href="https://bugs.python.org/issue7462">bpo-7462</a> and <a class="reference external" href="https://bugs.python.org/issue7622">bpo-7622</a>.)
</li>
<li>The <a class="reference internal" href="../library/pickle.html#module-pickle" title="pickle: Convert Python objects to streams of bytes and back."><code class="xref py py-mod docutils literal notranslate">pickle</code></a> and <code class="xref py py-mod docutils literal notranslate">cPickle</code> modules now automatically
intern the strings used for attribute names, reducing memory usage
of the objects resulting from unpickling. (Contributed by Jake
McGuire; <a class="reference external" href="https://bugs.python.org/issue5084">bpo-5084</a>.)
</li>
<li>The <code class="xref py py-mod docutils literal notranslate">cPickle</code> module now special-cases dictionaries,
nearly halving the time required to pickle them.
(Contributed by Collin Winter; <a class="reference external" href="https://bugs.python.org/issue5670">bpo-5670</a>.)
</li>
</ul>
</div>
</div>
<div class="section" id="new-and-improved-modules">
<h2>New and Improved Modules<a class="headerlink" href="#new-and-improved-modules" title="Permalink to this headline">¶</a></h2>
As in every release, Python’s standard library received a number of
enhancements and bug fixes. Here’s a partial list of the most notable
changes, sorted alphabetically by module name. Consult the
<code class="file docutils literal notranslate">Misc/NEWS</code> file in the source tree for a more complete list of
changes, or look through the Subversion logs for all the details.
<ul>
<li>The <a class="reference internal" href="../library/bdb.html#module-bdb" title="bdb: Debugger framework."><code class="xref py py-mod docutils literal notranslate">bdb</code></a> module’s base debugging class <a class="reference internal" href="../library/bdb.html#bdb.Bdb" title="bdb.Bdb"><code class="xref py py-class docutils literal notranslate">Bdb</code></a>
gained a feature for skipping modules. The constructor
now takes an iterable containing glob-style patterns such as
<code class="docutils literal notranslate">django.*</code>; the debugger will not step into stack frames
from a module that matches one of these patterns.
(Contributed by Maru Newby after a suggestion by
Senthil Kumaran; <a class="reference external" href="https://bugs.python.org/issue5142">bpo-5142</a>.)
</li>
<li>The <a class="reference internal" href="../library/binascii.html#module-binascii" title="binascii: Tools for converting between binary and various ASCII-encoded binary representations."><code class="xref py py-mod docutils literal notranslate">binascii</code></a> module now supports the buffer API, so it can be
used with <a class="reference internal" href="../library/stdtypes.html#memoryview" title="memoryview"><code class="xref py py-class docutils literal notranslate">memoryview</code></a> instances and other similar buffer objects.
(Backported from 3.x by Florent Xicluna; <a class="reference external" href="https://bugs.python.org/issue7703">bpo-7703</a>.)
</li>
<li>Updated module: the <code class="xref py py-mod docutils literal notranslate">bsddb</code> module has been updated from 4.7.2devel9
to version 4.8.4 of
<a class="reference external" href="https://www.jcea.es/programacion/pybsddb.htm">the pybsddb package</a>.
The new version features better Python 3.x compatibility, various bug fixes,
and adds several new BerkeleyDB flags and methods.
(Updated by Jesús Cea Avión; <a class="reference external" href="https://bugs.python.org/issue8156">bpo-8156</a>. The pybsddb
changelog can be read at <a class="reference external" href="http://hg.jcea.es/pybsddb/file/tip/ChangeLog">http://hg.jcea.es/pybsddb/file/tip/ChangeLog</a>.)
</li>
<li>The <a class="reference internal" href="../library/bz2.html#module-bz2" title="bz2: Interfaces for bzip2 compression and decompression."><code class="xref py py-mod docutils literal notranslate">bz2</code></a> module’s <a class="reference internal" href="../library/bz2.html#bz2.BZ2File" title="bz2.BZ2File"><code class="xref py py-class docutils literal notranslate">BZ2File</code></a> now supports the context
management protocol, so you can write <code class="docutils literal notranslate">with bz2.BZ2File(...) as f:</code>.
(Contributed by Hagen Fürstenau; <a class="reference external" href="https://bugs.python.org/issue3860">bpo-3860</a>.)
</li>
<li>New class: the <a class="reference internal" href="../library/collections.html#collections.Counter" title="collections.Counter"><code class="xref py py-class docutils literal notranslate">Counter</code></a> class in the <a class="reference internal" href="../library/collections.html#module-collections" title="collections: Container datatypes"><code class="xref py py-mod docutils literal notranslate">collections</code></a>
module is useful for tallying data. <a class="reference internal" href="../library/collections.html#collections.Counter" title="collections.Counter"><code class="xref py py-class docutils literal notranslate">Counter</code></a> instances
behave mostly like dictionaries but return zero for missing keys instead of
raising a <a class="reference internal" href="../library/exceptions.html#KeyError" title="KeyError"><code class="xref py py-exc docutils literal notranslate">KeyError</code></a>:
<div class="highlight-pycon notranslate"><div class="highlight"><pre>&gt;&gt;&gt; from collections import Counter
&gt;&gt;&gt; c = Counter()
&gt;&gt;&gt; for letter in &#39;here is a sample of english text&#39;:
... c[letter] += 1
...
&gt;&gt;&gt; c 
Counter({&#39; &#39;: 6, &#39;e&#39;: 5, &#39;s&#39;: 3, &#39;a&#39;: 2, &#39;i&#39;: 2, &#39;h&#39;: 2,
&#39;l&#39;: 2, &#39;t&#39;: 2, &#39;g&#39;: 1, &#39;f&#39;: 1, &#39;m&#39;: 1, &#39;o&#39;: 1, &#39;n&#39;: 1,
&#39;p&#39;: 1, &#39;r&#39;: 1, &#39;x&#39;: 1})
&gt;&gt;&gt; c[&#39;e&#39;]
5
&gt;&gt;&gt; c[&#39;z&#39;]
0
</pre></div>
</div>
There are three additional <a class="reference internal" href="../library/collections.html#collections.Counter" title="collections.Counter"><code class="xref py py-class docutils literal notranslate">Counter</code></a> methods.
<a class="reference internal" href="../library/collections.html#collections.Counter.most_common" title="collections.Counter.most_common"><code class="xref py py-meth docutils literal notranslate">most_common()</code></a> returns the N most common
elements and their counts. <a class="reference internal" href="../library/collections.html#collections.Counter.elements" title="collections.Counter.elements"><code class="xref py py-meth docutils literal notranslate">elements()</code></a>
returns an iterator over the contained elements, repeating each
element as many times as its count.
<a class="reference internal" href="../library/collections.html#collections.Counter.subtract" title="collections.Counter.subtract"><code class="xref py py-meth docutils literal notranslate">subtract()</code></a> takes an iterable and
subtracts one for each element instead of adding; if the argument is
a dictionary or another <code class="xref py py-class docutils literal notranslate">Counter</code>, the counts are
subtracted.
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; c.most_common(5)
[(&#39; &#39;, 6), (&#39;e&#39;, 5), (&#39;s&#39;, 3), (&#39;a&#39;, 2), (&#39;i&#39;, 2)]
&gt;&gt;&gt; c.elements() -&gt;
 &#39;a&#39;, &#39;a&#39;, &#39; &#39;, &#39; &#39;, &#39; &#39;, &#39; &#39;, &#39; &#39;, &#39; &#39;,
 &#39;e&#39;, &#39;e&#39;, &#39;e&#39;, &#39;e&#39;, &#39;e&#39;, &#39;g&#39;, &#39;f&#39;, &#39;i&#39;, &#39;i&#39;,
 &#39;h&#39;, &#39;h&#39;, &#39;m&#39;, &#39;l&#39;, &#39;l&#39;, &#39;o&#39;, &#39;n&#39;, &#39;p&#39;, &#39;s&#39;,
 &#39;s&#39;, &#39;s&#39;, &#39;r&#39;, &#39;t&#39;, &#39;t&#39;, &#39;x&#39;
&gt;&gt;&gt; c[&#39;e&#39;]
5
&gt;&gt;&gt; c.subtract(&#39;very heavy on the letter e&#39;)
&gt;&gt;&gt; c[&#39;e&#39;] # Count is now lower
-1
</pre></div>
</div>
Contributed by Raymond Hettinger; <a class="reference external" href="https://bugs.python.org/issue1696199">bpo-1696199</a>.
New class: <a class="reference internal" href="../library/collections.html#collections.OrderedDict" title="collections.OrderedDict"><code class="xref py py-class docutils literal notranslate">OrderedDict</code></a> is described in the earlier
section <a class="reference internal" href="#pep-0372">PEP 372: Adding an Ordered Dictionary to collections</a>.
New method: The <a class="reference internal" href="../library/collections.html#collections.deque" title="collections.deque"><code class="xref py py-class docutils literal notranslate">deque</code></a> data type now has a
<a class="reference internal" href="../library/collections.html#collections.deque.count" title="collections.deque.count"><code class="xref py py-meth docutils literal notranslate">count()</code></a> method that returns the number of
contained elements equal to the supplied argument x, and a
<a class="reference internal" href="../library/collections.html#collections.deque.reverse" title="collections.deque.reverse"><code class="xref py py-meth docutils literal notranslate">reverse()</code></a> method that reverses the elements
of the deque in-place. <a class="reference internal" href="../library/collections.html#collections.deque" title="collections.deque"><code class="xref py py-class docutils literal notranslate">deque</code></a> also exposes its maximum
length as the read-only <a class="reference internal" href="../library/collections.html#collections.deque.maxlen" title="collections.deque.maxlen"><code class="xref py py-attr docutils literal notranslate">maxlen</code></a> attribute.
(Both features added by Raymond Hettinger.)
The <a class="reference internal" href="../library/collections.html#collections.namedtuple" title="collections.namedtuple"><code class="xref py py-class docutils literal notranslate">namedtuple</code></a> class now has an optional rename parameter.
If rename is true, field names that are invalid because they’ve
been repeated or aren’t legal Python identifiers will be
renamed to legal names that are derived from the field’s
position within the list of fields:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; from collections import namedtuple
&gt;&gt;&gt; T = namedtuple(&#39;T&#39;, [&#39;field1&#39;, &#39;$illegal&#39;, &#39;for&#39;, &#39;field2&#39;], rename=True)
&gt;&gt;&gt; T._fields
(&#39;field1&#39;, &#39;_1&#39;, &#39;_2&#39;, &#39;field2&#39;)
</pre></div>
</div>
(Added by Raymond Hettinger; <a class="reference external" href="https://bugs.python.org/issue1818">bpo-1818</a>.)
Finally, the <code class="xref py py-class docutils literal notranslate">Mapping</code> abstract base class now
returns <a class="reference internal" href="../library/constants.html#NotImplemented" title="NotImplemented"><code class="xref py py-const docutils literal notranslate">NotImplemented</code></a> if a mapping is compared to
another type that isn’t a <code class="xref py py-class docutils literal notranslate">Mapping</code>.
(Fixed by Daniel Stutzbach; <a class="reference external" href="https://bugs.python.org/issue8729">bpo-8729</a>.)
</li>
<li>Constructors for the parsing classes in the <code class="xref py py-mod docutils literal notranslate">ConfigParser</code> module now
take an allow_no_value parameter, defaulting to false; if true,
options without values will be allowed. For example:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; import ConfigParser, StringIO
&gt;&gt;&gt; sample_config = &quot;&quot;&quot;
... [mysqld]
... user = mysql
... pid-file = /var/run/mysqld/mysqld.pid
... skip-bdb
... &quot;&quot;&quot;
&gt;&gt;&gt; config = ConfigParser.RawConfigParser(allow_no_value=True)
&gt;&gt;&gt; config.readfp(StringIO.StringIO(sample_config))
&gt;&gt;&gt; config.get(&#39;mysqld&#39;, &#39;user&#39;)
&#39;mysql&#39;
&gt;&gt;&gt; print config.get(&#39;mysqld&#39;, &#39;skip-bdb&#39;)
None
&gt;&gt;&gt; print config.get(&#39;mysqld&#39;, &#39;unknown&#39;)
Traceback (most recent call last):
 ...
NoOptionError: No option &#39;unknown&#39; in section: &#39;mysqld&#39;
</pre></div>
</div>
(Contributed by Mats Kindahl; <a class="reference external" href="https://bugs.python.org/issue7005">bpo-7005</a>.)
</li>
<li>Deprecated function: <code class="xref py py-func docutils literal notranslate">contextlib.nested()</code>, which allows
handling more than one context manager with a single <a class="reference internal" href="../reference/compound_stmts.html#with"><code class="xref std std-keyword docutils literal notranslate">with</code></a>
statement, has been deprecated, because the <a class="reference internal" href="../reference/compound_stmts.html#with"><code class="xref std std-keyword docutils literal notranslate">with</code></a> statement
now supports multiple context managers.
</li>
<li>The <code class="xref py py-mod docutils literal notranslate">cookielib</code> module now ignores cookies that have an invalid
version field, one that doesn’t contain an integer value. (Fixed by
John J. Lee; <a class="reference external" href="https://bugs.python.org/issue3924">bpo-3924</a>.)
</li>
<li>The <a class="reference internal" href="../library/copy.html#module-copy" title="copy: Shallow and deep copy operations."><code class="xref py py-mod docutils literal notranslate">copy</code></a> module’s <a class="reference internal" href="../library/copy.html#copy.deepcopy" title="copy.deepcopy"><code class="xref py py-func docutils literal notranslate">deepcopy()</code></a> function will now
correctly copy bound instance methods. (Implemented by
Robert Collins; <a class="reference external" href="https://bugs.python.org/issue1515">bpo-1515</a>.)
</li>
<li>The <a class="reference internal" href="../library/ctypes.html#module-ctypes" title="ctypes: A foreign function library for Python."><code class="xref py py-mod docutils literal notranslate">ctypes</code></a> module now always converts <code class="docutils literal notranslate">None</code> to a C NULL
pointer for arguments declared as pointers. (Changed by Thomas
Heller; <a class="reference external" href="https://bugs.python.org/issue4606">bpo-4606</a>.) The underlying <a class="reference external" href="https://sourceware.org/libffi/">libffi library</a> has been updated to version
3.0.9, containing various fixes for different platforms. (Updated
by Matthias Klose; <a class="reference external" href="https://bugs.python.org/issue8142">bpo-8142</a>.)
</li>
<li>New method: the <a class="reference internal" href="../library/datetime.html#module-datetime" title="datetime: Basic date and time types."><code class="xref py py-mod docutils literal notranslate">datetime</code></a> module’s <a class="reference internal" href="../library/datetime.html#datetime.timedelta" title="datetime.timedelta"><code class="xref py py-class docutils literal notranslate">timedelta</code></a> class
gained a <a class="reference internal" href="../library/datetime.html#datetime.timedelta.total_seconds" title="datetime.timedelta.total_seconds"><code class="xref py py-meth docutils literal notranslate">total_seconds()</code></a> method that returns the
number of seconds in the duration. (Contributed by Brian Quinlan; <a class="reference external" href="https://bugs.python.org/issue5788">bpo-5788</a>.)
</li>
<li>New method: the <a class="reference internal" href="../library/decimal.html#decimal.Decimal" title="decimal.Decimal"><code class="xref py py-class docutils literal notranslate">Decimal</code></a> class gained a
<a class="reference internal" href="../library/decimal.html#decimal.Decimal.from_float" title="decimal.Decimal.from_float"><code class="xref py py-meth docutils literal notranslate">from_float()</code></a> class method that performs an exact
conversion of a floating-point number to a <a class="reference internal" href="../library/decimal.html#decimal.Decimal" title="decimal.Decimal"><code class="xref py py-class docutils literal notranslate">Decimal</code></a>.
This exact conversion strives for the
closest decimal approximation to the floating-point representation’s value;
the resulting decimal value will therefore still include the inaccuracy,
if any.
For example, <code class="docutils literal notranslate">Decimal.from_float(0.1)</code> returns
<code class="docutils literal notranslate">Decimal('0.1000000000000000055511151231257827021181583404541015625')</code>.
(Implemented by Raymond Hettinger; <a class="reference external" href="https://bugs.python.org/issue4796">bpo-4796</a>.)
Comparing instances of <a class="reference internal" href="../library/decimal.html#decimal.Decimal" title="decimal.Decimal"><code class="xref py py-class docutils literal notranslate">Decimal</code></a> with floating-point
numbers now produces sensible results based on the numeric values
of the operands. Previously such comparisons would fall back to
Python’s default rules for comparing objects, which produced arbitrary
results based on their type. Note that you still cannot combine
<code class="xref py py-class docutils literal notranslate">Decimal</code> and floating-point in other operations such as addition,
since you should be explicitly choosing how to convert between float and
<a class="reference internal" href="../library/decimal.html#decimal.Decimal" title="decimal.Decimal"><code class="xref py py-class docutils literal notranslate">Decimal</code></a>. (Fixed by Mark Dickinson; <a class="reference external" href="https://bugs.python.org/issue2531">bpo-2531</a>.)
The constructor for <a class="reference internal" href="../library/decimal.html#decimal.Decimal" title="decimal.Decimal"><code class="xref py py-class docutils literal notranslate">Decimal</code></a> now accepts
floating-point numbers (added by Raymond Hettinger; <a class="reference external" href="https://bugs.python.org/issue8257">bpo-8257</a>)
and non-European Unicode characters such as Arabic-Indic digits
(contributed by Mark Dickinson; <a class="reference external" href="https://bugs.python.org/issue6595">bpo-6595</a>).
Most of the methods of the <a class="reference internal" href="../library/decimal.html#decimal.Context" title="decimal.Context"><code class="xref py py-class docutils literal notranslate">Context</code></a> class now accept integers
as well as <a class="reference internal" href="../library/decimal.html#decimal.Decimal" title="decimal.Decimal"><code class="xref py py-class docutils literal notranslate">Decimal</code></a> instances; the only exceptions are the
<a class="reference internal" href="../library/decimal.html#decimal.Context.canonical" title="decimal.Context.canonical"><code class="xref py py-meth docutils literal notranslate">canonical()</code></a> and <a class="reference internal" href="../library/decimal.html#decimal.Context.is_canonical" title="decimal.Context.is_canonical"><code class="xref py py-meth docutils literal notranslate">is_canonical()</code></a>
methods. (Patch by Juan José Conti; <a class="reference external" href="https://bugs.python.org/issue7633">bpo-7633</a>.)
When using <a class="reference internal" href="../library/decimal.html#decimal.Decimal" title="decimal.Decimal"><code class="xref py py-class docutils literal notranslate">Decimal</code></a> instances with a string’s
<a class="reference internal" href="../library/stdtypes.html#str.format" title="str.format"><code class="xref py py-meth docutils literal notranslate">format()</code></a> method, the default alignment was previously
left-alignment. This has been changed to right-alignment, which is
more sensible for numeric types. (Changed by Mark Dickinson; <a class="reference external" href="https://bugs.python.org/issue6857">bpo-6857</a>.)
Comparisons involving a signaling NaN value (or <code class="docutils literal notranslate">sNAN</code>) now signal
<code class="xref py py-const docutils literal notranslate">InvalidOperation</code> instead of silently returning a true or
false value depending on the comparison operator. Quiet NaN values
(or <code class="docutils literal notranslate">NaN</code>) are now hashable. (Fixed by Mark Dickinson;
<a class="reference external" href="https://bugs.python.org/issue7279">bpo-7279</a>.)
</li>
<li>The <a class="reference internal" href="../library/difflib.html#module-difflib" title="difflib: Helpers for computing differences between objects."><code class="xref py py-mod docutils literal notranslate">difflib</code></a> module now produces output that is more
compatible with modern diff/patch tools
through one small change, using a tab character instead of spaces as
a separator in the header giving the filename. (Fixed by Anatoly
Techtonik; <a class="reference external" href="https://bugs.python.org/issue7585">bpo-7585</a>.)
</li>
<li>The Distutils <code class="docutils literal notranslate">sdist</code> command now always regenerates the
<code class="file docutils literal notranslate">MANIFEST</code> file, since even if the <code class="file docutils literal notranslate">MANIFEST.in</code> or
<code class="file docutils literal notranslate">setup.py</code> files haven’t been modified, the user might have
created some new files that should be included.
(Fixed by Tarek Ziadé; <a class="reference external" href="https://bugs.python.org/issue8688">bpo-8688</a>.)
</li>
<li>The <a class="reference internal" href="../library/doctest.html#module-doctest" title="doctest: Test pieces of code within docstrings."><code class="xref py py-mod docutils literal notranslate">doctest</code></a> module’s <code class="xref py py-const docutils literal notranslate">IGNORE_EXCEPTION_DETAIL</code> flag
will now ignore the name of the module containing the exception
being tested. (Patch by Lennart Regebro; <a class="reference external" href="https://bugs.python.org/issue7490">bpo-7490</a>.)
</li>
<li>The <a class="reference internal" href="../library/email.html#module-email" title="email: Package supporting the parsing, manipulating, and generating email messages."><code class="xref py py-mod docutils literal notranslate">email</code></a> module’s <a class="reference internal" href="../library/email.compat32-message.html#email.message.Message" title="email.message.Message"><code class="xref py py-class docutils literal notranslate">Message</code></a> class will
now accept a Unicode-valued payload, automatically converting the
payload to the encoding specified by <code class="xref py py-attr docutils literal notranslate">output_charset</code>.
(Added by R. David Murray; <a class="reference external" href="https://bugs.python.org/issue1368247">bpo-1368247</a>.)
</li>
<li>The <a class="reference internal" href="../library/fractions.html#fractions.Fraction" title="fractions.Fraction"><code class="xref py py-class docutils literal notranslate">Fraction</code></a> class now accepts a single float or
<a class="reference internal" href="../library/decimal.html#decimal.Decimal" title="decimal.Decimal"><code class="xref py py-class docutils literal notranslate">Decimal</code></a> instance, or two rational numbers, as
arguments to its constructor. (Implemented by Mark Dickinson;
rationals added in <a class="reference external" href="https://bugs.python.org/issue5812">bpo-5812</a>, and float/decimal in
<a class="reference external" href="https://bugs.python.org/issue8294">bpo-8294</a>.)
Ordering comparisons (<code class="docutils literal notranslate">&lt;</code>, <code class="docutils literal notranslate">&lt;=</code>, <code class="docutils literal notranslate">&gt;</code>, <code class="docutils literal notranslate">&gt;=</code>) between
fractions and complex numbers now raise a <a class="reference internal" href="../library/exceptions.html#TypeError" title="TypeError"><code class="xref py py-exc docutils literal notranslate">TypeError</code></a>.
This fixes an oversight, making the <a class="reference internal" href="../library/fractions.html#fractions.Fraction" title="fractions.Fraction"><code class="xref py py-class docutils literal notranslate">Fraction</code></a>
match the other numeric types.
</li>
<li>New class: <a class="reference internal" href="../library/ftplib.html#ftplib.FTP_TLS" title="ftplib.FTP_TLS"><code class="xref py py-class docutils literal notranslate">FTP_TLS</code></a> in
the <a class="reference internal" href="../library/ftplib.html#module-ftplib" title="ftplib: FTP protocol client (requires sockets)."><code class="xref py py-mod docutils literal notranslate">ftplib</code></a> module provides secure FTP
connections using TLS encapsulation of authentication as well as
subsequent control and data transfers.
(Contributed by Giampaolo Rodola; <a class="reference external" href="https://bugs.python.org/issue2054">bpo-2054</a>.)
The <a class="reference internal" href="../library/ftplib.html#ftplib.FTP.storbinary" title="ftplib.FTP.storbinary"><code class="xref py py-meth docutils literal notranslate">storbinary()</code></a> method for binary uploads can now restart
uploads thanks to an added rest parameter (patch by Pablo Mouzo;
<a class="reference external" href="https://bugs.python.org/issue6845">bpo-6845</a>.)
</li>
<li>New class decorator: <a class="reference internal" href="../library/functools.html#functools.total_ordering" title="functools.total_ordering"><code class="xref py py-func docutils literal notranslate">total_ordering()</code></a> in the <a class="reference internal" href="../library/functools.html#module-functools" title="functools: Higher-order functions and operations on callable objects."><code class="xref py py-mod docutils literal notranslate">functools</code></a>
module takes a class that defines an <a class="reference internal" href="../reference/datamodel.html#object.__eq__" title="object.__eq__"><code class="xref py py-meth docutils literal notranslate">__eq__()</code></a> method and one of
<a class="reference internal" href="../reference/datamodel.html#object.__lt__" title="object.__lt__"><code class="xref py py-meth docutils literal notranslate">__lt__()</code></a>, <a class="reference internal" href="../reference/datamodel.html#object.__le__" title="object.__le__"><code class="xref py py-meth docutils literal notranslate">__le__()</code></a>, <a class="reference internal" href="../reference/datamodel.html#object.__gt__" title="object.__gt__"><code class="xref py py-meth docutils literal notranslate">__gt__()</code></a>, or <a class="reference internal" href="../reference/datamodel.html#object.__ge__" title="object.__ge__"><code class="xref py py-meth docutils literal notranslate">__ge__()</code></a>,
and generates the missing comparison methods. Since the
<code class="xref py py-meth docutils literal notranslate">__cmp__()</code> method is being deprecated in Python 3.x,
this decorator makes it easier to define ordered classes.
(Added by Raymond Hettinger; <a class="reference external" href="https://bugs.python.org/issue5479">bpo-5479</a>.)
New function: <a class="reference internal" href="../library/functools.html#functools.cmp_to_key" title="functools.cmp_to_key"><code class="xref py py-func docutils literal notranslate">cmp_to_key()</code></a> will take an old-style comparison
function that expects two arguments and return a new callable that
can be used as the key parameter to functions such as
<a class="reference internal" href="../library/functions.html#sorted" title="sorted"><code class="xref py py-func docutils literal notranslate">sorted()</code></a>, <a class="reference internal" href="../library/functions.html#min" title="min"><code class="xref py py-func docutils literal notranslate">min()</code></a> and <a class="reference internal" href="../library/functions.html#max" title="max"><code class="xref py py-func docutils literal notranslate">max()</code></a>, etc. The primary
intended use is to help with making code compatible with Python 3.x.
(Added by Raymond Hettinger.)
</li>
<li>New function: the <a class="reference internal" href="../library/gc.html#module-gc" title="gc: Interface to the cycle-detecting garbage collector."><code class="xref py py-mod docutils literal notranslate">gc</code></a> module’s <a class="reference internal" href="../library/gc.html#gc.is_tracked" title="gc.is_tracked"><code class="xref py py-func docutils literal notranslate">is_tracked()</code></a> returns
true if a given instance is tracked by the garbage collector, false
otherwise. (Contributed by Antoine Pitrou; <a class="reference external" href="https://bugs.python.org/issue4688">bpo-4688</a>.)
</li>
<li>The <a class="reference internal" href="../library/gzip.html#module-gzip" title="gzip: Interfaces for gzip compression and decompression using file objects."><code class="xref py py-mod docutils literal notranslate">gzip</code></a> module’s <a class="reference internal" href="../library/gzip.html#gzip.GzipFile" title="gzip.GzipFile"><code class="xref py py-class docutils literal notranslate">GzipFile</code></a> now supports the context
management protocol, so you can write <code class="docutils literal notranslate">with gzip.GzipFile(...) as f:</code>
(contributed by Hagen Fürstenau; <a class="reference external" href="https://bugs.python.org/issue3860">bpo-3860</a>), and it now implements
the <a class="reference internal" href="../library/io.html#io.BufferedIOBase" title="io.BufferedIOBase"><code class="xref py py-class docutils literal notranslate">io.BufferedIOBase</code></a> ABC, so you can wrap it with
<a class="reference internal" href="../library/io.html#io.BufferedReader" title="io.BufferedReader"><code class="xref py py-class docutils literal notranslate">io.BufferedReader</code></a> for faster processing
(contributed by Nir Aides; <a class="reference external" href="https://bugs.python.org/issue7471">bpo-7471</a>).
It’s also now possible to override the modification time
recorded in a gzipped file by providing an optional timestamp to
the constructor. (Contributed by Jacques Frechet; <a class="reference external" href="https://bugs.python.org/issue4272">bpo-4272</a>.)
Files in gzip format can be padded with trailing zero bytes; the
<a class="reference internal" href="../library/gzip.html#module-gzip" title="gzip: Interfaces for gzip compression and decompression using file objects."><code class="xref py py-mod docutils literal notranslate">gzip</code></a> module will now consume these trailing bytes. (Fixed by
Tadek Pietraszek and Brian Curtin; <a class="reference external" href="https://bugs.python.org/issue2846">bpo-2846</a>.)
</li>
<li>New attribute: the <a class="reference internal" href="../library/hashlib.html#module-hashlib" title="hashlib: Secure hash and message digest algorithms."><code class="xref py py-mod docutils literal notranslate">hashlib</code></a> module now has an <code class="xref py py-attr docutils literal notranslate">algorithms</code>
attribute containing a tuple naming the supported algorithms.
In Python 2.7, <code class="docutils literal notranslate">hashlib.algorithms</code> contains
<code class="docutils literal notranslate">('md5', 'sha1', 'sha224', 'sha256', 'sha384', 'sha512')</code>.
(Contributed by Carl Chenet; <a class="reference external" href="https://bugs.python.org/issue7418">bpo-7418</a>.)
</li>
<li>The default <code class="xref py py-class docutils literal notranslate">HTTPResponse</code> class used by the <code class="xref py py-mod docutils literal notranslate">httplib</code> module now
supports buffering, resulting in much faster reading of HTTP responses.
(Contributed by Kristján Valur Jónsson; <a class="reference external" href="https://bugs.python.org/issue4879">bpo-4879</a>.)
The <code class="xref py py-class docutils literal notranslate">HTTPConnection</code> and <code class="xref py py-class docutils literal notranslate">HTTPSConnection</code> classes
now support a source_address parameter, a <code class="docutils literal notranslate">(host, port)</code> 2-tuple
giving the source address that will be used for the connection.
(Contributed by Eldon Ziegler; <a class="reference external" href="https://bugs.python.org/issue3972">bpo-3972</a>.)
</li>
<li>The <code class="xref py py-mod docutils literal notranslate">ihooks</code> module now supports relative imports. Note that
<code class="xref py py-mod docutils literal notranslate">ihooks</code> is an older module for customizing imports,
superseded by the <code class="xref py py-mod docutils literal notranslate">imputil</code> module added in Python 2.0.
(Relative import support added by Neil Schemenauer.)
</li>
<li>The <a class="reference internal" href="../library/imaplib.html#module-imaplib" title="imaplib: IMAP4 protocol client (requires sockets)."><code class="xref py py-mod docutils literal notranslate">imaplib</code></a> module now supports IPv6 addresses.
(Contributed by Derek Morr; <a class="reference external" href="https://bugs.python.org/issue1655">bpo-1655</a>.)
</li>
<li>New function: the <a class="reference internal" href="../library/inspect.html#module-inspect" title="inspect: Extract information and source code from live objects."><code class="xref py py-mod docutils literal notranslate">inspect</code></a> module’s <a class="reference internal" href="../library/inspect.html#inspect.getcallargs" title="inspect.getcallargs"><code class="xref py py-func docutils literal notranslate">getcallargs()</code></a>
takes a callable and its positional and keyword arguments,
and figures out which of the callable’s parameters will receive each argument,
returning a dictionary mapping argument names to their values. For example:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; from inspect import getcallargs
&gt;&gt;&gt; def f(a, b=1, *pos, **named):
... pass
&gt;&gt;&gt; getcallargs(f, 1, 2, 3)
{&#39;a&#39;: 1, &#39;b&#39;: 2, &#39;pos&#39;: (3,), &#39;named&#39;: {}}
&gt;&gt;&gt; getcallargs(f, a=2, x=4)
{&#39;a&#39;: 2, &#39;b&#39;: 1, &#39;pos&#39;: (), &#39;named&#39;: {&#39;x&#39;: 4}}
&gt;&gt;&gt; getcallargs(f)
Traceback (most recent call last):
...
TypeError: f() takes at least 1 argument (0 given)
</pre></div>
</div>
Contributed by George Sakkis; <a class="reference external" href="https://bugs.python.org/issue3135">bpo-3135</a>.
</li>
<li>Updated module: The <a class="reference internal" href="../library/io.html#module-io" title="io: Core tools for working with streams."><code class="xref py py-mod docutils literal notranslate">io</code></a> library has been upgraded to the version shipped with
Python 3.1. For 3.1, the I/O library was entirely rewritten in C
and is 2 to 20 times faster depending on the task being performed. The
original Python version was renamed to the <code class="xref py py-mod docutils literal notranslate">_pyio</code> module.
One minor resulting change: the <a class="reference internal" href="../library/io.html#io.TextIOBase" title="io.TextIOBase"><code class="xref py py-class docutils literal notranslate">io.TextIOBase</code></a> class now
has an <code class="xref py py-attr docutils literal notranslate">errors</code> attribute giving the error setting
used for encoding and decoding errors (one of <code class="docutils literal notranslate">'strict'</code>, <code class="docutils literal notranslate">'replace'</code>,
<code class="docutils literal notranslate">'ignore'</code>).
The <a class="reference internal" href="../library/io.html#io.FileIO" title="io.FileIO"><code class="xref py py-class docutils literal notranslate">io.FileIO</code></a> class now raises an <a class="reference internal" href="../library/exceptions.html#OSError" title="OSError"><code class="xref py py-exc docutils literal notranslate">OSError</code></a> when passed
an invalid file descriptor. (Implemented by Benjamin Peterson;
<a class="reference external" href="https://bugs.python.org/issue4991">bpo-4991</a>.) The <a class="reference internal" href="../library/io.html#io.IOBase.truncate" title="io.IOBase.truncate"><code class="xref py py-meth docutils literal notranslate">truncate()</code></a> method now preserves the
file position; previously it would change the file position to the
end of the new file. (Fixed by Pascal Chambon; <a class="reference external" href="https://bugs.python.org/issue6939">bpo-6939</a>.)
</li>
<li>New function: <code class="docutils literal notranslate">itertools.compress(data, selectors)</code> takes two
iterators. Elements of data are returned if the corresponding
value in selectors is true:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>itertools.compress(&#39;ABCDEF&#39;, [1,0,1,0,1,1]) =&gt;
 A, C, E, F
</pre></div>
</div>
New function: <code class="docutils literal notranslate">itertools.combinations_with_replacement(iter, r)</code>
returns all the possible r-length combinations of elements from the
iterable iter. Unlike <a class="reference internal" href="../library/itertools.html#itertools.combinations" title="itertools.combinations"><code class="xref py py-func docutils literal notranslate">combinations()</code></a>, individual elements
can be repeated in the generated combinations:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>itertools.combinations_with_replacement(&#39;abc&#39;, 2) =&gt;
 (&#39;a&#39;, &#39;a&#39;), (&#39;a&#39;, &#39;b&#39;), (&#39;a&#39;, &#39;c&#39;),
 (&#39;b&#39;, &#39;b&#39;), (&#39;b&#39;, &#39;c&#39;), (&#39;c&#39;, &#39;c&#39;)
</pre></div>
</div>
Note that elements are treated as unique depending on their position
in the input, not their actual values.
The <a class="reference internal" href="../library/itertools.html#itertools.count" title="itertools.count"><code class="xref py py-func docutils literal notranslate">itertools.count()</code></a> function now has a step argument that
allows incrementing by values other than 1. <a class="reference internal" href="../library/itertools.html#itertools.count" title="itertools.count"><code class="xref py py-func docutils literal notranslate">count()</code></a> also
now allows keyword arguments, and using non-integer values such as
floats or <a class="reference internal" href="../library/decimal.html#decimal.Decimal" title="decimal.Decimal"><code class="xref py py-class docutils literal notranslate">Decimal</code></a> instances. (Implemented by Raymond
Hettinger; <a class="reference external" href="https://bugs.python.org/issue5032">bpo-5032</a>.)
<a class="reference internal" href="../library/itertools.html#itertools.combinations" title="itertools.combinations"><code class="xref py py-func docutils literal notranslate">itertools.combinations()</code></a> and <a class="reference internal" href="../library/itertools.html#itertools.product" title="itertools.product"><code class="xref py py-func docutils literal notranslate">itertools.product()</code></a>
previously raised <a class="reference internal" href="../library/exceptions.html#ValueError" title="ValueError"><code class="xref py py-exc docutils literal notranslate">ValueError</code></a> for values of r larger than
the input iterable. This was deemed a specification error, so they
now return an empty iterator. (Fixed by Raymond Hettinger; <a class="reference external" href="https://bugs.python.org/issue4816">bpo-4816</a>.)
</li>
<li>Updated module: The <a class="reference internal" href="../library/json.html#module-json" title="json: Encode and decode the JSON format."><code class="xref py py-mod docutils literal notranslate">json</code></a> module was upgraded to version 2.0.9 of the
simplejson package, which includes a C extension that makes
encoding and decoding faster.
(Contributed by Bob Ippolito; <a class="reference external" href="https://bugs.python.org/issue4136">bpo-4136</a>.)
To support the new <a class="reference internal" href="../library/collections.html#collections.OrderedDict" title="collections.OrderedDict"><code class="xref py py-class docutils literal notranslate">collections.OrderedDict</code></a> type, <a class="reference internal" href="../library/json.html#json.load" title="json.load"><code class="xref py py-func docutils literal notranslate">json.load()</code></a>
now has an optional object_pairs_hook parameter that will be called
with any object literal that decodes to a list of pairs.
(Contributed by Raymond Hettinger; <a class="reference external" href="https://bugs.python.org/issue5381">bpo-5381</a>.)
</li>
<li>The <a class="reference internal" href="../library/mailbox.html#module-mailbox" title="mailbox: Manipulate mailboxes in various formats"><code class="xref py py-mod docutils literal notranslate">mailbox</code></a> module’s <a class="reference internal" href="../library/mailbox.html#mailbox.Maildir" title="mailbox.Maildir"><code class="xref py py-class docutils literal notranslate">Maildir</code></a> class now records the
timestamp on the directories it reads, and only re-reads them if the
modification time has subsequently changed. This improves
performance by avoiding unneeded directory scans. (Fixed by
A.M. Kuchling and Antoine Pitrou; <a class="reference external" href="https://bugs.python.org/issue1607951">bpo-1607951</a>, <a class="reference external" href="https://bugs.python.org/issue6896">bpo-6896</a>.)
</li>
<li>New functions: the <a class="reference internal" href="../library/math.html#module-math" title="math: Mathematical functions (sin() etc.)."><code class="xref py py-mod docutils literal notranslate">math</code></a> module gained
<a class="reference internal" href="../library/math.html#math.erf" title="math.erf"><code class="xref py py-func docutils literal notranslate">erf()</code></a> and <a class="reference internal" href="../library/math.html#math.erfc" title="math.erfc"><code class="xref py py-func docutils literal notranslate">erfc()</code></a> for the error function and the complementary error function,
<a class="reference internal" href="../library/math.html#math.expm1" title="math.expm1"><code class="xref py py-func docutils literal notranslate">expm1()</code></a> which computes <code class="docutils literal notranslate">e**x - 1</code> with more precision than
using <a class="reference internal" href="../library/math.html#math.exp" title="math.exp"><code class="xref py py-func docutils literal notranslate">exp()</code></a> and subtracting 1,
<a class="reference internal" href="../library/math.html#math.gamma" title="math.gamma"><code class="xref py py-func docutils literal notranslate">gamma()</code></a> for the Gamma function, and
<a class="reference internal" href="../library/math.html#math.lgamma" title="math.lgamma"><code class="xref py py-func docutils literal notranslate">lgamma()</code></a> for the natural log of the Gamma function.
(Contributed by Mark Dickinson and nirinA raseliarison; <a class="reference external" href="https://bugs.python.org/issue3366">bpo-3366</a>.)
</li>
<li>The <a class="reference internal" href="../library/multiprocessing.html#module-multiprocessing" title="multiprocessing: Process-based parallelism."><code class="xref py py-mod docutils literal notranslate">multiprocessing</code></a> module’s <code class="xref py py-class docutils literal notranslate">Manager*</code> classes
can now be passed a callable that will be called whenever
a subprocess is started, along with a set of arguments that will be
passed to the callable.
(Contributed by lekma; <a class="reference external" href="https://bugs.python.org/issue5585">bpo-5585</a>.)
The <code class="xref py py-class docutils literal notranslate">Pool</code> class, which controls a pool of worker processes,
now has an optional maxtasksperchild parameter. Worker processes
will perform the specified number of tasks and then exit, causing the
<code class="xref py py-class docutils literal notranslate">Pool</code> to start a new worker. This is useful if tasks may leak
memory or other resources, or if some tasks will cause the worker to
become very large.
(Contributed by Charles Cazabon; <a class="reference external" href="https://bugs.python.org/issue6963">bpo-6963</a>.)
</li>
<li>The <a class="reference internal" href="../library/nntplib.html#module-nntplib" title="nntplib: NNTP protocol client (requires sockets)."><code class="xref py py-mod docutils literal notranslate">nntplib</code></a> module now supports IPv6 addresses.
(Contributed by Derek Morr; <a class="reference external" href="https://bugs.python.org/issue1664">bpo-1664</a>.)
</li>
<li>New functions: the <a class="reference internal" href="../library/os.html#module-os" title="os: Miscellaneous operating system interfaces."><code class="xref py py-mod docutils literal notranslate">os</code></a> module wraps the following POSIX system
calls: <a class="reference internal" href="../library/os.html#os.getresgid" title="os.getresgid"><code class="xref py py-func docutils literal notranslate">getresgid()</code></a> and <a class="reference internal" href="../library/os.html#os.getresuid" title="os.getresuid"><code class="xref py py-func docutils literal notranslate">getresuid()</code></a>, which return the
real, effective, and saved GIDs and UIDs;
<a class="reference internal" href="../library/os.html#os.setresgid" title="os.setresgid"><code class="xref py py-func docutils literal notranslate">setresgid()</code></a> and <a class="reference internal" href="../library/os.html#os.setresuid" title="os.setresuid"><code class="xref py py-func docutils literal notranslate">setresuid()</code></a>, which set
real, effective, and saved GIDs and UIDs to new values;
<a class="reference internal" href="../library/os.html#os.initgroups" title="os.initgroups"><code class="xref py py-func docutils literal notranslate">initgroups()</code></a>, which initialize the group access list
for the current process. (GID/UID functions
contributed by Travis H.; <a class="reference external" href="https://bugs.python.org/issue6508">bpo-6508</a>. Support for initgroups added
by Jean-Paul Calderone; <a class="reference external" href="https://bugs.python.org/issue7333">bpo-7333</a>.)
The <a class="reference internal" href="../library/os.html#os.fork" title="os.fork"><code class="xref py py-func docutils literal notranslate">os.fork()</code></a> function now re-initializes the import lock in
the child process; this fixes problems on Solaris when <a class="reference internal" href="../library/os.html#os.fork" title="os.fork"><code class="xref py py-func docutils literal notranslate">fork()</code></a>
is called from a thread. (Fixed by Zsolt Cserna; <a class="reference external" href="https://bugs.python.org/issue7242">bpo-7242</a>.)
</li>
<li>In the <a class="reference internal" href="../library/os.path.html#module-os.path" title="os.path: Operations on pathnames."><code class="xref py py-mod docutils literal notranslate">os.path</code></a> module, the <a class="reference internal" href="../library/os.path.html#os.path.normpath" title="os.path.normpath"><code class="xref py py-func docutils literal notranslate">normpath()</code></a> and
<a class="reference internal" href="../library/os.path.html#os.path.abspath" title="os.path.abspath"><code class="xref py py-func docutils literal notranslate">abspath()</code></a> functions now preserve Unicode; if their input path
is a Unicode string, the return value is also a Unicode string.
(<a class="reference internal" href="../library/os.path.html#os.path.normpath" title="os.path.normpath"><code class="xref py py-meth docutils literal notranslate">normpath()</code></a> fixed by Matt Giuca in <a class="reference external" href="https://bugs.python.org/issue5827">bpo-5827</a>;
<a class="reference internal" href="../library/os.path.html#os.path.abspath" title="os.path.abspath"><code class="xref py py-meth docutils literal notranslate">abspath()</code></a> fixed by Ezio Melotti in <a class="reference external" href="https://bugs.python.org/issue3426">bpo-3426</a>.)
</li>
<li>The <a class="reference internal" href="../library/pydoc.html#module-pydoc" title="pydoc: Documentation generator and online help system."><code class="xref py py-mod docutils literal notranslate">pydoc</code></a> module now has help for the various symbols that Python
uses. You can now do <code class="docutils literal notranslate">help('&lt;&lt;')</code> or <code class="docutils literal notranslate">help('&#64;')</code>, for example.
(Contributed by David Laban; <a class="reference external" href="https://bugs.python.org/issue4739">bpo-4739</a>.)
</li>
<li>The <a class="reference internal" href="../library/re.html#module-re" title="re: Regular expression operations."><code class="xref py py-mod docutils literal notranslate">re</code></a> module’s <a class="reference internal" href="../library/re.html#re.split" title="re.split"><code class="xref py py-func docutils literal notranslate">split()</code></a>, <a class="reference internal" href="../library/re.html#re.sub" title="re.sub"><code class="xref py py-func docutils literal notranslate">sub()</code></a>, and <a class="reference internal" href="../library/re.html#re.subn" title="re.subn"><code class="xref py py-func docutils literal notranslate">subn()</code></a>
now accept an optional flags argument, for consistency with the
other functions in the module. (Added by Gregory P. Smith.)
</li>
<li>New function: <a class="reference internal" href="../library/runpy.html#runpy.run_path" title="runpy.run_path"><code class="xref py py-func docutils literal notranslate">run_path()</code></a> in the <a class="reference internal" href="../library/runpy.html#module-runpy" title="runpy: Locate and run Python modules without importing them first."><code class="xref py py-mod docutils literal notranslate">runpy</code></a> module
will execute the code at a provided path argument. path can be
the path of a Python source file (<code class="file docutils literal notranslate">example.py</code>), a compiled
bytecode file (<code class="file docutils literal notranslate">example.pyc</code>), a directory
(<code class="file docutils literal notranslate">./package/</code>), or a zip archive (<code class="file docutils literal notranslate">example.zip</code>). If a
directory or zip path is provided, it will be added to the front of
<code class="docutils literal notranslate">sys.path</code> and the module <a class="reference internal" href="../library/__main__.html#module-__main__" title="__main__: The environment where the top-level script is run."><code class="xref py py-mod docutils literal notranslate">__main__</code></a> will be imported. It’s
expected that the directory or zip contains a <code class="file docutils literal notranslate">__main__.py</code>;
if it doesn’t, some other <code class="file docutils literal notranslate">__main__.py</code> might be imported from
a location later in <code class="docutils literal notranslate">sys.path</code>. This makes more of the machinery
of <a class="reference internal" href="../library/runpy.html#module-runpy" title="runpy: Locate and run Python modules without importing them first."><code class="xref py py-mod docutils literal notranslate">runpy</code></a> available to scripts that want to mimic the way
Python’s command line processes an explicit path name.
(Added by Nick Coghlan; <a class="reference external" href="https://bugs.python.org/issue6816">bpo-6816</a>.)
</li>
<li>New function: in the <a class="reference internal" href="../library/shutil.html#module-shutil" title="shutil: High-level file operations, including copying."><code class="xref py py-mod docutils literal notranslate">shutil</code></a> module, <a class="reference internal" href="../library/shutil.html#shutil.make_archive" title="shutil.make_archive"><code class="xref py py-func docutils literal notranslate">make_archive()</code></a>
takes a filename, archive type (zip or tar-format), and a directory
path, and creates an archive containing the directory’s contents.
(Added by Tarek Ziadé.)
<a class="reference internal" href="../library/shutil.html#module-shutil" title="shutil: High-level file operations, including copying."><code class="xref py py-mod docutils literal notranslate">shutil</code></a>’s <a class="reference internal" href="../library/shutil.html#shutil.copyfile" title="shutil.copyfile"><code class="xref py py-func docutils literal notranslate">copyfile()</code></a> and <a class="reference internal" href="../library/shutil.html#shutil.copytree" title="shutil.copytree"><code class="xref py py-func docutils literal notranslate">copytree()</code></a>
functions now raise a <code class="xref py py-exc docutils literal notranslate">SpecialFileError</code> exception when
asked to copy a named pipe. Previously the code would treat
named pipes like a regular file by opening them for reading, and
this would block indefinitely. (Fixed by Antoine Pitrou; <a class="reference external" href="https://bugs.python.org/issue3002">bpo-3002</a>.)
</li>
<li>The <a class="reference internal" href="../library/signal.html#module-signal" title="signal: Set handlers for asynchronous events."><code class="xref py py-mod docutils literal notranslate">signal</code></a> module no longer re-installs the signal handler
unless this is truly necessary, which fixes a bug that could make it
impossible to catch the EINTR signal robustly. (Fixed by
Charles-Francois Natali; <a class="reference external" href="https://bugs.python.org/issue8354">bpo-8354</a>.)
</li>
<li>New functions: in the <a class="reference internal" href="../library/site.html#module-site" title="site: Module responsible for site-specific configuration."><code class="xref py py-mod docutils literal notranslate">site</code></a> module, three new functions
return various site- and user-specific paths.
<a class="reference internal" href="../library/site.html#site.getsitepackages" title="site.getsitepackages"><code class="xref py py-func docutils literal notranslate">getsitepackages()</code></a> returns a list containing all
global site-packages directories,
<a class="reference internal" href="../library/site.html#site.getusersitepackages" title="site.getusersitepackages"><code class="xref py py-func docutils literal notranslate">getusersitepackages()</code></a> returns the path of the user’s
site-packages directory, and
<a class="reference internal" href="../library/site.html#site.getuserbase" title="site.getuserbase"><code class="xref py py-func docutils literal notranslate">getuserbase()</code></a> returns the value of the <code class="xref std std-envvar docutils literal notranslate">USER_BASE</code>
environment variable, giving the path to a directory that can be used
to store data.
(Contributed by Tarek Ziadé; <a class="reference external" href="https://bugs.python.org/issue6693">bpo-6693</a>.)
The <a class="reference internal" href="../library/site.html#module-site" title="site: Module responsible for site-specific configuration."><code class="xref py py-mod docutils literal notranslate">site</code></a> module now reports exceptions occurring
when the <code class="xref py py-mod docutils literal notranslate">sitecustomize</code> module is imported, and will no longer
catch and swallow the <a class="reference internal" href="../library/exceptions.html#KeyboardInterrupt" title="KeyboardInterrupt"><code class="xref py py-exc docutils literal notranslate">KeyboardInterrupt</code></a> exception. (Fixed by
Victor Stinner; <a class="reference external" href="https://bugs.python.org/issue3137">bpo-3137</a>.)
</li>
<li>The <a class="reference internal" href="../library/socket.html#socket.create_connection" title="socket.create_connection"><code class="xref py py-func docutils literal notranslate">create_connection()</code></a> function
gained a source_address parameter, a <code class="docutils literal notranslate">(host, port)</code> 2-tuple
giving the source address that will be used for the connection.
(Contributed by Eldon Ziegler; <a class="reference external" href="https://bugs.python.org/issue3972">bpo-3972</a>.)
The <a class="reference internal" href="../library/socket.html#socket.socket.recv_into" title="socket.socket.recv_into"><code class="xref py py-meth docutils literal notranslate">recv_into()</code></a> and <a class="reference internal" href="../library/socket.html#socket.socket.recvfrom_into" title="socket.socket.recvfrom_into"><code class="xref py py-meth docutils literal notranslate">recvfrom_into()</code></a>
methods will now write into objects that support the buffer API, most usefully
the <a class="reference internal" href="../library/stdtypes.html#bytearray" title="bytearray"><code class="xref py py-class docutils literal notranslate">bytearray</code></a> and <a class="reference internal" href="../library/stdtypes.html#memoryview" title="memoryview"><code class="xref py py-class docutils literal notranslate">memoryview</code></a> objects. (Implemented by
Antoine Pitrou; <a class="reference external" href="https://bugs.python.org/issue8104">bpo-8104</a>.)
</li>
<li>The <code class="xref py py-mod docutils literal notranslate">SocketServer</code> module’s <code class="xref py py-class docutils literal notranslate">TCPServer</code> class now
supports socket timeouts and disabling the Nagle algorithm.
The <code class="xref py py-attr docutils literal notranslate">disable_nagle_algorithm</code> class attribute
defaults to <code class="docutils literal notranslate">False</code>; if overridden to be true,
new request connections will have the TCP_NODELAY option set to
prevent buffering many small sends into a single TCP packet.
The <code class="xref py py-attr docutils literal notranslate">timeout</code> class attribute can hold
a timeout in seconds that will be applied to the request socket; if
no request is received within that time, <code class="xref py py-meth docutils literal notranslate">handle_timeout()</code>
will be called and <code class="xref py py-meth docutils literal notranslate">handle_request()</code> will return.
(Contributed by Kristján Valur Jónsson; <a class="reference external" href="https://bugs.python.org/issue6192">bpo-6192</a> and <a class="reference external" href="https://bugs.python.org/issue6267">bpo-6267</a>.)
</li>
<li>Updated module: the <a class="reference internal" href="../library/sqlite3.html#module-sqlite3" title="sqlite3: A DB-API 2.0 implementation using SQLite 3.x."><code class="xref py py-mod docutils literal notranslate">sqlite3</code></a> module has been updated to
version 2.6.0 of the <a class="reference external" href="https://github.com/ghaering/pysqlite">pysqlite package</a>. Version 2.6.0 includes a number of bugfixes, and adds
the ability to load SQLite extensions from shared libraries.
Call the <code class="docutils literal notranslate">enable_load_extension(True)</code> method to enable extensions,
and then call <a class="reference internal" href="../library/sqlite3.html#sqlite3.Connection.load_extension" title="sqlite3.Connection.load_extension"><code class="xref py py-meth docutils literal notranslate">load_extension()</code></a> to load a particular shared library.
(Updated by Gerhard Häring.)
</li>
<li>The <a class="reference internal" href="../library/ssl.html#module-ssl" title="ssl: TLS/SSL wrapper for socket objects"><code class="xref py py-mod docutils literal notranslate">ssl</code></a> module’s <a class="reference internal" href="../library/ssl.html#ssl.SSLSocket" title="ssl.SSLSocket"><code class="xref py py-class docutils literal notranslate">SSLSocket</code></a> objects now support the
buffer API, which fixed a test suite failure (fix by Antoine Pitrou;
<a class="reference external" href="https://bugs.python.org/issue7133">bpo-7133</a>) and automatically set
OpenSSL’s <code class="xref c c-macro docutils literal notranslate">SSL_MODE_AUTO_RETRY</code>, which will prevent an error
code being returned from <code class="xref py py-meth docutils literal notranslate">recv()</code> operations that trigger an SSL
renegotiation (fix by Antoine Pitrou; <a class="reference external" href="https://bugs.python.org/issue8222">bpo-8222</a>).
The <a class="reference internal" href="../library/ssl.html#ssl.wrap_socket" title="ssl.wrap_socket"><code class="xref py py-func docutils literal notranslate">ssl.wrap_socket()</code></a> constructor function now takes a
ciphers argument that’s a string listing the encryption algorithms
to be allowed; the format of the string is described
<a class="reference external" href="https://www.openssl.org/docs/manmaster/man1/ciphers.html#CIPHER-LIST-FORMAT">in the OpenSSL documentation</a>.
(Added by Antoine Pitrou; <a class="reference external" href="https://bugs.python.org/issue8322">bpo-8322</a>.)
Another change makes the extension load all of OpenSSL’s ciphers and
digest algorithms so that they’re all available. Some SSL
certificates couldn’t be verified, reporting an “unknown algorithm”
error. (Reported by Beda Kosata, and fixed by Antoine Pitrou;
<a class="reference external" href="https://bugs.python.org/issue8484">bpo-8484</a>.)
The version of OpenSSL being used is now available as the module
attributes <a class="reference internal" href="../library/ssl.html#ssl.OPENSSL_VERSION" title="ssl.OPENSSL_VERSION"><code class="xref py py-data docutils literal notranslate">ssl.OPENSSL_VERSION</code></a> (a string),
<a class="reference internal" href="../library/ssl.html#ssl.OPENSSL_VERSION_INFO" title="ssl.OPENSSL_VERSION_INFO"><code class="xref py py-data docutils literal notranslate">ssl.OPENSSL_VERSION_INFO</code></a> (a 5-tuple), and
<a class="reference internal" href="../library/ssl.html#ssl.OPENSSL_VERSION_NUMBER" title="ssl.OPENSSL_VERSION_NUMBER"><code class="xref py py-data docutils literal notranslate">ssl.OPENSSL_VERSION_NUMBER</code></a> (an integer). (Added by Antoine
Pitrou; <a class="reference external" href="https://bugs.python.org/issue8321">bpo-8321</a>.)
</li>
<li>The <a class="reference internal" href="../library/struct.html#module-struct" title="struct: Interpret bytes as packed binary data."><code class="xref py py-mod docutils literal notranslate">struct</code></a> module will no longer silently ignore overflow
errors when a value is too large for a particular integer format
code (one of <code class="docutils literal notranslate">bBhHiIlLqQ</code>); it now always raises a
<a class="reference internal" href="../library/struct.html#struct.error" title="struct.error"><code class="xref py py-exc docutils literal notranslate">struct.error</code></a> exception. (Changed by Mark Dickinson;
<a class="reference external" href="https://bugs.python.org/issue1523">bpo-1523</a>.) The <a class="reference internal" href="../library/struct.html#struct.pack" title="struct.pack"><code class="xref py py-func docutils literal notranslate">pack()</code></a> function will also
attempt to use <a class="reference internal" href="../reference/datamodel.html#object.__index__" title="object.__index__"><code class="xref py py-meth docutils literal notranslate">__index__()</code></a> to convert and pack non-integers
before trying the <a class="reference internal" href="../reference/datamodel.html#object.__int__" title="object.__int__"><code class="xref py py-meth docutils literal notranslate">__int__()</code></a> method or reporting an error.
(Changed by Mark Dickinson; <a class="reference external" href="https://bugs.python.org/issue8300">bpo-8300</a>.)
</li>
<li>New function: the <a class="reference internal" href="../library/subprocess.html#module-subprocess" title="subprocess: Subprocess management."><code class="xref py py-mod docutils literal notranslate">subprocess</code></a> module’s
<a class="reference internal" href="../library/subprocess.html#subprocess.check_output" title="subprocess.check_output"><code class="xref py py-func docutils literal notranslate">check_output()</code></a> runs a command with a specified set of arguments
and returns the command’s output as a string when the command runs without
error, or raises a <a class="reference internal" href="../library/subprocess.html#subprocess.CalledProcessError" title="subprocess.CalledProcessError"><code class="xref py py-exc docutils literal notranslate">CalledProcessError</code></a> exception otherwise.
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; subprocess.check_output([&#39;df&#39;, &#39;-h&#39;, &#39;.&#39;])
&#39;Filesystem Size Used Avail Capacity Mounted on\n
/dev/disk0s2 52G 49G 3.0G 94% /\n&#39;

&gt;&gt;&gt; subprocess.check_output([&#39;df&#39;, &#39;-h&#39;, &#39;/bogus&#39;])
 ...
subprocess.CalledProcessError: Command &#39;[&#39;df&#39;, &#39;-h&#39;, &#39;/bogus&#39;]&#39; returned non-zero exit status 1
</pre></div>
</div>
(Contributed by Gregory P. Smith.)
The <a class="reference internal" href="../library/subprocess.html#module-subprocess" title="subprocess: Subprocess management."><code class="xref py py-mod docutils literal notranslate">subprocess</code></a> module will now retry its internal system calls
on receiving an <code class="xref py py-const docutils literal notranslate">EINTR</code> signal. (Reported by several people; final
patch by Gregory P. Smith in <a class="reference external" href="https://bugs.python.org/issue1068268">bpo-1068268</a>.)
</li>
<li>New function: <a class="reference internal" href="../library/symtable.html#symtable.Symbol.is_declared_global" title="symtable.Symbol.is_declared_global"><code class="xref py py-func docutils literal notranslate">is_declared_global()</code></a> in the <a class="reference internal" href="../library/symtable.html#module-symtable" title="symtable: Interface to the compiler's internal symbol tables."><code class="xref py py-mod docutils literal notranslate">symtable</code></a> module
returns true for variables that are explicitly declared to be global,
false for ones that are implicitly global.
(Contributed by Jeremy Hylton.)
</li>
<li>The <a class="reference internal" href="../library/syslog.html#module-syslog" title="syslog: An interface to the Unix syslog library routines. (Unix)"><code class="xref py py-mod docutils literal notranslate">syslog</code></a> module will now use the value of <code class="docutils literal notranslate">sys.argv[0]</code> as the
identifier instead of the previous default value of <code class="docutils literal notranslate">'python'</code>.
(Changed by Sean Reifschneider; <a class="reference external" href="https://bugs.python.org/issue8451">bpo-8451</a>.)
</li>
<li>The <code class="docutils literal notranslate">sys.version_info</code> value is now a named tuple, with attributes
named <code class="xref py py-attr docutils literal notranslate">major</code>, <code class="xref py py-attr docutils literal notranslate">minor</code>, <code class="xref py py-attr docutils literal notranslate">micro</code>,
<code class="xref py py-attr docutils literal notranslate">releaselevel</code>, and <code class="xref py py-attr docutils literal notranslate">serial</code>. (Contributed by Ross
Light; <a class="reference external" href="https://bugs.python.org/issue4285">bpo-4285</a>.)
<a class="reference internal" href="../library/sys.html#sys.getwindowsversion" title="sys.getwindowsversion"><code class="xref py py-func docutils literal notranslate">sys.getwindowsversion()</code></a> also returns a named tuple,
with attributes named <code class="xref py py-attr docutils literal notranslate">major</code>, <code class="xref py py-attr docutils literal notranslate">minor</code>, <code class="xref py py-attr docutils literal notranslate">build</code>,
<a class="reference internal" href="../library/platform.html#module-platform" title="platform: Retrieves as much platform identifying data as possible."><code class="xref py py-attr docutils literal notranslate">platform</code></a>, <code class="xref py py-attr docutils literal notranslate">service_pack</code>, <code class="xref py py-attr docutils literal notranslate">service_pack_major</code>,
<code class="xref py py-attr docutils literal notranslate">service_pack_minor</code>, <code class="xref py py-attr docutils literal notranslate">suite_mask</code>, and
<code class="xref py py-attr docutils literal notranslate">product_type</code>. (Contributed by Brian Curtin; <a class="reference external" href="https://bugs.python.org/issue7766">bpo-7766</a>.)
</li>
<li>The <a class="reference internal" href="../library/tarfile.html#module-tarfile" title="tarfile: Read and write tar-format archive files."><code class="xref py py-mod docutils literal notranslate">tarfile</code></a> module’s default error handling has changed, to
no longer suppress fatal errors. The default error level was previously 0,
which meant that errors would only result in a message being written to the
debug log, but because the debug log is not activated by default,
these errors go unnoticed. The default error level is now 1,
which raises an exception if there’s an error.
(Changed by Lars Gustäbel; <a class="reference external" href="https://bugs.python.org/issue7357">bpo-7357</a>.)
<a class="reference internal" href="../library/tarfile.html#module-tarfile" title="tarfile: Read and write tar-format archive files."><code class="xref py py-mod docutils literal notranslate">tarfile</code></a> now supports filtering the <a class="reference internal" href="../library/tarfile.html#tarfile.TarInfo" title="tarfile.TarInfo"><code class="xref py py-class docutils literal notranslate">TarInfo</code></a>
objects being added to a tar file. When you call <a class="reference internal" href="../library/tarfile.html#tarfile.TarFile.add" title="tarfile.TarFile.add"><code class="xref py py-meth docutils literal notranslate">add()</code></a>,
you may supply an optional filter argument
that’s a callable. The filter callable will be passed the
<a class="reference internal" href="../library/tarfile.html#tarfile.TarInfo" title="tarfile.TarInfo"><code class="xref py py-class docutils literal notranslate">TarInfo</code></a> for every file being added, and can modify and return it.
If the callable returns <code class="docutils literal notranslate">None</code>, the file will be excluded from the
resulting archive. This is more powerful than the existing
exclude argument, which has therefore been deprecated.
(Added by Lars Gustäbel; <a class="reference external" href="https://bugs.python.org/issue6856">bpo-6856</a>.)
The <a class="reference internal" href="../library/tarfile.html#tarfile.TarFile" title="tarfile.TarFile"><code class="xref py py-class docutils literal notranslate">TarFile</code></a> class also now supports the context management protocol.
(Added by Lars Gustäbel; <a class="reference external" href="https://bugs.python.org/issue7232">bpo-7232</a>.)
</li>
<li>The <a class="reference internal" href="../library/threading.html#threading.Event.wait" title="threading.Event.wait"><code class="xref py py-meth docutils literal notranslate">wait()</code></a> method of the <a class="reference internal" href="../library/threading.html#threading.Event" title="threading.Event"><code class="xref py py-class docutils literal notranslate">threading.Event</code></a> class
now returns the internal flag on exit. This means the method will usually
return true because <a class="reference internal" href="../library/threading.html#threading.Event.wait" title="threading.Event.wait"><code class="xref py py-meth docutils literal notranslate">wait()</code></a> is supposed to block until the
internal flag becomes true. The return value will only be false if
a timeout was provided and the operation timed out.
(Contributed by Tim Lesher; <a class="reference external" href="https://bugs.python.org/issue1674032">bpo-1674032</a>.)
</li>
<li>The Unicode database provided by the <a class="reference internal" href="../library/unicodedata.html#module-unicodedata" title="unicodedata: Access the Unicode Database."><code class="xref py py-mod docutils literal notranslate">unicodedata</code></a> module is
now used internally to determine which characters are numeric,
whitespace, or represent line breaks. The database also
includes information from the <code class="file docutils literal notranslate">Unihan.txt</code> data file (patch
by Anders Chrigström and Amaury Forgeot d’Arc; <a class="reference external" href="https://bugs.python.org/issue1571184">bpo-1571184</a>)
and has been updated to version 5.2.0 (updated by
Florent Xicluna; <a class="reference external" href="https://bugs.python.org/issue8024">bpo-8024</a>).
</li>
<li>The <code class="xref py py-mod docutils literal notranslate">urlparse</code> module’s <code class="xref py py-func docutils literal notranslate">urlsplit()</code> now handles
unknown URL schemes in a fashion compliant with <a class="rfc reference external" href="https://tools.ietf.org/html/rfc3986.html">RFC 3986</a>: if the
URL is of the form <code class="docutils literal notranslate">&quot;&lt;something&gt;://...&quot;</code>, the text before the
<code class="docutils literal notranslate">://</code> is treated as the scheme, even if it’s a made-up scheme that
the module doesn’t know about. This change may break code that
worked around the old behaviour. For example, Python 2.6.4 or 2.5
will return the following:
<div class="highlight-pycon notranslate"><div class="highlight"><pre>&gt;&gt;&gt; import urlparse
&gt;&gt;&gt; urlparse.urlsplit(&#39;invented://host/filename?query&#39;)
(&#39;invented&#39;, &#39;&#39;, &#39;//host/filename?query&#39;, &#39;&#39;, &#39;&#39;)
</pre></div>
</div>
Python 2.7 (and Python 2.6.5) will return:
<div class="highlight-pycon notranslate"><div class="highlight"><pre>&gt;&gt;&gt; import urlparse
&gt;&gt;&gt; urlparse.urlsplit(&#39;invented://host/filename?query&#39;)
(&#39;invented&#39;, &#39;host&#39;, &#39;/filename?query&#39;, &#39;&#39;, &#39;&#39;)
</pre></div>
</div>
(Python 2.7 actually produces slightly different output, since it
returns a named tuple instead of a standard tuple.)
The <code class="xref py py-mod docutils literal notranslate">urlparse</code> module also supports IPv6 literal addresses as defined by
<a class="rfc reference external" href="https://tools.ietf.org/html/rfc2732.html">RFC 2732</a> (contributed by Senthil Kumaran; <a class="reference external" href="https://bugs.python.org/issue2987">bpo-2987</a>).
<div class="highlight-pycon notranslate"><div class="highlight"><pre>&gt;&gt;&gt; urlparse.urlparse(&#39;http://[1080::8:800:200C:417A]/foo&#39;)
ParseResult(scheme=&#39;http&#39;, netloc=&#39;[1080::8:800:200C:417A]&#39;,
 path=&#39;/foo&#39;, params=&#39;&#39;, query=&#39;&#39;, fragment=&#39;&#39;)
</pre></div>
</div>
</li>
<li>New class: the <a class="reference internal" href="../library/weakref.html#weakref.WeakSet" title="weakref.WeakSet"><code class="xref py py-class docutils literal notranslate">WeakSet</code></a> class in the <a class="reference internal" href="../library/weakref.html#module-weakref" title="weakref: Support for weak references and weak dictionaries."><code class="xref py py-mod docutils literal notranslate">weakref</code></a>
module is a set that only holds weak references to its elements; elements
will be removed once there are no references pointing to them.
(Originally implemented in Python 3.x by Raymond Hettinger, and backported
to 2.7 by Michael Foord.)
</li>
<li>The ElementTree library, <code class="xref py py-mod docutils literal notranslate">xml.etree</code>, no longer escapes
ampersands and angle brackets when outputting an XML processing
instruction (which looks like <code class="docutils literal notranslate">&lt;?xml-stylesheet href=&quot;#style1&quot;?&gt;</code>)
or comment (which looks like <code class="docutils literal notranslate">&lt;!-- comment --&gt;</code>).
(Patch by Neil Muller; <a class="reference external" href="https://bugs.python.org/issue2746">bpo-2746</a>.)
</li>
<li>The XML-RPC client and server, provided by the <code class="xref py py-mod docutils literal notranslate">xmlrpclib</code> and
<code class="xref py py-mod docutils literal notranslate">SimpleXMLRPCServer</code> modules, have improved performance by
supporting HTTP/1.1 keep-alive and by optionally using gzip encoding
to compress the XML being exchanged. The gzip compression is
controlled by the <code class="xref py py-attr docutils literal notranslate">encode_threshold</code> attribute of
<code class="xref py py-class docutils literal notranslate">SimpleXMLRPCRequestHandler</code>, which contains a size in bytes;
responses larger than this will be compressed.
(Contributed by Kristján Valur Jónsson; <a class="reference external" href="https://bugs.python.org/issue6267">bpo-6267</a>.)
</li>
<li>The <a class="reference internal" href="../library/zipfile.html#module-zipfile" title="zipfile: Read and write ZIP-format archive files."><code class="xref py py-mod docutils literal notranslate">zipfile</code></a> module’s <a class="reference internal" href="../library/zipfile.html#zipfile.ZipFile" title="zipfile.ZipFile"><code class="xref py py-class docutils literal notranslate">ZipFile</code></a> now supports the context
management protocol, so you can write <code class="docutils literal notranslate">with zipfile.ZipFile(...) as f:</code>.
(Contributed by Brian Curtin; <a class="reference external" href="https://bugs.python.org/issue5511">bpo-5511</a>.)
<a class="reference internal" href="../library/zipfile.html#module-zipfile" title="zipfile: Read and write ZIP-format archive files."><code class="xref py py-mod docutils literal notranslate">zipfile</code></a> now also supports archiving empty directories and
extracts them correctly. (Fixed by Kuba Wieczorek; <a class="reference external" href="https://bugs.python.org/issue4710">bpo-4710</a>.)
Reading files out of an archive is faster, and interleaving
<a class="reference internal" href="../library/zipfile.html#zipfile.ZipFile.read" title="zipfile.ZipFile.read"><code class="xref py py-meth docutils literal notranslate">read()</code></a> and <code class="xref py py-meth docutils literal notranslate">readline()</code> now works correctly.
(Contributed by Nir Aides; <a class="reference external" href="https://bugs.python.org/issue7610">bpo-7610</a>.)
The <a class="reference internal" href="../library/zipfile.html#zipfile.is_zipfile" title="zipfile.is_zipfile"><code class="xref py py-func docutils literal notranslate">is_zipfile()</code></a> function now
accepts a file object, in addition to the path names accepted in earlier
versions. (Contributed by Gabriel Genellina; <a class="reference external" href="https://bugs.python.org/issue4756">bpo-4756</a>.)
The <a class="reference internal" href="../library/zipfile.html#zipfile.ZipFile.writestr" title="zipfile.ZipFile.writestr"><code class="xref py py-meth docutils literal notranslate">writestr()</code></a> method now has an optional compress_type parameter
that lets you override the default compression method specified in the
<a class="reference internal" href="../library/zipfile.html#zipfile.ZipFile" title="zipfile.ZipFile"><code class="xref py py-class docutils literal notranslate">ZipFile</code></a> constructor. (Contributed by Ronald Oussoren;
<a class="reference external" href="https://bugs.python.org/issue6003">bpo-6003</a>.)
</li>
</ul>
<div class="section" id="new-module-importlib">
<h3>New module: importlib<a class="headerlink" href="#new-module-importlib" title="Permalink to this headline">¶</a></h3>
Python 3.1 includes the <a class="reference internal" href="../library/importlib.html#module-importlib" title="importlib: The implementation of the import machinery."><code class="xref py py-mod docutils literal notranslate">importlib</code></a> package, a re-implementation
of the logic underlying Python’s <a class="reference internal" href="../reference/simple_stmts.html#import"><code class="xref std std-keyword docutils literal notranslate">import</code></a> statement.
<a class="reference internal" href="../library/importlib.html#module-importlib" title="importlib: The implementation of the import machinery."><code class="xref py py-mod docutils literal notranslate">importlib</code></a> is useful for implementors of Python interpreters and
to users who wish to write new importers that can participate in the
import process. Python 2.7 doesn’t contain the complete
<a class="reference internal" href="../library/importlib.html#module-importlib" title="importlib: The implementation of the import machinery."><code class="xref py py-mod docutils literal notranslate">importlib</code></a> package, but instead has a tiny subset that contains
a single function, <a class="reference internal" href="../library/importlib.html#importlib.import_module" title="importlib.import_module"><code class="xref py py-func docutils literal notranslate">import_module()</code></a>.
<code class="docutils literal notranslate">import_module(name, package=None)</code> imports a module. name is
a string containing the module or package’s name. It’s possible to do
relative imports by providing a string that begins with a <code class="docutils literal notranslate">.</code>
character, such as <code class="docutils literal notranslate">..utils.errors</code>. For relative imports, the
package argument must be provided and is the name of the package that
will be used as the anchor for
the relative import. <a class="reference internal" href="../library/importlib.html#importlib.import_module" title="importlib.import_module"><code class="xref py py-func docutils literal notranslate">import_module()</code></a> both inserts the imported
module into <code class="docutils literal notranslate">sys.modules</code> and returns the module object.
Here are some examples:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>&gt;&gt;&gt; from importlib import import_module
&gt;&gt;&gt; anydbm = import_module(&#39;anydbm&#39;) # Standard absolute import
&gt;&gt;&gt; anydbm
&lt;module &#39;anydbm&#39; from &#39;/p/python/Lib/anydbm.py&#39;&gt;
&gt;&gt;&gt; # Relative import
&gt;&gt;&gt; file_util = import_module(&#39;..file_util&#39;, &#39;distutils.command&#39;)
&gt;&gt;&gt; file_util
&lt;module &#39;distutils.file_util&#39; from &#39;/python/Lib/distutils/file_util.pyc&#39;&gt;
</pre></div>
</div>
<a class="reference internal" href="../library/importlib.html#module-importlib" title="importlib: The implementation of the import machinery."><code class="xref py py-mod docutils literal notranslate">importlib</code></a> was implemented by Brett Cannon and introduced in
Python 3.1.
</div>
<div class="section" id="new-module-sysconfig">
<h3>New module: sysconfig<a class="headerlink" href="#new-module-sysconfig" title="Permalink to this headline">¶</a></h3>
The <a class="reference internal" href="../library/sysconfig.html#module-sysconfig" title="sysconfig: Python's configuration information"><code class="xref py py-mod docutils literal notranslate">sysconfig</code></a> module has been pulled out of the Distutils
package, becoming a new top-level module in its own right.
<a class="reference internal" href="../library/sysconfig.html#module-sysconfig" title="sysconfig: Python's configuration information"><code class="xref py py-mod docutils literal notranslate">sysconfig</code></a> provides functions for getting information about
Python’s build process: compiler switches, installation paths, the
platform name, and whether Python is running from its source
directory.
Some of the functions in the module are:
<ul class="simple">
<li><a class="reference internal" href="../library/sysconfig.html#sysconfig.get_config_var" title="sysconfig.get_config_var"><code class="xref py py-func docutils literal notranslate">get_config_var()</code></a> returns variables from Python’s
Makefile and the <code class="file docutils literal notranslate">pyconfig.h</code> file.</li>
<li><a class="reference internal" href="../library/sysconfig.html#sysconfig.get_config_vars" title="sysconfig.get_config_vars"><code class="xref py py-func docutils literal notranslate">get_config_vars()</code></a> returns a dictionary containing
all of the configuration variables.</li>
<li><a class="reference internal" href="../library/sysconfig.html#sysconfig.get_path" title="sysconfig.get_path"><code class="xref py py-func docutils literal notranslate">get_path()</code></a> returns the configured path for
a particular type of module: the standard library,
site-specific modules, platform-specific modules, etc.</li>
<li><a class="reference internal" href="../library/sysconfig.html#sysconfig.is_python_build" title="sysconfig.is_python_build"><code class="xref py py-func docutils literal notranslate">is_python_build()</code></a> returns true if you’re running a
binary from a Python source tree, and false otherwise.</li>
</ul>
Consult the <a class="reference internal" href="../library/sysconfig.html#module-sysconfig" title="sysconfig: Python's configuration information"><code class="xref py py-mod docutils literal notranslate">sysconfig</code></a> documentation for more details and for
a complete list of functions.
The Distutils package and <a class="reference internal" href="../library/sysconfig.html#module-sysconfig" title="sysconfig: Python's configuration information"><code class="xref py py-mod docutils literal notranslate">sysconfig</code></a> are now maintained by Tarek
Ziadé, who has also started a Distutils2 package (source repository at
<a class="reference external" href="https://hg.python.org/distutils2/">https://hg.python.org/distutils2/</a>) for developing a next-generation
version of Distutils.
</div>
<div class="section" id="ttk-themed-widgets-for-tk">
<h3>ttk: Themed Widgets for Tk<a class="headerlink" href="#ttk-themed-widgets-for-tk" title="Permalink to this headline">¶</a></h3>
Tcl/Tk 8.5 includes a set of themed widgets that re-implement basic Tk
widgets but have a more customizable appearance and can therefore more
closely resemble the native platform’s widgets. This widget
set was originally called Tile, but was renamed to Ttk (for “themed Tk”)
on being added to Tcl/Tck release 8.5.
To learn more, read the <code class="xref py py-mod docutils literal notranslate">ttk</code> module documentation. You may also
wish to read the Tcl/Tk manual page describing the
Ttk theme engine, available at
<a class="reference external" href="https://www.tcl.tk/man/tcl8.5/TkCmd/ttk_intro.htm">https://www.tcl.tk/man/tcl8.5/TkCmd/ttk_intro.htm</a>. Some
screenshots of the Python/Ttk code in use are at
<a class="reference external" href="https://code.google.com/archive/p/python-ttk/wikis/Screenshots.wiki">https://code.google.com/archive/p/python-ttk/wikis/Screenshots.wiki</a>.
The <code class="xref py py-mod docutils literal notranslate">ttk</code> module was written by Guilherme Polo and added in
<a class="reference external" href="https://bugs.python.org/issue2983">bpo-2983</a>. An alternate version called <code class="docutils literal notranslate">Tile.py</code>, written by
Martin Franklin and maintained by Kevin Walzer, was proposed for
inclusion in <a class="reference external" href="https://bugs.python.org/issue2618">bpo-2618</a>, but the authors argued that Guilherme
Polo’s work was more comprehensive.
</div>
<div class="section" id="updated-module-unittest">
<h3>Updated module: unittest<a class="headerlink" href="#updated-module-unittest" title="Permalink to this headline">¶</a></h3>
The <a class="reference internal" href="../library/unittest.html#module-unittest" title="unittest: Unit testing framework for Python."><code class="xref py py-mod docutils literal notranslate">unittest</code></a> module was greatly enhanced; many
new features were added. Most of these features were implemented
by Michael Foord, unless otherwise noted. The enhanced version of
the module is downloadable separately for use with Python versions 2.4 to 2.6,
packaged as the <code class="xref py py-mod docutils literal notranslate">unittest2</code> package, from
<a class="reference external" href="https://pypi.org/project/unittest2">https://pypi.org/project/unittest2</a>.
When used from the command line, the module can automatically discover
tests. It’s not as fancy as <a class="reference external" href="http://pytest.org">py.test</a> or
<a class="reference external" href="https://nose.readthedocs.io/">nose</a>, but provides a
simple way to run tests kept within a set of package directories. For example,
the following command will search the <code class="file docutils literal notranslate">test/</code> subdirectory for
any importable test files named <code class="docutils literal notranslate">test*.py</code>:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>python -m unittest discover -s test
</pre></div>
</div>
Consult the <a class="reference internal" href="../library/unittest.html#module-unittest" title="unittest: Unit testing framework for Python."><code class="xref py py-mod docutils literal notranslate">unittest</code></a> module documentation for more details.
(Developed in <a class="reference external" href="https://bugs.python.org/issue6001">bpo-6001</a>.)
The <a class="reference internal" href="../library/unittest.html#unittest.main" title="unittest.main"><code class="xref py py-func docutils literal notranslate">main()</code></a> function supports some other new options:
<ul>
<li><a class="reference internal" href="../library/unittest.html#cmdoption-unittest-b"><code class="xref std std-option docutils literal notranslate">-b</code></a> or <code class="xref std std-option docutils literal notranslate">--buffer</code> will buffer the standard output
and standard error streams during each test. If the test passes,
any resulting output will be discarded; on failure, the buffered
output will be displayed.
</li>
<li><a class="reference internal" href="../library/unittest.html#cmdoption-unittest-c"><code class="xref std std-option docutils literal notranslate">-c</code></a> or <code class="xref std std-option docutils literal notranslate">--catch</code> will cause the control-C interrupt
to be handled more gracefully. Instead of interrupting the test
process immediately, the currently running test will be completed
and then the partial results up to the interruption will be reported.
If you’re impatient, a second press of control-C will cause an immediate
interruption.
This control-C handler tries to avoid causing problems when the code
being tested or the tests being run have defined a signal handler of
their own, by noticing that a signal handler was already set and
calling it. If this doesn’t work for you, there’s a
<a class="reference internal" href="../library/unittest.html#unittest.removeHandler" title="unittest.removeHandler"><code class="xref py py-func docutils literal notranslate">removeHandler()</code></a> decorator that can be used to mark tests that
should have the control-C handling disabled.
</li>
<li><a class="reference internal" href="../library/unittest.html#cmdoption-unittest-f"><code class="xref std std-option docutils literal notranslate">-f</code></a> or <code class="xref std std-option docutils literal notranslate">--failfast</code> makes
test execution stop immediately when a test fails instead of
continuing to execute further tests. (Suggested by Cliff Dyer and
implemented by Michael Foord; <a class="reference external" href="https://bugs.python.org/issue8074">bpo-8074</a>.)
</li>
</ul>
The progress messages now show ‘x’ for expected failures
and ‘u’ for unexpected successes when run in verbose mode.
(Contributed by Benjamin Peterson.)
Test cases can raise the <a class="reference internal" href="../library/unittest.html#unittest.SkipTest" title="unittest.SkipTest"><code class="xref py py-exc docutils literal notranslate">SkipTest</code></a> exception to skip a
test (<a class="reference external" href="https://bugs.python.org/issue1034053">bpo-1034053</a>).
The error messages for <a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertEqual" title="unittest.TestCase.assertEqual"><code class="xref py py-meth docutils literal notranslate">assertEqual()</code></a>,
<a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertTrue" title="unittest.TestCase.assertTrue"><code class="xref py py-meth docutils literal notranslate">assertTrue()</code></a>, and <a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertFalse" title="unittest.TestCase.assertFalse"><code class="xref py py-meth docutils literal notranslate">assertFalse()</code></a>
failures now provide more information. If you set the
<a class="reference internal" href="../library/unittest.html#unittest.TestCase.longMessage" title="unittest.TestCase.longMessage"><code class="xref py py-attr docutils literal notranslate">longMessage</code></a> attribute of your <a class="reference internal" href="../library/unittest.html#unittest.TestCase" title="unittest.TestCase"><code class="xref py py-class docutils literal notranslate">TestCase</code></a> classes to
true, both the standard error message and any additional message you
provide will be printed for failures. (Added by Michael Foord; <a class="reference external" href="https://bugs.python.org/issue5663">bpo-5663</a>.)
The <a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertRaises" title="unittest.TestCase.assertRaises"><code class="xref py py-meth docutils literal notranslate">assertRaises()</code></a> method now
returns a context handler when called without providing a callable
object to run. For example, you can write this:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>with self.assertRaises(KeyError):
 {}[&#39;foo&#39;]
</pre></div>
</div>
(Implemented by Antoine Pitrou; <a class="reference external" href="https://bugs.python.org/issue4444">bpo-4444</a>.)
Module- and class-level setup and teardown fixtures are now supported.
Modules can contain <code class="xref py py-func docutils literal notranslate">setUpModule()</code> and <code class="xref py py-func docutils literal notranslate">tearDownModule()</code>
functions. Classes can have <a class="reference internal" href="../library/unittest.html#unittest.TestCase.setUpClass" title="unittest.TestCase.setUpClass"><code class="xref py py-meth docutils literal notranslate">setUpClass()</code></a> and
<a class="reference internal" href="../library/unittest.html#unittest.TestCase.tearDownClass" title="unittest.TestCase.tearDownClass"><code class="xref py py-meth docutils literal notranslate">tearDownClass()</code></a> methods that must be defined as class methods
(using <code class="docutils literal notranslate">&#64;classmethod</code> or equivalent). These functions and
methods are invoked when the test runner switches to a test case in a
different module or class.
The methods <a class="reference internal" href="../library/unittest.html#unittest.TestCase.addCleanup" title="unittest.TestCase.addCleanup"><code class="xref py py-meth docutils literal notranslate">addCleanup()</code></a> and
<a class="reference internal" href="../library/unittest.html#unittest.TestCase.doCleanups" title="unittest.TestCase.doCleanups"><code class="xref py py-meth docutils literal notranslate">doCleanups()</code></a> were added.
<a class="reference internal" href="../library/unittest.html#unittest.TestCase.addCleanup" title="unittest.TestCase.addCleanup"><code class="xref py py-meth docutils literal notranslate">addCleanup()</code></a> lets you add cleanup functions that
will be called unconditionally (after <a class="reference internal" href="../library/unittest.html#unittest.TestCase.setUp" title="unittest.TestCase.setUp"><code class="xref py py-meth docutils literal notranslate">setUp()</code></a> if
<a class="reference internal" href="../library/unittest.html#unittest.TestCase.setUp" title="unittest.TestCase.setUp"><code class="xref py py-meth docutils literal notranslate">setUp()</code></a> fails, otherwise after <a class="reference internal" href="../library/unittest.html#unittest.TestCase.tearDown" title="unittest.TestCase.tearDown"><code class="xref py py-meth docutils literal notranslate">tearDown()</code></a>). This allows
for much simpler resource allocation and deallocation during tests
(<a class="reference external" href="https://bugs.python.org/issue5679">bpo-5679</a>).
A number of new methods were added that provide more specialized
tests. Many of these methods were written by Google engineers
for use in their test suites; Gregory P. Smith, Michael Foord, and
GvR worked on merging them into Python’s version of <a class="reference internal" href="../library/unittest.html#module-unittest" title="unittest: Unit testing framework for Python."><code class="xref py py-mod docutils literal notranslate">unittest</code></a>.
<ul class="simple">
<li><a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertIsNone" title="unittest.TestCase.assertIsNone"><code class="xref py py-meth docutils literal notranslate">assertIsNone()</code></a> and <a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertIsNotNone" title="unittest.TestCase.assertIsNotNone"><code class="xref py py-meth docutils literal notranslate">assertIsNotNone()</code></a> take one
expression and verify that the result is or is not <code class="docutils literal notranslate">None</code>.</li>
<li><a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertIs" title="unittest.TestCase.assertIs"><code class="xref py py-meth docutils literal notranslate">assertIs()</code></a> and <a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertIsNot" title="unittest.TestCase.assertIsNot"><code class="xref py py-meth docutils literal notranslate">assertIsNot()</code></a>
take two values and check whether the two values evaluate to the same object or not.
(Added by Michael Foord; <a class="reference external" href="https://bugs.python.org/issue2578">bpo-2578</a>.)</li>
<li><a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertIsInstance" title="unittest.TestCase.assertIsInstance"><code class="xref py py-meth docutils literal notranslate">assertIsInstance()</code></a> and
<a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertNotIsInstance" title="unittest.TestCase.assertNotIsInstance"><code class="xref py py-meth docutils literal notranslate">assertNotIsInstance()</code></a> check whether
the resulting object is an instance of a particular class, or of
one of a tuple of classes. (Added by Georg Brandl; <a class="reference external" href="https://bugs.python.org/issue7031">bpo-7031</a>.)</li>
<li><a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertGreater" title="unittest.TestCase.assertGreater"><code class="xref py py-meth docutils literal notranslate">assertGreater()</code></a>, <a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertGreaterEqual" title="unittest.TestCase.assertGreaterEqual"><code class="xref py py-meth docutils literal notranslate">assertGreaterEqual()</code></a>,
<a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertLess" title="unittest.TestCase.assertLess"><code class="xref py py-meth docutils literal notranslate">assertLess()</code></a>, and <a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertLessEqual" title="unittest.TestCase.assertLessEqual"><code class="xref py py-meth docutils literal notranslate">assertLessEqual()</code></a> compare
two quantities.</li>
<li><a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertMultiLineEqual" title="unittest.TestCase.assertMultiLineEqual"><code class="xref py py-meth docutils literal notranslate">assertMultiLineEqual()</code></a> compares two strings, and if they’re
not equal, displays a helpful comparison that highlights the
differences in the two strings. This comparison is now used by
default when Unicode strings are compared with <a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertEqual" title="unittest.TestCase.assertEqual"><code class="xref py py-meth docutils literal notranslate">assertEqual()</code></a>.</li>
<li><code class="xref py py-meth docutils literal notranslate">assertRegexpMatches()</code> and
<code class="xref py py-meth docutils literal notranslate">assertNotRegexpMatches()</code> checks whether the
first argument is a string matching or not matching the regular
expression provided as the second argument (<a class="reference external" href="https://bugs.python.org/issue8038">bpo-8038</a>).</li>
<li><code class="xref py py-meth docutils literal notranslate">assertRaisesRegexp()</code> checks whether a particular exception
is raised, and then also checks that the string representation of
the exception matches the provided regular expression.</li>
<li><a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertIn" title="unittest.TestCase.assertIn"><code class="xref py py-meth docutils literal notranslate">assertIn()</code></a> and <a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertNotIn" title="unittest.TestCase.assertNotIn"><code class="xref py py-meth docutils literal notranslate">assertNotIn()</code></a>
tests whether first is or is not in second.</li>
<li><code class="xref py py-meth docutils literal notranslate">assertItemsEqual()</code> tests whether two provided sequences
contain the same elements.</li>
<li><a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertSetEqual" title="unittest.TestCase.assertSetEqual"><code class="xref py py-meth docutils literal notranslate">assertSetEqual()</code></a> compares whether two sets are equal, and
only reports the differences between the sets in case of error.</li>
<li>Similarly, <a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertListEqual" title="unittest.TestCase.assertListEqual"><code class="xref py py-meth docutils literal notranslate">assertListEqual()</code></a> and <a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertTupleEqual" title="unittest.TestCase.assertTupleEqual"><code class="xref py py-meth docutils literal notranslate">assertTupleEqual()</code></a>
compare the specified types and explain any differences without necessarily
printing their full values; these methods are now used by default
when comparing lists and tuples using <a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertEqual" title="unittest.TestCase.assertEqual"><code class="xref py py-meth docutils literal notranslate">assertEqual()</code></a>.
More generally, <a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertSequenceEqual" title="unittest.TestCase.assertSequenceEqual"><code class="xref py py-meth docutils literal notranslate">assertSequenceEqual()</code></a> compares two sequences
and can optionally check whether both sequences are of a
particular type.</li>
<li><a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertDictEqual" title="unittest.TestCase.assertDictEqual"><code class="xref py py-meth docutils literal notranslate">assertDictEqual()</code></a> compares two dictionaries and reports the
differences; it’s now used by default when you compare two dictionaries
using <a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertEqual" title="unittest.TestCase.assertEqual"><code class="xref py py-meth docutils literal notranslate">assertEqual()</code></a>. <code class="xref py py-meth docutils literal notranslate">assertDictContainsSubset()</code> checks whether
all of the key/value pairs in first are found in second.</li>
<li><a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertAlmostEqual" title="unittest.TestCase.assertAlmostEqual"><code class="xref py py-meth docutils literal notranslate">assertAlmostEqual()</code></a> and <a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertNotAlmostEqual" title="unittest.TestCase.assertNotAlmostEqual"><code class="xref py py-meth docutils literal notranslate">assertNotAlmostEqual()</code></a> test
whether first and second are approximately equal. This method
can either round their difference to an optionally-specified number
of places (the default is 7) and compare it to zero, or require
the difference to be smaller than a supplied delta value.</li>
<li><a class="reference internal" href="../library/unittest.html#unittest.TestLoader.loadTestsFromName" title="unittest.TestLoader.loadTestsFromName"><code class="xref py py-meth docutils literal notranslate">loadTestsFromName()</code></a> properly honors the
<a class="reference internal" href="../library/unittest.html#unittest.TestLoader.suiteClass" title="unittest.TestLoader.suiteClass"><code class="xref py py-attr docutils literal notranslate">suiteClass</code></a> attribute of
the <a class="reference internal" href="../library/unittest.html#unittest.TestLoader" title="unittest.TestLoader"><code class="xref py py-class docutils literal notranslate">TestLoader</code></a>. (Fixed by Mark Roddy; <a class="reference external" href="https://bugs.python.org/issue6866">bpo-6866</a>.)</li>
<li>A new hook lets you extend the <a class="reference internal" href="../library/unittest.html#unittest.TestCase.assertEqual" title="unittest.TestCase.assertEqual"><code class="xref py py-meth docutils literal notranslate">assertEqual()</code></a> method to handle
new data types. The <a class="reference internal" href="../library/unittest.html#unittest.TestCase.addTypeEqualityFunc" title="unittest.TestCase.addTypeEqualityFunc"><code class="xref py py-meth docutils literal notranslate">addTypeEqualityFunc()</code></a> method takes a type
object and a function. The function will be used when both of the
objects being compared are of the specified type. This function
should compare the two objects and raise an exception if they don’t
match; it’s a good idea for the function to provide additional
information about why the two objects aren’t matching, much as the new
sequence comparison methods do.</li>
</ul>
<a class="reference internal" href="../library/unittest.html#unittest.main" title="unittest.main"><code class="xref py py-func docutils literal notranslate">unittest.main()</code></a> now takes an optional <code class="docutils literal notranslate">exit</code> argument. If
false, <a class="reference internal" href="../library/unittest.html#unittest.main" title="unittest.main"><code class="xref py py-func docutils literal notranslate">main()</code></a> doesn’t call <a class="reference internal" href="../library/sys.html#sys.exit" title="sys.exit"><code class="xref py py-func docutils literal notranslate">sys.exit()</code></a>, allowing
<a class="reference internal" href="../library/unittest.html#unittest.main" title="unittest.main"><code class="xref py py-func docutils literal notranslate">main()</code></a> to be used from the interactive interpreter.
(Contributed by J. Pablo Fernández; <a class="reference external" href="https://bugs.python.org/issue3379">bpo-3379</a>.)
<a class="reference internal" href="../library/unittest.html#unittest.TestResult" title="unittest.TestResult"><code class="xref py py-class docutils literal notranslate">TestResult</code></a> has new <a class="reference internal" href="../library/unittest.html#unittest.TestResult.startTestRun" title="unittest.TestResult.startTestRun"><code class="xref py py-meth docutils literal notranslate">startTestRun()</code></a> and
<a class="reference internal" href="../library/unittest.html#unittest.TestResult.stopTestRun" title="unittest.TestResult.stopTestRun"><code class="xref py py-meth docutils literal notranslate">stopTestRun()</code></a> methods that are called immediately before
and after a test run. (Contributed by Robert Collins; <a class="reference external" href="https://bugs.python.org/issue5728">bpo-5728</a>.)
With all these changes, the <code class="file docutils literal notranslate">unittest.py</code> was becoming awkwardly
large, so the module was turned into a package and the code split into
several files (by Benjamin Peterson). This doesn’t affect how the
module is imported or used.
<div class="admonition seealso">
See also
<dl class="last docutils">
<dt><a class="reference external" href="http://www.voidspace.org.uk/python/articles/unittest2.shtml">http://www.voidspace.org.uk/python/articles/unittest2.shtml</a></dt>
<dd>Describes the new features, how to use them, and the
rationale for various design decisions. (By Michael Foord.)</dd>
</dl>
</div>
</div>
<div class="section" id="updated-module-elementtree-1-3">
<h3>Updated module: ElementTree 1.3<a class="headerlink" href="#updated-module-elementtree-1-3" title="Permalink to this headline">¶</a></h3>
The version of the ElementTree library included with Python was updated to
version 1.3. Some of the new features are:
<ul>
<li>The various parsing functions now take a parser keyword argument
giving an <a class="reference internal" href="../library/xml.etree.elementtree.html#xml.etree.ElementTree.XMLParser" title="xml.etree.ElementTree.XMLParser"><code class="xref py py-class docutils literal notranslate">XMLParser</code></a> instance that will
be used. This makes it possible to override the file’s internal encoding:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>p = ET.XMLParser(encoding=&#39;utf-8&#39;)
t = ET.XML(&quot;&quot;&quot;&lt;root/&gt;&quot;&quot;&quot;, parser=p)
</pre></div>
</div>
Errors in parsing XML now raise a <code class="xref py py-exc docutils literal notranslate">ParseError</code> exception, whose
instances have a <code class="xref py py-attr docutils literal notranslate">position</code> attribute
containing a (line, column) tuple giving the location of the problem.
</li>
<li>ElementTree’s code for converting trees to a string has been
significantly reworked, making it roughly twice as fast in many
cases. The <a class="reference internal" href="../library/xml.etree.elementtree.html#xml.etree.ElementTree.ElementTree.write" title="xml.etree.ElementTree.ElementTree.write"><code class="xref py py-meth docutils literal notranslate">ElementTree.write()</code></a>
and <code class="xref py py-meth docutils literal notranslate">Element.write()</code> methods now have a method parameter that can be
“xml” (the default), “html”, or “text”. HTML mode will output empty
elements as <code class="docutils literal notranslate">&lt;empty&gt;&lt;/empty&gt;</code> instead of <code class="docutils literal notranslate">&lt;empty/&gt;</code>, and text
mode will skip over elements and only output the text chunks. If
you set the <code class="xref py py-attr docutils literal notranslate">tag</code> attribute of an element to <code class="docutils literal notranslate">None</code> but
leave its children in place, the element will be omitted when the
tree is written out, so you don’t need to do more extensive rearrangement
to remove a single element.
Namespace handling has also been improved. All <code class="docutils literal notranslate">xmlns:&lt;whatever&gt;</code>
declarations are now output on the root element, not scattered throughout
the resulting XML. You can set the default namespace for a tree
by setting the <code class="xref py py-attr docutils literal notranslate">default_namespace</code> attribute and can
register new prefixes with <a class="reference internal" href="../library/xml.etree.elementtree.html#xml.etree.ElementTree.register_namespace" title="xml.etree.ElementTree.register_namespace"><code class="xref py py-meth docutils literal notranslate">register_namespace()</code></a>. In XML mode,
you can use the true/false xml_declaration parameter to suppress the
XML declaration.
</li>
<li>New <a class="reference internal" href="../library/xml.etree.elementtree.html#xml.etree.ElementTree.Element" title="xml.etree.ElementTree.Element"><code class="xref py py-class docutils literal notranslate">Element</code></a> method:
<a class="reference internal" href="../library/xml.etree.elementtree.html#xml.etree.ElementTree.Element.extend" title="xml.etree.ElementTree.Element.extend"><code class="xref py py-meth docutils literal notranslate">extend()</code></a> appends the items from a
sequence to the element’s children. Elements themselves behave like
sequences, so it’s easy to move children from one element to
another:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>from xml.etree import ElementTree as ET

t = ET.XML(&quot;&quot;&quot;&lt;list&gt;
 &lt;item&gt;1&lt;/item&gt; &lt;item&gt;2&lt;/item&gt; &lt;item&gt;3&lt;/item&gt;
&lt;/list&gt;&quot;&quot;&quot;)
new = ET.XML(&#39;&lt;root/&gt;&#39;)
new.extend(t)

# Outputs &lt;root&gt;&lt;item&gt;1&lt;/item&gt;...&lt;/root&gt;
print ET.tostring(new)
</pre></div>
</div>
</li>
<li>New <code class="xref py py-class docutils literal notranslate">Element</code> method:
<a class="reference internal" href="../library/xml.etree.elementtree.html#xml.etree.ElementTree.Element.iter" title="xml.etree.ElementTree.Element.iter"><code class="xref py py-meth docutils literal notranslate">iter()</code></a> yields the children of the
element as a generator. It’s also possible to write <code class="docutils literal notranslate">for child in
elem:</code> to loop over an element’s children. The existing method
<code class="xref py py-meth docutils literal notranslate">getiterator()</code> is now deprecated, as is <code class="xref py py-meth docutils literal notranslate">getchildren()</code>
which constructs and returns a list of children.
</li>
<li>New <code class="xref py py-class docutils literal notranslate">Element</code> method:
<a class="reference internal" href="../library/xml.etree.elementtree.html#xml.etree.ElementTree.Element.itertext" title="xml.etree.ElementTree.Element.itertext"><code class="xref py py-meth docutils literal notranslate">itertext()</code></a> yields all chunks of
text that are descendants of the element. For example:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>t = ET.XML(&quot;&quot;&quot;&lt;list&gt;
 &lt;item&gt;1&lt;/item&gt; &lt;item&gt;2&lt;/item&gt; &lt;item&gt;3&lt;/item&gt;
&lt;/list&gt;&quot;&quot;&quot;)

# Outputs [&#39;\n &#39;, &#39;1&#39;, &#39; &#39;, &#39;2&#39;, &#39; &#39;, &#39;3&#39;, &#39;\n&#39;]
print list(t.itertext())
</pre></div>
</div>
</li>
<li>Deprecated: using an element as a Boolean (i.e., <code class="docutils literal notranslate">if elem:</code>) would
return true if the element had any children, or false if there were
no children. This behaviour is confusing – <code class="docutils literal notranslate">None</code> is false, but
so is a childless element? – so it will now trigger a
<a class="reference internal" href="../library/exceptions.html#FutureWarning" title="FutureWarning"><code class="xref py py-exc docutils literal notranslate">FutureWarning</code></a>. In your code, you should be explicit: write
<code class="docutils literal notranslate">len(elem) != 0</code> if you’re interested in the number of children,
or <code class="docutils literal notranslate">elem is not None</code>.
</li>
</ul>
Fredrik Lundh develops ElementTree and produced the 1.3 version;
you can read his article describing 1.3 at
<a class="reference external" href="http://effbot.org/zone/elementtree-13-intro.htm">http://effbot.org/zone/elementtree-13-intro.htm</a>.
Florent Xicluna updated the version included with
Python, after discussions on python-dev and in <a class="reference external" href="https://bugs.python.org/issue6472">bpo-6472</a>.)
</div>
</div>
<div class="section" id="build-and-c-api-changes">
<h2>Build and C API Changes<a class="headerlink" href="#build-and-c-api-changes" title="Permalink to this headline">¶</a></h2>
Changes to Python’s build process and to the C API include:
<ul>
<li>The latest release of the GNU Debugger, GDB 7, can be <a class="reference external" href="https://sourceware.org/gdb/current/onlinedocs/gdb/Python.html">scripted
using Python</a>.
When you begin debugging an executable program P, GDB will look for
a file named <code class="docutils literal notranslate">P-gdb.py</code> and automatically read it. Dave Malcolm
contributed a <code class="file docutils literal notranslate">python-gdb.py</code> that adds a number of
commands useful when debugging Python itself. For example,
<code class="docutils literal notranslate">py-up</code> and <code class="docutils literal notranslate">py-down</code> go up or down one Python stack frame,
which usually corresponds to several C stack frames. <code class="docutils literal notranslate">py-print</code>
prints the value of a Python variable, and <code class="docutils literal notranslate">py-bt</code> prints the
Python stack trace. (Added as a result of <a class="reference external" href="https://bugs.python.org/issue8032">bpo-8032</a>.)
</li>
<li>If you use the <code class="file docutils literal notranslate">.gdbinit</code> file provided with Python,
the “pyo” macro in the 2.7 version now works correctly when the thread being
debugged doesn’t hold the GIL; the macro now acquires it before printing.
(Contributed by Victor Stinner; <a class="reference external" href="https://bugs.python.org/issue3632">bpo-3632</a>.)
</li>
<li><a class="reference internal" href="../c-api/init.html#c.Py_AddPendingCall" title="Py_AddPendingCall"><code class="xref c c-func docutils literal notranslate">Py_AddPendingCall()</code></a> is now thread-safe, letting any
worker thread submit notifications to the main Python thread. This
is particularly useful for asynchronous IO operations.
(Contributed by Kristján Valur Jónsson; <a class="reference external" href="https://bugs.python.org/issue4293">bpo-4293</a>.)
</li>
<li>New function: <a class="reference internal" href="../c-api/code.html#c.PyCode_NewEmpty" title="PyCode_NewEmpty"><code class="xref c c-func docutils literal notranslate">PyCode_NewEmpty()</code></a> creates an empty code object;
only the filename, function name, and first line number are required.
This is useful for extension modules that are attempting to
construct a more useful traceback stack. Previously such
extensions needed to call <a class="reference internal" href="../c-api/code.html#c.PyCode_New" title="PyCode_New"><code class="xref c c-func docutils literal notranslate">PyCode_New()</code></a>, which had many
more arguments. (Added by Jeffrey Yasskin.)
</li>
<li>New function: <a class="reference internal" href="../c-api/exceptions.html#c.PyErr_NewExceptionWithDoc" title="PyErr_NewExceptionWithDoc"><code class="xref c c-func docutils literal notranslate">PyErr_NewExceptionWithDoc()</code></a> creates a new
exception class, just as the existing <a class="reference internal" href="../c-api/exceptions.html#c.PyErr_NewException" title="PyErr_NewException"><code class="xref c c-func docutils literal notranslate">PyErr_NewException()</code></a> does,
but takes an extra <code class="docutils literal notranslate">char *</code> argument containing the docstring for the
new exception class. (Added by ‘lekma’ on the Python bug tracker;
<a class="reference external" href="https://bugs.python.org/issue7033">bpo-7033</a>.)
</li>
<li>New function: <a class="reference internal" href="../c-api/reflection.html#c.PyFrame_GetLineNumber" title="PyFrame_GetLineNumber"><code class="xref c c-func docutils literal notranslate">PyFrame_GetLineNumber()</code></a> takes a frame object
and returns the line number that the frame is currently executing.
Previously code would need to get the index of the bytecode
instruction currently executing, and then look up the line number
corresponding to that address. (Added by Jeffrey Yasskin.)
</li>
<li>New functions: <a class="reference internal" href="../c-api/long.html#c.PyLong_AsLongAndOverflow" title="PyLong_AsLongAndOverflow"><code class="xref c c-func docutils literal notranslate">PyLong_AsLongAndOverflow()</code></a> and
<a class="reference internal" href="../c-api/long.html#c.PyLong_AsLongLongAndOverflow" title="PyLong_AsLongLongAndOverflow"><code class="xref c c-func docutils literal notranslate">PyLong_AsLongLongAndOverflow()</code></a> approximates a Python long
integer as a C <code class="xref c c-type docutils literal notranslate">long</code> or <code class="xref c c-type docutils literal notranslate">long long</code>.
If the number is too large to fit into
the output type, an overflow flag is set and returned to the caller.
(Contributed by Case Van Horsen; <a class="reference external" href="https://bugs.python.org/issue7528">bpo-7528</a> and <a class="reference external" href="https://bugs.python.org/issue7767">bpo-7767</a>.)
</li>
<li>New function: stemming from the rewrite of string-to-float conversion,
a new <a class="reference internal" href="../c-api/conversion.html#c.PyOS_string_to_double" title="PyOS_string_to_double"><code class="xref c c-func docutils literal notranslate">PyOS_string_to_double()</code></a> function was added. The old
<code class="xref c c-func docutils literal notranslate">PyOS_ascii_strtod()</code> and <code class="xref c c-func docutils literal notranslate">PyOS_ascii_atof()</code> functions
are now deprecated.
</li>
<li>New function: <a class="reference internal" href="../c-api/init.html#c.PySys_SetArgvEx" title="PySys_SetArgvEx"><code class="xref c c-func docutils literal notranslate">PySys_SetArgvEx()</code></a> sets the value of
<code class="docutils literal notranslate">sys.argv</code> and can optionally update <code class="docutils literal notranslate">sys.path</code> to include the
directory containing the script named by <code class="docutils literal notranslate">sys.argv[0]</code> depending
on the value of an updatepath parameter.
This function was added to close a security hole for applications
that embed Python. The old function, <a class="reference internal" href="../c-api/init.html#c.PySys_SetArgv" title="PySys_SetArgv"><code class="xref c c-func docutils literal notranslate">PySys_SetArgv()</code></a>, would
always update <code class="docutils literal notranslate">sys.path</code>, and sometimes it would add the current
directory. This meant that, if you ran an application embedding
Python in a directory controlled by someone else, attackers could
put a Trojan-horse module in the directory (say, a file named
<code class="file docutils literal notranslate">os.py</code>) that your application would then import and run.
If you maintain a C/C++ application that embeds Python, check
whether you’re calling <a class="reference internal" href="../c-api/init.html#c.PySys_SetArgv" title="PySys_SetArgv"><code class="xref c c-func docutils literal notranslate">PySys_SetArgv()</code></a> and carefully consider
whether the application should be using <a class="reference internal" href="../c-api/init.html#c.PySys_SetArgvEx" title="PySys_SetArgvEx"><code class="xref c c-func docutils literal notranslate">PySys_SetArgvEx()</code></a>
with updatepath set to false.
Security issue reported as <a class="reference external" href="https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2008-5983">CVE-2008-5983</a>;
discussed in <a class="reference external" href="https://bugs.python.org/issue5753">bpo-5753</a>, and fixed by Antoine Pitrou.
</li>
<li>New macros: the Python header files now define the following macros:
<code class="xref c c-macro docutils literal notranslate">Py_ISALNUM</code>,
<code class="xref c c-macro docutils literal notranslate">Py_ISALPHA</code>,
<code class="xref c c-macro docutils literal notranslate">Py_ISDIGIT</code>,
<code class="xref c c-macro docutils literal notranslate">Py_ISLOWER</code>,
<code class="xref c c-macro docutils literal notranslate">Py_ISSPACE</code>,
<code class="xref c c-macro docutils literal notranslate">Py_ISUPPER</code>,
<code class="xref c c-macro docutils literal notranslate">Py_ISXDIGIT</code>,
<code class="xref c c-macro docutils literal notranslate">Py_TOLOWER</code>, and <code class="xref c c-macro docutils literal notranslate">Py_TOUPPER</code>.
All of these functions are analogous to the C
standard macros for classifying characters, but ignore the current
locale setting, because in
several places Python needs to analyze characters in a
locale-independent way. (Added by Eric Smith;
<a class="reference external" href="https://bugs.python.org/issue5793">bpo-5793</a>.)
</li>
<li>Removed function: <code class="xref c c-macro docutils literal notranslate">PyEval_CallObject</code> is now only available
as a macro. A function version was being kept around to preserve
ABI linking compatibility, but that was in 1997; it can certainly be
deleted by now. (Removed by Antoine Pitrou; <a class="reference external" href="https://bugs.python.org/issue8276">bpo-8276</a>.)
</li>
<li>New format codes: the <code class="xref c c-func docutils literal notranslate">PyFormat_FromString()</code>,
<code class="xref c c-func docutils literal notranslate">PyFormat_FromStringV()</code>, and <a class="reference internal" href="../c-api/exceptions.html#c.PyErr_Format" title="PyErr_Format"><code class="xref c c-func docutils literal notranslate">PyErr_Format()</code></a> functions now
accept <code class="docutils literal notranslate">%lld</code> and <code class="docutils literal notranslate">%llu</code> format codes for displaying
C’s <code class="xref c c-type docutils literal notranslate">long long</code> types.
(Contributed by Mark Dickinson; <a class="reference external" href="https://bugs.python.org/issue7228">bpo-7228</a>.)
</li>
<li>The complicated interaction between threads and process forking has
been changed. Previously, the child process created by
<a class="reference internal" href="../library/os.html#os.fork" title="os.fork"><code class="xref py py-func docutils literal notranslate">os.fork()</code></a> might fail because the child is created with only a
single thread running, the thread performing the <a class="reference internal" href="../library/os.html#os.fork" title="os.fork"><code class="xref py py-func docutils literal notranslate">os.fork()</code></a>.
If other threads were holding a lock, such as Python’s import lock,
when the fork was performed, the lock would still be marked as
“held” in the new process. But in the child process nothing would
ever release the lock, since the other threads weren’t replicated,
and the child process would no longer be able to perform imports.
Python 2.7 acquires the import lock before performing an
<a class="reference internal" href="../library/os.html#os.fork" title="os.fork"><code class="xref py py-func docutils literal notranslate">os.fork()</code></a>, and will also clean up any locks created using the
<a class="reference internal" href="../library/threading.html#module-threading" title="threading: Thread-based parallelism."><code class="xref py py-mod docutils literal notranslate">threading</code></a> module. C extension modules that have internal
locks, or that call <code class="xref c c-func docutils literal notranslate">fork()</code> themselves, will not benefit
from this clean-up.
(Fixed by Thomas Wouters; <a class="reference external" href="https://bugs.python.org/issue1590864">bpo-1590864</a>.)
</li>
<li>The <a class="reference internal" href="../c-api/init.html#c.Py_Finalize" title="Py_Finalize"><code class="xref c c-func docutils literal notranslate">Py_Finalize()</code></a> function now calls the internal
<code class="xref py py-func docutils literal notranslate">threading._shutdown()</code> function; this prevents some exceptions from
being raised when an interpreter shuts down.
(Patch by Adam Olsen; <a class="reference external" href="https://bugs.python.org/issue1722344">bpo-1722344</a>.)
</li>
<li>When using the <a class="reference internal" href="../c-api/structures.html#c.PyMemberDef" title="PyMemberDef"><code class="xref c c-type docutils literal notranslate">PyMemberDef</code></a> structure to define attributes
of a type, Python will no longer let you try to delete or set a
<code class="xref py py-const docutils literal notranslate">T_STRING_INPLACE</code> attribute.
</li>
<li>Global symbols defined by the <a class="reference internal" href="../library/ctypes.html#module-ctypes" title="ctypes: A foreign function library for Python."><code class="xref py py-mod docutils literal notranslate">ctypes</code></a> module are now prefixed
with <code class="docutils literal notranslate">Py</code>, or with <code class="docutils literal notranslate">_ctypes</code>. (Implemented by Thomas
Heller; <a class="reference external" href="https://bugs.python.org/issue3102">bpo-3102</a>.)
</li>
<li>New configure option: the <code class="xref std std-option docutils literal notranslate">--with-system-expat</code> switch allows
building the <code class="xref py py-mod docutils literal notranslate">pyexpat</code> module to use the system Expat library.
(Contributed by Arfrever Frehtes Taifersar Arahesis; <a class="reference external" href="https://bugs.python.org/issue7609">bpo-7609</a>.)
</li>
<li>New configure option: the
<code class="xref std std-option docutils literal notranslate">--with-valgrind</code> option will now disable the pymalloc
allocator, which is difficult for the Valgrind memory-error detector
to analyze correctly.
Valgrind will therefore be better at detecting memory leaks and
overruns. (Contributed by James Henstridge; <a class="reference external" href="https://bugs.python.org/issue2422">bpo-2422</a>.)
</li>
<li>New configure option: you can now supply an empty string to
<code class="xref std std-option docutils literal notranslate">--with-dbmliborder=</code> in order to disable all of the various
DBM modules. (Added by Arfrever Frehtes Taifersar Arahesis;
<a class="reference external" href="https://bugs.python.org/issue6491">bpo-6491</a>.)
</li>
<li>The configure script now checks for floating-point rounding bugs
on certain 32-bit Intel chips and defines a <code class="xref c c-macro docutils literal notranslate">X87_DOUBLE_ROUNDING</code>
preprocessor definition. No code currently uses this definition,
but it’s available if anyone wishes to use it.
(Added by Mark Dickinson; <a class="reference external" href="https://bugs.python.org/issue2937">bpo-2937</a>.)
configure also now sets a <code class="xref std std-envvar docutils literal notranslate">LDCXXSHARED</code> Makefile
variable for supporting C++ linking. (Contributed by Arfrever
Frehtes Taifersar Arahesis; <a class="reference external" href="https://bugs.python.org/issue1222585">bpo-1222585</a>.)
</li>
<li>The build process now creates the necessary files for pkg-config
support. (Contributed by Clinton Roy; <a class="reference external" href="https://bugs.python.org/issue3585">bpo-3585</a>.)
</li>
<li>The build process now supports Subversion 1.7. (Contributed by
Arfrever Frehtes Taifersar Arahesis; <a class="reference external" href="https://bugs.python.org/issue6094">bpo-6094</a>.)
</li>
</ul>
<div class="section" id="capsules">
<h3>Capsules<a class="headerlink" href="#capsules" title="Permalink to this headline">¶</a></h3>
Python 3.1 adds a new C datatype, <a class="reference internal" href="../c-api/capsule.html#c.PyCapsule" title="PyCapsule"><code class="xref c c-type docutils literal notranslate">PyCapsule</code></a>, for providing a
C API to an extension module. A capsule is essentially the holder of
a C <code class="docutils literal notranslate">void *</code> pointer, and is made available as a module attribute; for
example, the <a class="reference internal" href="../library/socket.html#module-socket" title="socket: Low-level networking interface."><code class="xref py py-mod docutils literal notranslate">socket</code></a> module’s API is exposed as <code class="docutils literal notranslate">socket.CAPI</code>,
and <a class="reference internal" href="../library/unicodedata.html#module-unicodedata" title="unicodedata: Access the Unicode Database."><code class="xref py py-mod docutils literal notranslate">unicodedata</code></a> exposes <code class="docutils literal notranslate">ucnhash_CAPI</code>. Other extensions
can import the module, access its dictionary to get the capsule
object, and then get the <code class="docutils literal notranslate">void *</code> pointer, which will usually point
to an array of pointers to the module’s various API functions.
There is an existing data type already used for this,
<code class="xref c c-type docutils literal notranslate">PyCObject</code>, but it doesn’t provide type safety. Evil code
written in pure Python could cause a segmentation fault by taking a
<code class="xref c c-type docutils literal notranslate">PyCObject</code> from module A and somehow substituting it for the
<code class="xref c c-type docutils literal notranslate">PyCObject</code> in module B. Capsules know their own name,
and getting the pointer requires providing the name:
<div class="highlight-c notranslate"><div class="highlight"><pre>void *vtable;

if (!PyCapsule_IsValid(capsule, &quot;mymodule.CAPI&quot;) {
 PyErr_SetString(PyExc_ValueError, &quot;argument type invalid&quot;);
 return NULL;
}

vtable = PyCapsule_GetPointer(capsule, &quot;mymodule.CAPI&quot;);
</pre></div>
</div>
You are assured that <code class="docutils literal notranslate">vtable</code> points to whatever you’re expecting.
If a different capsule was passed in, <a class="reference internal" href="../c-api/capsule.html#c.PyCapsule_IsValid" title="PyCapsule_IsValid"><code class="xref c c-func docutils literal notranslate">PyCapsule_IsValid()</code></a> would
detect the mismatched name and return false. Refer to
<a class="reference internal" href="../extending/extending.html#using-capsules">Providing a C API for an Extension Module</a> for more information on using these objects.
Python 2.7 now uses capsules internally to provide various
extension-module APIs, but the <code class="xref c c-func docutils literal notranslate">PyCObject_AsVoidPtr()</code> was
modified to handle capsules, preserving compile-time compatibility
with the <code class="xref c c-type docutils literal notranslate">CObject</code> interface. Use of
<code class="xref c c-func docutils literal notranslate">PyCObject_AsVoidPtr()</code> will signal a
<a class="reference internal" href="../library/exceptions.html#PendingDeprecationWarning" title="PendingDeprecationWarning"><code class="xref py py-exc docutils literal notranslate">PendingDeprecationWarning</code></a>, which is silent by default.
Implemented in Python 3.1 and backported to 2.7 by Larry Hastings;
discussed in <a class="reference external" href="https://bugs.python.org/issue5630">bpo-5630</a>.
</div>
<div class="section" id="port-specific-changes-windows">
<h3>Port-Specific Changes: Windows<a class="headerlink" href="#port-specific-changes-windows" title="Permalink to this headline">¶</a></h3>
<ul class="simple">
<li>The <a class="reference internal" href="../library/msvcrt.html#module-msvcrt" title="msvcrt: Miscellaneous useful routines from the MS VC++ runtime. (Windows)"><code class="xref py py-mod docutils literal notranslate">msvcrt</code></a> module now contains some constants from
the <code class="file docutils literal notranslate">crtassem.h</code> header file:
<code class="xref py py-data docutils literal notranslate">CRT_ASSEMBLY_VERSION</code>,
<code class="xref py py-data docutils literal notranslate">VC_ASSEMBLY_PUBLICKEYTOKEN</code>,
and <code class="xref py py-data docutils literal notranslate">LIBRARIES_ASSEMBLY_NAME_PREFIX</code>.
(Contributed by David Cournapeau; <a class="reference external" href="https://bugs.python.org/issue4365">bpo-4365</a>.)</li>
<li>The <code class="xref py py-mod docutils literal notranslate">_winreg</code> module for accessing the registry now implements
the <code class="xref py py-func docutils literal notranslate">CreateKeyEx()</code> and <code class="xref py py-func docutils literal notranslate">DeleteKeyEx()</code>
functions, extended versions of previously-supported functions that
take several extra arguments. The <code class="xref py py-func docutils literal notranslate">DisableReflectionKey()</code>,
<code class="xref py py-func docutils literal notranslate">EnableReflectionKey()</code>, and <code class="xref py py-func docutils literal notranslate">QueryReflectionKey()</code>
were also tested and documented.
(Implemented by Brian Curtin: <a class="reference external" href="https://bugs.python.org/issue7347">bpo-7347</a>.)</li>
<li>The new <code class="xref c c-func docutils literal notranslate">_beginthreadex()</code> API is used to start threads, and
the native thread-local storage functions are now used.
(Contributed by Kristján Valur Jónsson; <a class="reference external" href="https://bugs.python.org/issue3582">bpo-3582</a>.)</li>
<li>The <a class="reference internal" href="../library/os.html#os.kill" title="os.kill"><code class="xref py py-func docutils literal notranslate">os.kill()</code></a> function now works on Windows. The signal value
can be the constants <code class="xref py py-const docutils literal notranslate">CTRL_C_EVENT</code>,
<code class="xref py py-const docutils literal notranslate">CTRL_BREAK_EVENT</code>, or any integer. The first two constants
will send <kbd class="kbd docutils literal notranslate">Control-C</kbd> and <kbd class="kbd docutils literal notranslate">Control-Break</kbd> keystroke events to
subprocesses; any other value will use the <code class="xref c c-func docutils literal notranslate">TerminateProcess()</code>
API. (Contributed by Miki Tebeka; <a class="reference external" href="https://bugs.python.org/issue1220212">bpo-1220212</a>.)</li>
<li>The <a class="reference internal" href="../library/os.html#os.listdir" title="os.listdir"><code class="xref py py-func docutils literal notranslate">os.listdir()</code></a> function now correctly fails
for an empty path. (Fixed by Hirokazu Yamamoto; <a class="reference external" href="https://bugs.python.org/issue5913">bpo-5913</a>.)</li>
<li>The <code class="xref py py-mod docutils literal notranslate">mimelib</code> module will now read the MIME database from
the Windows registry when initializing.
(Patch by Gabriel Genellina; <a class="reference external" href="https://bugs.python.org/issue4969">bpo-4969</a>.)</li>
</ul>
</div>
<div class="section" id="port-specific-changes-mac-os-x">
<h3>Port-Specific Changes: Mac OS X<a class="headerlink" href="#port-specific-changes-mac-os-x" title="Permalink to this headline">¶</a></h3>
<ul>
<li>The path <code class="docutils literal notranslate">/Library/Python/2.7/site-packages</code> is now appended to
<code class="docutils literal notranslate">sys.path</code>, in order to share added packages between the system
installation and a user-installed copy of the same version.
(Changed by Ronald Oussoren; <a class="reference external" href="https://bugs.python.org/issue4865">bpo-4865</a>.)
<blockquote>
<div><div class="versionchanged">
Changed in version 2.7.13: As of 2.7.13, this change was removed.
<code class="docutils literal notranslate">/Library/Python/2.7/site-packages</code>, the site-packages directory
used by the Apple-supplied system Python 2.7 is no longer appended to
<code class="docutils literal notranslate">sys.path</code> for user-installed Pythons such as from the python.org
installers. As of macOS 10.12, Apple changed how the system
site-packages directory is configured, which could cause installation
of pip components, like setuptools, to fail. Packages installed for
the system Python will no longer be shared with user-installed
Pythons. (<a class="reference external" href="https://bugs.python.org/issue28440">bpo-28440</a>)
</div>
</div></blockquote>
</li>
</ul>
</div>
<div class="section" id="port-specific-changes-freebsd">
<h3>Port-Specific Changes: FreeBSD<a class="headerlink" href="#port-specific-changes-freebsd" title="Permalink to this headline">¶</a></h3>
<ul class="simple">
<li>FreeBSD 7.1’s <code class="xref py py-const docutils literal notranslate">SO_SETFIB</code> constant, used with
<code class="xref py py-func docutils literal notranslate">getsockopt()</code>/<code class="xref py py-func docutils literal notranslate">setsockopt()</code> to select an
alternate routing table, is now available in the <a class="reference internal" href="../library/socket.html#module-socket" title="socket: Low-level networking interface."><code class="xref py py-mod docutils literal notranslate">socket</code></a>
module. (Added by Kyle VanderBeek; <a class="reference external" href="https://bugs.python.org/issue8235">bpo-8235</a>.)</li>
</ul>
</div>
</div>
<div class="section" id="other-changes-and-fixes">
<h2>Other Changes and Fixes<a class="headerlink" href="#other-changes-and-fixes" title="Permalink to this headline">¶</a></h2>
<ul class="simple">
<li>Two benchmark scripts, <code class="file docutils literal notranslate">iobench</code> and <code class="file docutils literal notranslate">ccbench</code>, were
added to the <code class="file docutils literal notranslate">Tools</code> directory. <code class="file docutils literal notranslate">iobench</code> measures the
speed of the built-in file I/O objects returned by <a class="reference internal" href="../library/functions.html#open" title="open"><code class="xref py py-func docutils literal notranslate">open()</code></a>
while performing various operations, and <code class="file docutils literal notranslate">ccbench</code> is a
concurrency benchmark that tries to measure computing throughput,
thread switching latency, and IO processing bandwidth when
performing several tasks using a varying number of threads.</li>
<li>The <code class="file docutils literal notranslate">Tools/i18n/msgfmt.py</code> script now understands plural
forms in <code class="file docutils literal notranslate">.po</code> files. (Fixed by Martin von Löwis;
<a class="reference external" href="https://bugs.python.org/issue5464">bpo-5464</a>.)</li>
<li>When importing a module from a <code class="file docutils literal notranslate">.pyc</code> or <code class="file docutils literal notranslate">.pyo</code> file
with an existing <code class="file docutils literal notranslate">.py</code> counterpart, the <code class="xref py py-attr docutils literal notranslate">co_filename</code>
attributes of the resulting code objects are overwritten when the
original filename is obsolete. This can happen if the file has been
renamed, moved, or is accessed through different paths. (Patch by
Ziga Seilnacht and Jean-Paul Calderone; <a class="reference external" href="https://bugs.python.org/issue1180193">bpo-1180193</a>.)</li>
<li>The <code class="file docutils literal notranslate">regrtest.py</code> script now takes a <code class="xref std std-option docutils literal notranslate">--randseed=</code>
switch that takes an integer that will be used as the random seed
for the <code class="xref std std-option docutils literal notranslate">-r</code> option that executes tests in random order.
The <code class="xref std std-option docutils literal notranslate">-r</code> option also reports the seed that was used
(Added by Collin Winter.)</li>
<li>Another <code class="file docutils literal notranslate">regrtest.py</code> switch is <code class="xref std std-option docutils literal notranslate">-j</code>, which
takes an integer specifying how many tests run in parallel. This
allows reducing the total runtime on multi-core machines.
This option is compatible with several other options, including the
<code class="xref std std-option docutils literal notranslate">-R</code> switch which is known to produce long runtimes.
(Added by Antoine Pitrou, <a class="reference external" href="https://bugs.python.org/issue6152">bpo-6152</a>.) This can also be used
with a new <code class="xref std std-option docutils literal notranslate">-F</code> switch that runs selected tests in a loop
until they fail. (Added by Antoine Pitrou; <a class="reference external" href="https://bugs.python.org/issue7312">bpo-7312</a>.)</li>
<li>When executed as a script, the <code class="file docutils literal notranslate">py_compile.py</code> module now
accepts <code class="docutils literal notranslate">'-'</code> as an argument, which will read standard input for
the list of filenames to be compiled. (Contributed by Piotr
Ożarowski; <a class="reference external" href="https://bugs.python.org/issue8233">bpo-8233</a>.)</li>
</ul>
</div>
<div class="section" id="porting-to-python-2-7">
<h2>Porting to Python 2.7<a class="headerlink" href="#porting-to-python-2-7" title="Permalink to this headline">¶</a></h2>
This section lists previously described changes and other bugfixes
that may require changes to your code:
<ul class="simple">
<li>The <a class="reference internal" href="../library/stdtypes.html#range" title="range"><code class="xref py py-func docutils literal notranslate">range()</code></a> function processes its arguments more
consistently; it will now call <a class="reference internal" href="../reference/datamodel.html#object.__int__" title="object.__int__"><code class="xref py py-meth docutils literal notranslate">__int__()</code></a> on non-float,
non-integer arguments that are supplied to it. (Fixed by Alexander
Belopolsky; <a class="reference external" href="https://bugs.python.org/issue1533">bpo-1533</a>.)</li>
<li>The string <a class="reference internal" href="../library/functions.html#format" title="format"><code class="xref py py-meth docutils literal notranslate">format()</code></a> method changed the default precision used
for floating-point and complex numbers from 6 decimal
places to 12, which matches the precision used by <a class="reference internal" href="../library/stdtypes.html#str" title="str"><code class="xref py py-func docutils literal notranslate">str()</code></a>.
(Changed by Eric Smith; <a class="reference external" href="https://bugs.python.org/issue5920">bpo-5920</a>.)</li>
<li>Because of an optimization for the <a class="reference internal" href="../reference/compound_stmts.html#with"><code class="xref std std-keyword docutils literal notranslate">with</code></a> statement, the special
methods <a class="reference internal" href="../reference/datamodel.html#object.__enter__" title="object.__enter__"><code class="xref py py-meth docutils literal notranslate">__enter__()</code></a> and <a class="reference internal" href="../reference/datamodel.html#object.__exit__" title="object.__exit__"><code class="xref py py-meth docutils literal notranslate">__exit__()</code></a> must belong to the object’s
type, and cannot be directly attached to the object’s instance. This
affects new-style classes (derived from <a class="reference internal" href="../library/functions.html#object" title="object"><code class="xref py py-class docutils literal notranslate">object</code></a>) and C extension
types. (<a class="reference external" href="https://bugs.python.org/issue6101">bpo-6101</a>.)</li>
<li>Due to a bug in Python 2.6, the exc_value parameter to
<a class="reference internal" href="../reference/datamodel.html#object.__exit__" title="object.__exit__"><code class="xref py py-meth docutils literal notranslate">__exit__()</code></a> methods was often the string representation of the
exception, not an instance. This was fixed in 2.7, so exc_value
will be an instance as expected. (Fixed by Florent Xicluna;
<a class="reference external" href="https://bugs.python.org/issue7853">bpo-7853</a>.)</li>
<li>When a restricted set of attributes were set using <code class="docutils literal notranslate">__slots__</code>,
deleting an unset attribute would not raise <a class="reference internal" href="../library/exceptions.html#AttributeError" title="AttributeError"><code class="xref py py-exc docutils literal notranslate">AttributeError</code></a>
as you would expect. Fixed by Benjamin Peterson; <a class="reference external" href="https://bugs.python.org/issue7604">bpo-7604</a>.)</li>
</ul>
In the standard library:
<ul>
<li>Operations with <a class="reference internal" href="../library/datetime.html#datetime.datetime" title="datetime.datetime"><code class="xref py py-class docutils literal notranslate">datetime</code></a> instances that resulted in a year
falling outside the supported range didn’t always raise
<a class="reference internal" href="../library/exceptions.html#OverflowError" title="OverflowError"><code class="xref py py-exc docutils literal notranslate">OverflowError</code></a>. Such errors are now checked more carefully
and will now raise the exception. (Reported by Mark Leander, patch
by Anand B. Pillai and Alexander Belopolsky; <a class="reference external" href="https://bugs.python.org/issue7150">bpo-7150</a>.)
</li>
<li>When using <a class="reference internal" href="../library/decimal.html#decimal.Decimal" title="decimal.Decimal"><code class="xref py py-class docutils literal notranslate">Decimal</code></a> instances with a string’s
<a class="reference internal" href="../library/functions.html#format" title="format"><code class="xref py py-meth docutils literal notranslate">format()</code></a> method, the default alignment was previously
left-alignment. This has been changed to right-alignment, which might
change the output of your programs.
(Changed by Mark Dickinson; <a class="reference external" href="https://bugs.python.org/issue6857">bpo-6857</a>.)
Comparisons involving a signaling NaN value (or <code class="docutils literal notranslate">sNAN</code>) now signal
<a class="reference internal" href="../library/decimal.html#decimal.InvalidOperation" title="decimal.InvalidOperation"><code class="xref py py-const docutils literal notranslate">InvalidOperation</code></a> instead of silently returning a true or
false value depending on the comparison operator. Quiet NaN values
(or <code class="docutils literal notranslate">NaN</code>) are now hashable. (Fixed by Mark Dickinson;
<a class="reference external" href="https://bugs.python.org/issue7279">bpo-7279</a>.)
</li>
<li>The ElementTree library, <code class="xref py py-mod docutils literal notranslate">xml.etree</code>, no longer escapes
ampersands and angle brackets when outputting an XML processing
instruction (which looks like <cite>&lt;?xml-stylesheet href=”#style1”?&gt;</cite>)
or comment (which looks like <cite>&lt;!– comment –&gt;</cite>).
(Patch by Neil Muller; <a class="reference external" href="https://bugs.python.org/issue2746">bpo-2746</a>.)
</li>
<li>The <code class="xref py py-meth docutils literal notranslate">readline()</code> method of <code class="xref py py-class docutils literal notranslate">StringIO</code> objects now does
nothing when a negative length is requested, as other file-like
objects do. (<a class="reference external" href="https://bugs.python.org/issue7348">bpo-7348</a>).
</li>
<li>The <a class="reference internal" href="../library/syslog.html#module-syslog" title="syslog: An interface to the Unix syslog library routines. (Unix)"><code class="xref py py-mod docutils literal notranslate">syslog</code></a> module will now use the value of <code class="docutils literal notranslate">sys.argv[0]</code> as the
identifier instead of the previous default value of <code class="docutils literal notranslate">'python'</code>.
(Changed by Sean Reifschneider; <a class="reference external" href="https://bugs.python.org/issue8451">bpo-8451</a>.)
</li>
<li>The <a class="reference internal" href="../library/tarfile.html#module-tarfile" title="tarfile: Read and write tar-format archive files."><code class="xref py py-mod docutils literal notranslate">tarfile</code></a> module’s default error handling has changed, to
no longer suppress fatal errors. The default error level was previously 0,
which meant that errors would only result in a message being written to the
debug log, but because the debug log is not activated by default,
these errors go unnoticed. The default error level is now 1,
which raises an exception if there’s an error.
(Changed by Lars Gustäbel; <a class="reference external" href="https://bugs.python.org/issue7357">bpo-7357</a>.)
</li>
<li>The <code class="xref py py-mod docutils literal notranslate">urlparse</code> module’s <code class="xref py py-func docutils literal notranslate">urlsplit()</code> now handles
unknown URL schemes in a fashion compliant with <a class="rfc reference external" href="https://tools.ietf.org/html/rfc3986.html">RFC 3986</a>: if the
URL is of the form <code class="docutils literal notranslate">&quot;&lt;something&gt;://...&quot;</code>, the text before the
<code class="docutils literal notranslate">://</code> is treated as the scheme, even if it’s a made-up scheme that
the module doesn’t know about. This change may break code that
worked around the old behaviour. For example, Python 2.6.4 or 2.5
will return the following:
<div class="highlight-pycon notranslate"><div class="highlight"><pre>&gt;&gt;&gt; import urlparse
&gt;&gt;&gt; urlparse.urlsplit(&#39;invented://host/filename?query&#39;)
(&#39;invented&#39;, &#39;&#39;, &#39;//host/filename?query&#39;, &#39;&#39;, &#39;&#39;)
</pre></div>
</div>
Python 2.7 (and Python 2.6.5) will return:
<div class="highlight-pycon notranslate"><div class="highlight"><pre>&gt;&gt;&gt; import urlparse
&gt;&gt;&gt; urlparse.urlsplit(&#39;invented://host/filename?query&#39;)
(&#39;invented&#39;, &#39;host&#39;, &#39;/filename?query&#39;, &#39;&#39;, &#39;&#39;)
</pre></div>
</div>
(Python 2.7 actually produces slightly different output, since it
returns a named tuple instead of a standard tuple.)
</li>
</ul>
For C extensions:
<ul class="simple">
<li>C extensions that use integer format codes with the <code class="docutils literal notranslate">PyArg_Parse*</code>
family of functions will now raise a <a class="reference internal" href="../library/exceptions.html#TypeError" title="TypeError"><code class="xref py py-exc docutils literal notranslate">TypeError</code></a> exception
instead of triggering a <a class="reference internal" href="../library/exceptions.html#DeprecationWarning" title="DeprecationWarning"><code class="xref py py-exc docutils literal notranslate">DeprecationWarning</code></a> (<a class="reference external" href="https://bugs.python.org/issue5080">bpo-5080</a>).</li>
<li>Use the new <a class="reference internal" href="../c-api/conversion.html#c.PyOS_string_to_double" title="PyOS_string_to_double"><code class="xref c c-func docutils literal notranslate">PyOS_string_to_double()</code></a> function instead of the old
<code class="xref c c-func docutils literal notranslate">PyOS_ascii_strtod()</code> and <code class="xref c c-func docutils literal notranslate">PyOS_ascii_atof()</code> functions,
which are now deprecated.</li>
</ul>
For applications that embed Python:
<ul class="simple">
<li>The <a class="reference internal" href="../c-api/init.html#c.PySys_SetArgvEx" title="PySys_SetArgvEx"><code class="xref c c-func docutils literal notranslate">PySys_SetArgvEx()</code></a> function was added, letting
applications close a security hole when the existing
<a class="reference internal" href="../c-api/init.html#c.PySys_SetArgv" title="PySys_SetArgv"><code class="xref c c-func docutils literal notranslate">PySys_SetArgv()</code></a> function was used. Check whether you’re
calling <a class="reference internal" href="../c-api/init.html#c.PySys_SetArgv" title="PySys_SetArgv"><code class="xref c c-func docutils literal notranslate">PySys_SetArgv()</code></a> and carefully consider whether the
application should be using <a class="reference internal" href="../c-api/init.html#c.PySys_SetArgvEx" title="PySys_SetArgvEx"><code class="xref c c-func docutils literal notranslate">PySys_SetArgvEx()</code></a> with
updatepath set to false.</li>
</ul>
</div>
<div class="section" id="new-features-added-to-python-2-7-maintenance-releases">
<h2>New Features Added to Python 2.7 Maintenance Releases<a class="headerlink" href="#new-features-added-to-python-2-7-maintenance-releases" title="Permalink to this headline">¶</a></h2>
New features may be added to Python 2.7 maintenance releases when the
situation genuinely calls for it. Any such additions must go through
the Python Enhancement Proposal process, and make a compelling case for why
they can’t be adequately addressed by either adding the new feature solely to
Python 3, or else by publishing it on the Python Package Index.
In addition to the specific proposals listed below, there is a general
exemption allowing new <code class="docutils literal notranslate">-3</code> warnings to be added in any Python 2.7
maintenance release.
<div class="section" id="two-new-environment-variables-for-debug-mode">
<h3>Two new environment variables for debug mode<a class="headerlink" href="#two-new-environment-variables-for-debug-mode" title="Permalink to this headline">¶</a></h3>
In debug mode, the <code class="docutils literal notranslate">[xxx refs]</code> statistic is not written by default, the
<code class="xref std std-envvar docutils literal notranslate">PYTHONSHOWREFCOUNT</code> environment variable now must also be set.
(Contributed by Victor Stinner; <a class="reference external" href="https://bugs.python.org/issue31733">bpo-31733</a>.)
When Python is compiled with <code class="docutils literal notranslate">COUNT_ALLOC</code> defined, allocation counts are no
longer dumped by default anymore: the <code class="xref std std-envvar docutils literal notranslate">PYTHONSHOWALLOCCOUNT</code> environment
variable must now also be set. Moreover, allocation counts are now dumped into
stderr, rather than stdout. (Contributed by Victor Stinner; <a class="reference external" href="https://bugs.python.org/issue31692">bpo-31692</a>.)
<div class="versionadded">
New in version 2.7.15.
</div>
</div>
<div class="section" id="pep-434-idle-enhancement-exception-for-all-branches">
<h3>PEP 434: IDLE Enhancement Exception for All Branches<a class="headerlink" href="#pep-434-idle-enhancement-exception-for-all-branches" title="Permalink to this headline">¶</a></h3>
<a class="pep reference external" href="https://www.python.org/dev/peps/pep-0434">PEP 434</a> describes a general exemption for changes made to the IDLE
development environment shipped along with Python. This exemption makes it
possible for the IDLE developers to provide a more consistent user
experience across all supported versions of Python 2 and 3.
For details of any IDLE changes, refer to the NEWS file for the specific
release.
</div>
<div class="section" id="pep-466-network-security-enhancements-for-python-2-7">
<h3>PEP 466: Network Security Enhancements for Python 2.7<a class="headerlink" href="#pep-466-network-security-enhancements-for-python-2-7" title="Permalink to this headline">¶</a></h3>
<a class="pep reference external" href="https://www.python.org/dev/peps/pep-0466">PEP 466</a> describes a number of network security enhancement proposals
that have been approved for inclusion in Python 2.7 maintenance releases,
with the first of those changes appearing in the Python 2.7.7 release.
<a class="pep reference external" href="https://www.python.org/dev/peps/pep-0466">PEP 466</a> related features added in Python 2.7.7:
<ul class="simple">
<li><a class="reference internal" href="../library/hmac.html#hmac.compare_digest" title="hmac.compare_digest"><code class="xref py py-func docutils literal notranslate">hmac.compare_digest()</code></a> was backported from Python 3 to make a timing
attack resistant comparison operation available to Python 2 applications.
(Contributed by Alex Gaynor; <a class="reference external" href="https://bugs.python.org/issue21306">bpo-21306</a>.)</li>
<li>OpenSSL 1.0.1g was upgraded in the official Windows installers published on
python.org. (Contributed by Zachary Ware; <a class="reference external" href="https://bugs.python.org/issue21462">bpo-21462</a>.)</li>
</ul>
<a class="pep reference external" href="https://www.python.org/dev/peps/pep-0466">PEP 466</a> related features added in Python 2.7.8:
<ul class="simple">
<li><a class="reference internal" href="../library/hashlib.html#hashlib.pbkdf2_hmac" title="hashlib.pbkdf2_hmac"><code class="xref py py-func docutils literal notranslate">hashlib.pbkdf2_hmac()</code></a> was backported from Python 3 to make a hashing
algorithm suitable for secure password storage broadly available to Python
2 applications. (Contributed by Alex Gaynor; <a class="reference external" href="https://bugs.python.org/issue21304">bpo-21304</a>.)</li>
<li>OpenSSL 1.0.1h was upgraded for the official Windows installers published on
python.org. (contributed by Zachary Ware in <a class="reference external" href="https://bugs.python.org/issue21671">bpo-21671</a> for CVE-2014-0224)</li>
</ul>
<a class="pep reference external" href="https://www.python.org/dev/peps/pep-0466">PEP 466</a> related features added in Python 2.7.9:
<ul>
<li>Most of Python 3.4’s <a class="reference internal" href="../library/ssl.html#module-ssl" title="ssl: TLS/SSL wrapper for socket objects"><code class="xref py py-mod docutils literal notranslate">ssl</code></a> module was backported. This means <a class="reference internal" href="../library/ssl.html#module-ssl" title="ssl: TLS/SSL wrapper for socket objects"><code class="xref py py-mod docutils literal notranslate">ssl</code></a>
now supports Server Name Indication, TLS1.x settings, access to the platform
certificate store, the <a class="reference internal" href="../library/ssl.html#ssl.SSLContext" title="ssl.SSLContext"><code class="xref py py-class docutils literal notranslate">SSLContext</code></a> class, and other
features. (Contributed by Alex Gaynor and David Reid; <a class="reference external" href="https://bugs.python.org/issue21308">bpo-21308</a>.)
Refer to the “Version added: 2.7.9” notes in the module documentation for
specific details.
</li>
<li><a class="reference internal" href="../library/os.html#os.urandom" title="os.urandom"><code class="xref py py-func docutils literal notranslate">os.urandom()</code></a> was changed to cache a file descriptor to <code class="docutils literal notranslate">/dev/urandom</code>
instead of reopening <code class="docutils literal notranslate">/dev/urandom</code> on every call. (Contributed by Alex
Gaynor; <a class="reference external" href="https://bugs.python.org/issue21305">bpo-21305</a>.)
</li>
<li><a class="reference internal" href="../library/hashlib.html#hashlib.algorithms_guaranteed" title="hashlib.algorithms_guaranteed"><code class="xref py py-data docutils literal notranslate">hashlib.algorithms_guaranteed</code></a> and
<a class="reference internal" href="../library/hashlib.html#hashlib.algorithms_available" title="hashlib.algorithms_available"><code class="xref py py-data docutils literal notranslate">hashlib.algorithms_available</code></a> were backported from Python 3 to make
it easier for Python 2 applications to select the strongest available hash
algorithm. (Contributed by Alex Gaynor in <a class="reference external" href="https://bugs.python.org/issue21307">bpo-21307</a>)
</li>
</ul>
</div>
<div class="section" id="pep-477-backport-ensurepip-pep-453-to-python-2-7">
<h3>PEP 477: Backport ensurepip (PEP 453) to Python 2.7<a class="headerlink" href="#pep-477-backport-ensurepip-pep-453-to-python-2-7" title="Permalink to this headline">¶</a></h3>
<a class="pep reference external" href="https://www.python.org/dev/peps/pep-0477">PEP 477</a> approves the inclusion of the <a class="pep reference external" href="https://www.python.org/dev/peps/pep-0453">PEP 453</a> ensurepip module and the
improved documentation that was enabled by it in the Python 2.7 maintenance
releases, appearing first in the Python 2.7.9 release.
<div class="section" id="bootstrapping-pip-by-default">
<h4>Bootstrapping pip By Default<a class="headerlink" href="#bootstrapping-pip-by-default" title="Permalink to this headline">¶</a></h4>
The new <a class="reference internal" href="../library/ensurepip.html#module-ensurepip" title="ensurepip: Bootstrapping the &quot;pip&quot; installer into an existing Python installation or virtual environment."><code class="xref py py-mod docutils literal notranslate">ensurepip</code></a> module (defined in <a class="pep reference external" href="https://www.python.org/dev/peps/pep-0453">PEP 453</a>) provides a standard
cross-platform mechanism to bootstrap the pip installer into Python
installations. The version of <code class="docutils literal notranslate">pip</code> included with Python 2.7.9 is <code class="docutils literal notranslate">pip</code>
1.5.6, and future 2.7.x maintenance releases will update the bundled version to
the latest version of <code class="docutils literal notranslate">pip</code> that is available at the time of creating the
release candidate.
By default, the commands <code class="docutils literal notranslate">pip</code>, <code class="docutils literal notranslate">pipX</code> and <code class="docutils literal notranslate">pipX.Y</code> will be installed on
all platforms (where X.Y stands for the version of the Python installation),
along with the <code class="docutils literal notranslate">pip</code> Python package and its dependencies.
For CPython <a class="reference internal" href="../using/unix.html#building-python-on-unix">source builds on POSIX systems</a>,
the <code class="docutils literal notranslate">make install</code> and <code class="docutils literal notranslate">make altinstall</code> commands do not bootstrap <code class="docutils literal notranslate">pip</code>
by default. This behaviour can be controlled through configure options, and
overridden through Makefile options.
On Windows and Mac OS X, the CPython installers now default to installing
<code class="docutils literal notranslate">pip</code> along with CPython itself (users may opt out of installing it
during the installation process). Window users will need to opt in to the
automatic <code class="docutils literal notranslate">PATH</code> modifications to have <code class="docutils literal notranslate">pip</code> available from the command
line by default, otherwise it can still be accessed through the Python
launcher for Windows as <code class="docutils literal notranslate">py -m pip</code>.
As <a class="reference external" href="https://www.python.org/dev/peps/pep-0477/#disabling-ensurepip-by-downstream-distributors">discussed in the PEP</a>, platform packagers may choose not to install
these commands by default, as long as, when invoked, they provide clear and
simple directions on how to install them on that platform (usually using
the system package manager).
</div>
<div class="section" id="documentation-changes">
<h4>Documentation Changes<a class="headerlink" href="#documentation-changes" title="Permalink to this headline">¶</a></h4>
As part of this change, the <a class="reference internal" href="../installing/index.html#installing-index">Installing Python Modules</a> and
<a class="reference internal" href="../distributing/index.html#distributing-index">Distributing Python Modules</a> sections of the documentation have been
completely redesigned as short getting started and FAQ documents. Most
packaging documentation has now been moved out to the Python Packaging
Authority maintained <a class="reference external" href="http://packaging.python.org">Python Packaging User Guide</a> and the documentation of the individual
projects.
However, as this migration is currently still incomplete, the legacy
versions of those guides remaining available as <a class="reference internal" href="../install/index.html#install-index">Installing Python Modules (Legacy version)</a>
and <a class="reference internal" href="../distutils/index.html#distutils-index">Distributing Python Modules (Legacy version)</a>.
<div class="admonition seealso">
See also
<dl class="last docutils">
<dt><a class="pep reference external" href="https://www.python.org/dev/peps/pep-0453">PEP 453</a> – Explicit bootstrapping of pip in Python installations</dt>
<dd>PEP written by Donald Stufft and Nick Coghlan, implemented by
Donald Stufft, Nick Coghlan, Martin von Löwis and Ned Deily.</dd>
</dl>
</div>
</div>
</div>
<div class="section" id="pep-476-enabling-certificate-verification-by-default-for-stdlib-http-clients">
<h3>PEP 476: Enabling certificate verification by default for stdlib http clients<a class="headerlink" href="#pep-476-enabling-certificate-verification-by-default-for-stdlib-http-clients" title="Permalink to this headline">¶</a></h3>
<a class="pep reference external" href="https://www.python.org/dev/peps/pep-0476">PEP 476</a> updated <code class="xref py py-mod docutils literal notranslate">httplib</code> and modules which use it, such as
<code class="xref py py-mod docutils literal notranslate">urllib2</code> and <code class="xref py py-mod docutils literal notranslate">xmlrpclib</code>, to now verify that the server
presents a certificate which is signed by a Certificate Authority in the
platform trust store and whose hostname matches the hostname being requested
by default, significantly improving security for many applications. This
change was made in the Python 2.7.9 release.
For applications which require the old previous behavior, they can pass an
alternate context:
<div class="highlight-python3 notranslate"><div class="highlight"><pre>import urllib2
import ssl

# This disables all verification
context = ssl._create_unverified_context()

# This allows using a specific certificate for the host, which doesn&#39;t need
# to be in the trust store
context = ssl.create_default_context(cafile=&quot;/path/to/file.crt&quot;)

urllib2.urlopen(&quot;https://invalid-cert&quot;, context=context)
</pre></div>
</div>
</div>
<div class="section" id="pep-493-https-verification-migration-tools-for-python-2-7">
<h3>PEP 493: HTTPS verification migration tools for Python 2.7<a class="headerlink" href="#pep-493-https-verification-migration-tools-for-python-2-7" title="Permalink to this headline">¶</a></h3>
<a class="pep reference external" href="https://www.python.org/dev/peps/pep-0493">PEP 493</a> provides additional migration tools to support a more incremental
infrastructure upgrade process for environments containing applications and
services relying on the historically permissive processing of server
certificates when establishing client HTTPS connections. These additions were
made in the Python 2.7.12 release.
These tools are intended for use in cases where affected applications and
services can’t be modified to explicitly pass a more permissive SSL context
when establishing the connection.
For applications and services which can’t be modified at all, the new
<code class="docutils literal notranslate">PYTHONHTTPSVERIFY</code> environment variable may be set to <code class="docutils literal notranslate">0</code> to revert an
entire Python process back to the default permissive behaviour of Python 2.7.8
and earlier.
For cases where the connection establishment code can’t be modified, but the
overall application can be, the new <code class="xref py py-func docutils literal notranslate">ssl._https_verify_certificates()</code>
function can be used to adjust the default behaviour at runtime.
</div>
<div class="section" id="new-make-regen-all-build-target">
<h3>New <code class="docutils literal notranslate">make regen-all</code> build target<a class="headerlink" href="#new-make-regen-all-build-target" title="Permalink to this headline">¶</a></h3>
To simplify cross-compilation, and to ensure that CPython can reliably be
compiled without requiring an existing version of Python to already be
available, the autotools-based build system no longer attempts to implicitly
recompile generated files based on file modification times.
Instead, a new <code class="docutils literal notranslate">make regen-all</code> command has been added to force regeneration
of these files when desired (e.g. after an initial version of Python has
already been built based on the pregenerated versions).
More selective regeneration targets are also defined - see
<a class="reference external" href="https://github.com/python/cpython/tree/3.6/Makefile.pre.in">Makefile.pre.in</a> for details.
(Contributed by Victor Stinner in <a class="reference external" href="https://bugs.python.org/issue23404">bpo-23404</a>.)
<div class="versionadded">
New in version 2.7.14.
</div>
</div>
<div class="section" id="removal-of-make-touch-build-target">
<h3>Removal of <code class="docutils literal notranslate">make touch</code> build target<a class="headerlink" href="#removal-of-make-touch-build-target" title="Permalink to this headline">¶</a></h3>
The <code class="docutils literal notranslate">make touch</code> build target previously used to request implicit regeneration
of generated files by updating their modification times has been removed.
It has been replaced by the new <code class="docutils literal notranslate">make regen-all</code> target.
(Contributed by Victor Stinner in <a class="reference external" href="https://bugs.python.org/issue23404">bpo-23404</a>.)
<div class="versionchanged">
Changed in version 2.7.14.
</div>
</div>
</div>
<div class="section" id="acknowledgements">
<h2>Acknowledgements<a class="headerlink" href="#acknowledgements" title="Permalink to this headline">¶</a></h2>
The author would like to thank the following people for offering
suggestions, corrections and assistance with various drafts of this
article: Nick Coghlan, Philip Jenvey, Ryan Lovett, R. David Murray,
Hugh Secker-Walker.
</div>
</div>

</div>
 </div>
 </div>
 <div class="sphinxsidebar" role="navigation" aria-label="main navigation">
 <div class="sphinxsidebarwrapper">
 <h3><a href="../contents.html">Table Of Contents</a></h3>
 <ul>
<li><a class="reference internal" href="#">What’s New in Python 2.7</a><ul>
<li><a class="reference internal" href="#the-future-for-python-2-x">The Future for Python 2.x</a></li>
<li><a class="reference internal" href="#changes-to-the-handling-of-deprecation-warnings">Changes to the Handling of Deprecation Warnings</a></li>
<li><a class="reference internal" href="#python-3-1-features">Python 3.1 Features</a></li>
<li><a class="reference internal" href="#pep-372-adding-an-ordered-dictionary-to-collections">PEP 372: Adding an Ordered Dictionary to collections</a></li>
<li><a class="reference internal" href="#pep-378-format-specifier-for-thousands-separator">PEP 378: Format Specifier for Thousands Separator</a></li>
<li><a class="reference internal" href="#pep-389-the-argparse-module-for-parsing-command-lines">PEP 389: The argparse Module for Parsing Command Lines</a></li>
<li><a class="reference internal" href="#pep-391-dictionary-based-configuration-for-logging">PEP 391: Dictionary-Based Configuration For Logging</a></li>
<li><a class="reference internal" href="#pep-3106-dictionary-views">PEP 3106: Dictionary Views</a></li>
<li><a class="reference internal" href="#pep-3137-the-memoryview-object">PEP 3137: The memoryview Object</a></li>
<li><a class="reference internal" href="#other-language-changes">Other Language Changes</a><ul>
<li><a class="reference internal" href="#interpreter-changes">Interpreter Changes</a></li>
<li><a class="reference internal" href="#optimizations">Optimizations</a></li>
</ul>
</li>
<li><a class="reference internal" href="#new-and-improved-modules">New and Improved Modules</a><ul>
<li><a class="reference internal" href="#new-module-importlib">New module: importlib</a></li>
<li><a class="reference internal" href="#new-module-sysconfig">New module: sysconfig</a></li>
<li><a class="reference internal" href="#ttk-themed-widgets-for-tk">ttk: Themed Widgets for Tk</a></li>
<li><a class="reference internal" href="#updated-module-unittest">Updated module: unittest</a></li>
<li><a class="reference internal" href="#updated-module-elementtree-1-3">Updated module: ElementTree 1.3</a></li>
</ul>
</li>
<li><a class="reference internal" href="#build-and-c-api-changes">Build and C API Changes</a><ul>
<li><a class="reference internal" href="#capsules">Capsules</a></li>
<li><a class="reference internal" href="#port-specific-changes-windows">Port-Specific Changes: Windows</a></li>
<li><a class="reference internal" href="#port-specific-changes-mac-os-x">Port-Specific Changes: Mac OS X</a></li>
<li><a class="reference internal" href="#port-specific-changes-freebsd">Port-Specific Changes: FreeBSD</a></li>
</ul>
</li>
<li><a class="reference internal" href="#other-changes-and-fixes">Other Changes and Fixes</a></li>
<li><a class="reference internal" href="#porting-to-python-2-7">Porting to Python 2.7</a></li>
<li><a class="reference internal" href="#new-features-added-to-python-2-7-maintenance-releases">New Features Added to Python 2.7 Maintenance Releases</a><ul>
<li><a class="reference internal" href="#two-new-environment-variables-for-debug-mode">Two new environment variables for debug mode</a></li>
<li><a class="reference internal" href="#pep-434-idle-enhancement-exception-for-all-branches">PEP 434: IDLE Enhancement Exception for All Branches</a></li>
<li><a class="reference internal" href="#pep-466-network-security-enhancements-for-python-2-7">PEP 466: Network Security Enhancements for Python 2.7</a></li>
<li><a class="reference internal" href="#pep-477-backport-ensurepip-pep-453-to-python-2-7">PEP 477: Backport ensurepip (PEP 453) to Python 2.7</a><ul>
<li><a class="reference internal" href="#bootstrapping-pip-by-default">Bootstrapping pip By Default</a></li>
<li><a class="reference internal" href="#documentation-changes">Documentation Changes</a></li>
</ul>
</li>
<li><a class="reference internal" href="#pep-476-enabling-certificate-verification-by-default-for-stdlib-http-clients">PEP 476: Enabling certificate verification by default for stdlib http clients</a></li>
<li><a class="reference internal" href="#pep-493-https-verification-migration-tools-for-python-2-7">PEP 493: HTTPS verification migration tools for Python 2.7</a></li>
<li><a class="reference internal" href="#new-make-regen-all-build-target">New <code class="docutils literal notranslate">make regen-all</code> build target</a></li>
<li><a class="reference internal" href="#removal-of-make-touch-build-target">Removal of <code class="docutils literal notranslate">make touch</code> build target</a></li>
</ul>
</li>
<li><a class="reference internal" href="#acknowledgements">Acknowledgements</a></li>
</ul>
</li>
</ul>

<h4>Previous topic</h4>
 <a href="3.0.html"
 title="previous chapter">What’s New In Python 3.0</a>
 <h4>Next topic</h4>
 <a href="2.6.html"
 title="next chapter">What’s New in Python 2.6</a>
 <div role="note" aria-label="source link">
 <h3>This Page</h3>
 <ul class="this-page-menu">
 <li><a href="../bugs.html">Report a Bug</a></li>
 <li>
 <a href="https://github.com/python/cpython/blob/3.6/Doc/whatsnew/2.7.rst"
 rel="nofollow">Show Source
 </a>
 </li>
 </ul>
 </div>
 </div>
 </div>
 <div class="clearer"></div>
 </div> 
 <div class="related" role="navigation" aria-label="related navigation">
 <h3>Navigation</h3>
 <ul>
 <li class="right" style="margin-right: 10px">
 <a href="../genindex.html" title="General Index"
 >index</a></li>
 <li class="right" >
 <a href="../py-modindex.html" title="Python Module Index"
 >modules</a> |</li>
 <li class="right" >
 <a href="2.6.html" title="What’s New in Python 2.6"
 >next</a> |</li>
 <li class="right" >
 <a href="3.0.html" title="What’s New In Python 3.0"
 >previous</a> |</li>
 <li><img src="../_static/py.png" alt=""
 style="vertical-align: middle; margin-top: -1px"/></li>
 <li><a href="https://www.python.org/">Python</a> &#187;</li>
 <li>
 <a href="../index.html">3.6.7 Documentation</a> &#187;
 </li>

<li class="nav-item nav-item-1"><a href="index.html" >What’s New in Python</a> &#187;</li>
 <li class="right">

</ul>
 </div> 
 <div class="footer">
 &copy; <a href="../copyright.html">Copyright</a> 2001-2023, Python Software Foundation.
 
 The Python Software Foundation is a non-profit corporation.
 <a href="https://www.python.org/psf/donations/">Please donate.</a>
 
 Last updated on Dec 18, 2023.
 <a href="../bugs.html">Found a bug</a>?
 
 Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.7.6.
 </div>

</body>
</html>

MR SKB