<?xml version="1.0" encoding="utf-8" ?>
<!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" xml:lang="en" lang="en">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<meta name="generator" content="Docutils 0.5: http://docutils.sourceforge.net/" />
<title>The decorator module</title>
<meta name="author" content="Michele Simionato" />
<style type="text/css">
.highlight { background: #f8f8f8; }
.highlight .c { color: #408080; font-style: italic } /* Comment */
.highlight .err { border: 1px solid #FF0000 } /* Error */
.highlight .k { color: #008000; font-weight: bold } /* Keyword */
.highlight .o { color: #666666 } /* Operator */
.highlight .cm { color: #408080; font-style: italic } /* Comment.Multiline */
.highlight .cp { color: #BC7A00 } /* Comment.Preproc */
.highlight .c1 { color: #408080; font-style: italic } /* Comment.Single */
.highlight .cs { color: #408080; font-style: italic } /* Comment.Special */
.highlight .gd { color: #A00000 } /* Generic.Deleted */
.highlight .ge { font-style: italic } /* Generic.Emph */
.highlight .gr { color: #FF0000 } /* Generic.Error */
.highlight .gh { color: #000080; font-weight: bold } /* Generic.Heading */
.highlight .gi { color: #00A000 } /* Generic.Inserted */
.highlight .go { color: #808080 } /* Generic.Output */
.highlight .gp { color: #000080; font-weight: bold } /* Generic.Prompt */
.highlight .gs { font-weight: bold } /* Generic.Strong */
.highlight .gu { color: #800080; font-weight: bold } /* Generic.Subheading */
.highlight .gt { color: #0040D0 } /* Generic.Traceback */
.highlight .kc { color: #008000; font-weight: bold } /* Keyword.Constant */
.highlight .kd { color: #008000; font-weight: bold } /* Keyword.Declaration */
.highlight .kp { color: #008000 } /* Keyword.Pseudo */
.highlight .kr { color: #008000; font-weight: bold } /* Keyword.Reserved */
.highlight .kt { color: #008000; font-weight: bold } /* Keyword.Type */
.highlight .m { color: #666666 } /* Literal.Number */
.highlight .s { color: #BA2121 } /* Literal.String */
.highlight .na { color: #7D9029 } /* Name.Attribute */
.highlight .nb { color: #008000 } /* Name.Builtin */
.highlight .nc { color: #0000FF; font-weight: bold } /* Name.Class */
.highlight .no { color: #880000 } /* Name.Constant */
.highlight .nd { color: #AA22FF } /* Name.Decorator */
.highlight .ni { color: #999999; font-weight: bold } /* Name.Entity */
.highlight .ne { color: #D2413A; font-weight: bold } /* Name.Exception */
.highlight .nf { color: #0000FF } /* Name.Function */
.highlight .nl { color: #A0A000 } /* Name.Label */
.highlight .nn { color: #0000FF; font-weight: bold } /* Name.Namespace */
.highlight .nt { color: #008000; font-weight: bold } /* Name.Tag */
.highlight .nv { color: #19177C } /* Name.Variable */
.highlight .ow { color: #AA22FF; font-weight: bold } /* Operator.Word */
.highlight .w { color: #bbbbbb } /* Text.Whitespace */
.highlight .mf { color: #666666 } /* Literal.Number.Float */
.highlight .mh { color: #666666 } /* Literal.Number.Hex */
.highlight .mi { color: #666666 } /* Literal.Number.Integer */
.highlight .mo { color: #666666 } /* Literal.Number.Oct */
.highlight .sb { color: #BA2121 } /* Literal.String.Backtick */
.highlight .sc { color: #BA2121 } /* Literal.String.Char */
.highlight .sd { color: #BA2121; font-style: italic } /* Literal.String.Doc */
.highlight .s2 { color: #BA2121 } /* Literal.String.Double */
.highlight .se { color: #BB6622; font-weight: bold } /* Literal.String.Escape */
.highlight .sh { color: #BA2121 } /* Literal.String.Heredoc */
.highlight .si { color: #BB6688; font-weight: bold } /* Literal.String.Interpol */
.highlight .sx { color: #008000 } /* Literal.String.Other */
.highlight .sr { color: #BB6688 } /* Literal.String.Regex */
.highlight .s1 { color: #BA2121 } /* Literal.String.Single */
.highlight .ss { color: #19177C } /* Literal.String.Symbol */
.highlight .bp { color: #008000 } /* Name.Builtin.Pseudo */
.highlight .vc { color: #19177C } /* Name.Variable.Class */
.highlight .vg { color: #19177C } /* Name.Variable.Global */
.highlight .vi { color: #19177C } /* Name.Variable.Instance */
.highlight .il { color: #666666 } /* Literal.Number.Integer.Long */
</style>
</head>
<body>
<div class="document" id="the-decorator-module">
<h1 class="title">The <tt class="docutils literal"><span class="pre">decorator</span></tt> module</h1>
<table class="docinfo" frame="void" rules="none">
<col class="docinfo-name" />
<col class="docinfo-content" />
<tbody valign="top">
<tr><th class="docinfo-name">Author:</th>
<td>Michele Simionato</td></tr>
<tr class="field"><th class="docinfo-name">E-mail:</th><td class="field-body"><a class="reference external" href="mailto:michele.simionato@gmail.com">michele.simionato@gmail.com</a></td>
</tr>
<tr><th class="docinfo-name">Version:</th>
<td>3.1.1 (2009-08-25)</td></tr>
<tr class="field"><th class="docinfo-name">Requires:</th><td class="field-body">Python 2.4+</td>
</tr>
<tr class="field"><th class="docinfo-name">Download page:</th><td class="field-body"><a class="reference external" href="http://pypi.python.org/pypi/decorator/3.1.1">http://pypi.python.org/pypi/decorator/3.1.1</a></td>
</tr>
<tr class="field"><th class="docinfo-name">Installation:</th><td class="field-body"><tt class="docutils literal"><span class="pre">easy_install</span> <span class="pre">decorator</span></tt></td>
</tr>
<tr class="field"><th class="docinfo-name">License:</th><td class="field-body">BSD license</td>
</tr>
</tbody>
</table>
<div class="contents topic" id="contents">
<p class="topic-title first">Contents</p>
<ul class="simple">
<li><a class="reference internal" href="#introduction" id="id3">Introduction</a></li>
<li><a class="reference internal" href="#definitions" id="id4">Definitions</a></li>
<li><a class="reference internal" href="#statement-of-the-problem" id="id5">Statement of the problem</a></li>
<li><a class="reference internal" href="#the-solution" id="id6">The solution</a></li>
<li><a class="reference internal" href="#a-trace-decorator" id="id7">A <tt class="docutils literal"><span class="pre">trace</span></tt> decorator</a></li>
<li><a class="reference internal" href="#decorator-is-a-decorator" id="id8"><tt class="docutils literal"><span class="pre">decorator</span></tt> is a decorator</a></li>
<li><a class="reference internal" href="#blocking" id="id9"><tt class="docutils literal"><span class="pre">blocking</span></tt></a></li>
<li><a class="reference internal" href="#async" id="id10"><tt class="docutils literal"><span class="pre">async</span></tt></a></li>
<li><a class="reference internal" href="#the-functionmaker-class" id="id11">The <tt class="docutils literal"><span class="pre">FunctionMaker</span></tt> class</a></li>
<li><a class="reference internal" href="#getting-the-source-code" id="id12">Getting the source code</a></li>
<li><a class="reference internal" href="#dealing-with-third-party-decorators" id="id13">Dealing with third party decorators</a></li>
<li><a class="reference internal" href="#caveats-and-limitations" id="id14">Caveats and limitations</a></li>
<li><a class="reference internal" href="#compatibility-notes" id="id15">Compatibility notes</a></li>
<li><a class="reference internal" href="#licence" id="id16">LICENCE</a></li>
</ul>
</div>
<div class="section" id="introduction">
<h1><a class="toc-backref" href="#id3">Introduction</a></h1>
<p>Python decorators are an interesting example of why syntactic sugar
matters. In principle, their introduction in Python 2.4 changed
nothing, since they do not provide any new functionality which was not
already present in the language. In practice, their introduction has
significantly changed the way we structure our programs in Python. I
believe the change is for the best, and that decorators are a great
idea since:</p>
<ul class="simple">
<li>decorators help reducing boilerplate code;</li>
<li>decorators help separation of concerns;</li>
<li>decorators enhance readability and maintenability;</li>
<li>decorators are explicit.</li>
</ul>
<p>Still, as of now, writing custom decorators correctly requires
some experience and it is not as easy as it could be. For instance,
typical implementations of decorators involve nested functions, and
we all know that flat is better than nested.</p>
<p>The aim of the <tt class="docutils literal"><span class="pre">decorator</span></tt> module it to simplify the usage of
decorators for the average programmer, and to popularize decorators by
showing various non-trivial examples. Of course, as all techniques,
decorators can be abused (I have seen that) and you should not try to
solve every problem with a decorator, just because you can.</p>
<p>You may find the source code for all the examples
discussed here in the <tt class="docutils literal"><span class="pre">documentation.py</span></tt> file, which contains
this documentation in the form of doctests.</p>
</div>
<div class="section" id="definitions">
<h1><a class="toc-backref" href="#id4">Definitions</a></h1>
<p>Technically speaking, any Python object which can be called with one argument
can be used as a decorator. However, this definition is somewhat too large
to be really useful. It is more convenient to split the generic class of
decorators in two subclasses:</p>
<ul class="simple">
<li><em>signature-preserving</em> decorators, i.e. callable objects taking a
function as input and returning a function <em>with the same
signature</em> as output;</li>
<li><em>signature-changing</em> decorators, i.e. decorators that change
the signature of their input function, or decorators returning
non-callable objects.</li>
</ul>
<p>Signature-changing decorators have their use: for instance the
builtin classes <tt class="docutils literal"><span class="pre">staticmethod</span></tt> and <tt class="docutils literal"><span class="pre">classmethod</span></tt> are in this
group, since they take functions and return descriptor objects which
are not functions, nor callables.</p>
<p>However, signature-preserving decorators are more common and easier to
reason about; in particular signature-preserving decorators can be
composed together whereas other decorators in general cannot.</p>
<p>Writing signature-preserving decorators from scratch is not that
obvious, especially if one wants to define proper decorators that
can accept functions with any signature. A simple example will clarify
the issue.</p>
</div>
<div class="section" id="statement-of-the-problem">
<h1><a class="toc-backref" href="#id5">Statement of the problem</a></h1>
<p>A very common use case for decorators is the memoization of functions.
A <tt class="docutils literal"><span class="pre">memoize</span></tt> decorator works by caching
the result of the function call in a dictionary, so that the next time
the function is called with the same input parameters the result is retrieved
from the cache and not recomputed. There are many implementations of
<tt class="docutils literal"><span class="pre">memoize</span></tt> in <a class="reference external" href="http://www.python.org/moin/PythonDecoratorLibrary">http://www.python.org/moin/PythonDecoratorLibrary</a>,
but they do not preserve the signature.
A simple implementation for Python 2.5 could be the following (notice
that in general it is impossible to memoize correctly something
that depends on non-hashable arguments):</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="k">def</span> <span class="nf">memoize25</span><span class="p">(</span><span class="n">func</span><span class="p">):</span>
<span class="n">func</span><span class="o">.</span><span class="n">cache</span> <span class="o">=</span> <span class="p">{}</span>
<span class="k">def</span> <span class="nf">memoize</span><span class="p">(</span><span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">):</span>
<span class="k">if</span> <span class="n">kw</span><span class="p">:</span> <span class="c"># frozenset is used to ensure hashability</span>
<span class="n">key</span> <span class="o">=</span> <span class="n">args</span><span class="p">,</span> <span class="n">frozenset</span><span class="p">(</span><span class="n">kw</span><span class="o">.</span><span class="n">iteritems</span><span class="p">())</span>
<span class="k">else</span><span class="p">:</span>
<span class="n">key</span> <span class="o">=</span> <span class="n">args</span>
<span class="n">cache</span> <span class="o">=</span> <span class="n">func</span><span class="o">.</span><span class="n">cache</span>
<span class="k">if</span> <span class="n">key</span> <span class="ow">in</span> <span class="n">cache</span><span class="p">:</span>
<span class="k">return</span> <span class="n">cache</span><span class="p">[</span><span class="n">key</span><span class="p">]</span>
<span class="k">else</span><span class="p">:</span>
<span class="n">cache</span><span class="p">[</span><span class="n">key</span><span class="p">]</span> <span class="o">=</span> <span class="n">result</span> <span class="o">=</span> <span class="n">func</span><span class="p">(</span><span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">)</span>
<span class="k">return</span> <span class="n">result</span>
<span class="k">return</span> <span class="n">functools</span><span class="o">.</span><span class="n">update_wrapper</span><span class="p">(</span><span class="n">memoize</span><span class="p">,</span> <span class="n">func</span><span class="p">)</span>
</pre></div>
</div>
<p>Here we used the <a class="reference external" href="http://www.python.org/doc/2.5.2/lib/module-functools.html">functools.update_wrapper</a> utility, which has
been added in Python 2.5 expressly to simplify the definition of decorators
(in older versions of Python you need to copy the function attributes
<tt class="docutils literal"><span class="pre">__name__</span></tt>, <tt class="docutils literal"><span class="pre">__doc__</span></tt>, <tt class="docutils literal"><span class="pre">__module__</span></tt> and <tt class="docutils literal"><span class="pre">__dict__</span></tt>
from the original function to the decorated function by hand).</p>
<p>The implementation above works in the sense that the decorator
can accept functions with generic signatures; unfortunately this
implementation does <em>not</em> define a signature-preserving decorator, since in
general <tt class="docutils literal"><span class="pre">memoize25</span></tt> returns a function with a
<em>different signature</em> from the original function.</p>
<p>Consider for instance the following case:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="nd">@memoize25</span>
<span class="o">...</span> <span class="k">def</span> <span class="nf">f1</span><span class="p">(</span><span class="n">x</span><span class="p">):</span>
<span class="o">...</span> <span class="n">time</span><span class="o">.</span><span class="n">sleep</span><span class="p">(</span><span class="mf">1</span><span class="p">)</span> <span class="c"># simulate some long computation</span>
<span class="o">...</span> <span class="k">return</span> <span class="n">x</span>
</pre></div>
</div>
<p>Here the original function takes a single argument named <tt class="docutils literal"><span class="pre">x</span></tt>,
but the decorated function takes any number of arguments and
keyword arguments:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="kn">from</span> <span class="nn">inspect</span> <span class="kn">import</span> <span class="n">getargspec</span>
<span class="o">>>></span> <span class="k">print</span> <span class="n">getargspec</span><span class="p">(</span><span class="n">f1</span><span class="p">)</span>
<span class="p">([],</span> <span class="s">'args'</span><span class="p">,</span> <span class="s">'kw'</span><span class="p">,</span> <span class="bp">None</span><span class="p">)</span>
</pre></div>
</div>
<p>This means that introspection tools such as pydoc will give
wrong informations about the signature of <tt class="docutils literal"><span class="pre">f1</span></tt>. This is pretty bad:
pydoc will tell you that the function accepts a generic signature
<tt class="docutils literal"><span class="pre">*args</span></tt>, <tt class="docutils literal"><span class="pre">**kw</span></tt>, but when you try to call the function with more than an
argument, you will get an error:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="n">f1</span><span class="p">(</span><span class="mf">0</span><span class="p">,</span> <span class="mf">1</span><span class="p">)</span>
<span class="n">Traceback</span> <span class="p">(</span><span class="n">most</span> <span class="n">recent</span> <span class="n">call</span> <span class="n">last</span><span class="p">):</span>
<span class="o">...</span>
<span class="ne">TypeError</span><span class="p">:</span> <span class="n">f1</span><span class="p">()</span> <span class="n">takes</span> <span class="n">exactly</span> <span class="mf">1</span> <span class="n">argument</span> <span class="p">(</span><span class="mf">2</span> <span class="n">given</span><span class="p">)</span>
</pre></div>
</div>
</div>
<div class="section" id="the-solution">
<h1><a class="toc-backref" href="#id6">The solution</a></h1>
<p>The solution is to provide a generic factory of generators, which
hides the complexity of making signature-preserving decorators
from the application programmer. The <tt class="docutils literal"><span class="pre">decorator</span></tt> function in
the <tt class="docutils literal"><span class="pre">decorator</span></tt> module is such a factory:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="kn">from</span> <span class="nn">decorator</span> <span class="kn">import</span> <span class="n">decorator</span>
</pre></div>
</div>
<p><tt class="docutils literal"><span class="pre">decorator</span></tt> takes two arguments, a caller function describing the
functionality of the decorator and a function to be decorated; it
returns the decorated function. The caller function must have
signature <tt class="docutils literal"><span class="pre">(f,</span> <span class="pre">*args,</span> <span class="pre">**kw)</span></tt> and it must call the original function <tt class="docutils literal"><span class="pre">f</span></tt>
with arguments <tt class="docutils literal"><span class="pre">args</span></tt> and <tt class="docutils literal"><span class="pre">kw</span></tt>, implementing the wanted capability,
i.e. memoization in this case:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="k">def</span> <span class="nf">_memoize</span><span class="p">(</span><span class="n">func</span><span class="p">,</span> <span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">):</span>
<span class="k">if</span> <span class="n">kw</span><span class="p">:</span> <span class="c"># frozenset is used to ensure hashability</span>
<span class="n">key</span> <span class="o">=</span> <span class="n">args</span><span class="p">,</span> <span class="n">frozenset</span><span class="p">(</span><span class="n">kw</span><span class="o">.</span><span class="n">iteritems</span><span class="p">())</span>
<span class="k">else</span><span class="p">:</span>
<span class="n">key</span> <span class="o">=</span> <span class="n">args</span>
<span class="n">cache</span> <span class="o">=</span> <span class="n">func</span><span class="o">.</span><span class="n">cache</span> <span class="c"># attributed added by memoize</span>
<span class="k">if</span> <span class="n">key</span> <span class="ow">in</span> <span class="n">cache</span><span class="p">:</span>
<span class="k">return</span> <span class="n">cache</span><span class="p">[</span><span class="n">key</span><span class="p">]</span>
<span class="k">else</span><span class="p">:</span>
<span class="n">cache</span><span class="p">[</span><span class="n">key</span><span class="p">]</span> <span class="o">=</span> <span class="n">result</span> <span class="o">=</span> <span class="n">func</span><span class="p">(</span><span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">)</span>
<span class="k">return</span> <span class="n">result</span>
</pre></div>
</div>
<p>At this point you can define your decorator as follows:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="k">def</span> <span class="nf">memoize</span><span class="p">(</span><span class="n">f</span><span class="p">):</span>
<span class="n">f</span><span class="o">.</span><span class="n">cache</span> <span class="o">=</span> <span class="p">{}</span>
<span class="k">return</span> <span class="n">decorator</span><span class="p">(</span><span class="n">_memoize</span><span class="p">,</span> <span class="n">f</span><span class="p">)</span>
</pre></div>
</div>
<p>The difference with respect to the Python 2.5 approach, which is based
on nested functions, is that the decorator module forces you to lift
the inner function at the outer level (<em>flat is better than nested</em>).
Moreover, you are forced to pass explicitly the function you want to
decorate to the caller function.</p>
<p>Here is a test of usage:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="nd">@memoize</span>
<span class="o">...</span> <span class="k">def</span> <span class="nf">heavy_computation</span><span class="p">():</span>
<span class="o">...</span> <span class="n">time</span><span class="o">.</span><span class="n">sleep</span><span class="p">(</span><span class="mf">2</span><span class="p">)</span>
<span class="o">...</span> <span class="k">return</span> <span class="s">"done"</span>
<span class="o">>>></span> <span class="k">print</span> <span class="n">heavy_computation</span><span class="p">()</span> <span class="c"># the first time it will take 2 seconds</span>
<span class="n">done</span>
<span class="o">>>></span> <span class="k">print</span> <span class="n">heavy_computation</span><span class="p">()</span> <span class="c"># the second time it will be instantaneous</span>
<span class="n">done</span>
</pre></div>
</div>
<p>The signature of <tt class="docutils literal"><span class="pre">heavy_computation</span></tt> is the one you would expect:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="k">print</span> <span class="n">getargspec</span><span class="p">(</span><span class="n">heavy_computation</span><span class="p">)</span>
<span class="p">([],</span> <span class="bp">None</span><span class="p">,</span> <span class="bp">None</span><span class="p">,</span> <span class="bp">None</span><span class="p">)</span>
</pre></div>
</div>
</div>
<div class="section" id="a-trace-decorator">
<h1><a class="toc-backref" href="#id7">A <tt class="docutils literal"><span class="pre">trace</span></tt> decorator</a></h1>
<p>As an additional example, here is how you can define a trivial
<tt class="docutils literal"><span class="pre">trace</span></tt> decorator, which prints a message everytime the traced
function is called:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="k">def</span> <span class="nf">_trace</span><span class="p">(</span><span class="n">f</span><span class="p">,</span> <span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">):</span>
<span class="k">print</span> <span class="s">"calling </span><span class="si">%s</span><span class="s"> with args </span><span class="si">%s</span><span class="s">, </span><span class="si">%s</span><span class="s">"</span> <span class="o">%</span> <span class="p">(</span><span class="n">f</span><span class="o">.</span><span class="n">__name__</span><span class="p">,</span> <span class="n">args</span><span class="p">,</span> <span class="n">kw</span><span class="p">)</span>
<span class="k">return</span> <span class="n">f</span><span class="p">(</span><span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">)</span>
</pre></div>
</div>
<div class="codeblock python">
<div class="highlight"><pre><span class="k">def</span> <span class="nf">trace</span><span class="p">(</span><span class="n">f</span><span class="p">):</span>
<span class="k">return</span> <span class="n">decorator</span><span class="p">(</span><span class="n">_trace</span><span class="p">,</span> <span class="n">f</span><span class="p">)</span>
</pre></div>
</div>
<p>Here is an example of usage:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="nd">@trace</span>
<span class="o">...</span> <span class="k">def</span> <span class="nf">f1</span><span class="p">(</span><span class="n">x</span><span class="p">):</span>
<span class="o">...</span> <span class="k">pass</span>
</pre></div>
</div>
<p>It is immediate to verify that <tt class="docutils literal"><span class="pre">f1</span></tt> works</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="n">f1</span><span class="p">(</span><span class="mf">0</span><span class="p">)</span>
<span class="n">calling</span> <span class="n">f1</span> <span class="k">with</span> <span class="n">args</span> <span class="p">(</span><span class="mf">0</span><span class="p">,),</span> <span class="p">{}</span>
</pre></div>
</div>
<p>and it that it has the correct signature:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="k">print</span> <span class="n">getargspec</span><span class="p">(</span><span class="n">f1</span><span class="p">)</span>
<span class="p">([</span><span class="s">'x'</span><span class="p">],</span> <span class="bp">None</span><span class="p">,</span> <span class="bp">None</span><span class="p">,</span> <span class="bp">None</span><span class="p">)</span>
</pre></div>
</div>
<p>The same decorator works with functions of any signature:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="nd">@trace</span>
<span class="o">...</span> <span class="k">def</span> <span class="nf">f</span><span class="p">(</span><span class="n">x</span><span class="p">,</span> <span class="n">y</span><span class="o">=</span><span class="mf">1</span><span class="p">,</span> <span class="n">z</span><span class="o">=</span><span class="mf">2</span><span class="p">,</span> <span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">):</span>
<span class="o">...</span> <span class="k">pass</span>
<span class="o">>>></span> <span class="n">f</span><span class="p">(</span><span class="mf">0</span><span class="p">,</span> <span class="mf">3</span><span class="p">)</span>
<span class="n">calling</span> <span class="n">f</span> <span class="k">with</span> <span class="n">args</span> <span class="p">(</span><span class="mf">0</span><span class="p">,</span> <span class="mf">3</span><span class="p">,</span> <span class="mf">2</span><span class="p">),</span> <span class="p">{}</span>
<span class="o">>>></span> <span class="k">print</span> <span class="n">getargspec</span><span class="p">(</span><span class="n">f</span><span class="p">)</span>
<span class="p">([</span><span class="s">'x'</span><span class="p">,</span> <span class="s">'y'</span><span class="p">,</span> <span class="s">'z'</span><span class="p">],</span> <span class="s">'args'</span><span class="p">,</span> <span class="s">'kw'</span><span class="p">,</span> <span class="p">(</span><span class="mf">1</span><span class="p">,</span> <span class="mf">2</span><span class="p">))</span>
</pre></div>
</div>
<p>That includes even functions with exotic signatures like the following:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="nd">@trace</span>
<span class="o">...</span> <span class="k">def</span> <span class="nf">exotic_signature</span><span class="p">((</span><span class="n">x</span><span class="p">,</span> <span class="n">y</span><span class="p">)</span><span class="o">=</span><span class="p">(</span><span class="mf">1</span><span class="p">,</span><span class="mf">2</span><span class="p">)):</span> <span class="k">return</span> <span class="n">x</span><span class="o">+</span><span class="n">y</span>
<span class="o">>>></span> <span class="k">print</span> <span class="n">getargspec</span><span class="p">(</span><span class="n">exotic_signature</span><span class="p">)</span>
<span class="p">([[</span><span class="s">'x'</span><span class="p">,</span> <span class="s">'y'</span><span class="p">]],</span> <span class="bp">None</span><span class="p">,</span> <span class="bp">None</span><span class="p">,</span> <span class="p">((</span><span class="mf">1</span><span class="p">,</span> <span class="mf">2</span><span class="p">),))</span>
<span class="o">>>></span> <span class="n">exotic_signature</span><span class="p">()</span>
<span class="n">calling</span> <span class="n">exotic_signature</span> <span class="k">with</span> <span class="n">args</span> <span class="p">((</span><span class="mf">1</span><span class="p">,</span> <span class="mf">2</span><span class="p">),),</span> <span class="p">{}</span>
<span class="mf">3</span>
</pre></div>
</div>
<p>Notice that the support for exotic signatures has been deprecated
in Python 2.6 and removed in Python 3.0.</p>
</div>
<div class="section" id="decorator-is-a-decorator">
<h1><a class="toc-backref" href="#id8"><tt class="docutils literal"><span class="pre">decorator</span></tt> is a decorator</a></h1>
<p>It may be annoying to write a caller function (like the <tt class="docutils literal"><span class="pre">_trace</span></tt>
function above) and then a trivial wrapper
(<tt class="docutils literal"><span class="pre">def</span> <span class="pre">trace(f):</span> <span class="pre">return</span> <span class="pre">decorator(_trace,</span> <span class="pre">f)</span></tt>) every time. For this reason,
the <tt class="docutils literal"><span class="pre">decorator</span></tt> module provides an easy shortcut to convert
the caller function into a signature-preserving decorator:
you can just call <tt class="docutils literal"><span class="pre">decorator</span></tt> with a single argument.
In our example you can just write <tt class="docutils literal"><span class="pre">trace</span> <span class="pre">=</span> <span class="pre">decorator(_trace)</span></tt>.
The <tt class="docutils literal"><span class="pre">decorator</span></tt> function can also be used as a signature-changing
decorator, just as <tt class="docutils literal"><span class="pre">classmethod</span></tt> and <tt class="docutils literal"><span class="pre">staticmethod</span></tt>.
However, <tt class="docutils literal"><span class="pre">classmethod</span></tt> and <tt class="docutils literal"><span class="pre">staticmethod</span></tt> return generic
objects which are not callable, while <tt class="docutils literal"><span class="pre">decorator</span></tt> returns
signature-preserving decorators, i.e. functions of a single argument.
For instance, you can write directly</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="nd">@decorator</span>
<span class="o">...</span> <span class="k">def</span> <span class="nf">trace</span><span class="p">(</span><span class="n">f</span><span class="p">,</span> <span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">):</span>
<span class="o">...</span> <span class="k">print</span> <span class="s">"calling </span><span class="si">%s</span><span class="s"> with args </span><span class="si">%s</span><span class="s">, </span><span class="si">%s</span><span class="s">"</span> <span class="o">%</span> <span class="p">(</span><span class="n">f</span><span class="o">.</span><span class="n">func_name</span><span class="p">,</span> <span class="n">args</span><span class="p">,</span> <span class="n">kw</span><span class="p">)</span>
<span class="o">...</span> <span class="k">return</span> <span class="n">f</span><span class="p">(</span><span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">)</span>
</pre></div>
</div>
<p>and now <tt class="docutils literal"><span class="pre">trace</span></tt> will be a decorator. Actually <tt class="docutils literal"><span class="pre">trace</span></tt> is a <tt class="docutils literal"><span class="pre">partial</span></tt>
object which can be used as a decorator:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="n">trace</span>
<span class="o"><</span><span class="n">function</span> <span class="n">trace</span> <span class="n">at</span> <span class="mf">0</span><span class="n">x</span><span class="o">...></span>
</pre></div>
</div>
<p>Here is an example of usage:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="nd">@trace</span>
<span class="o">...</span> <span class="k">def</span> <span class="nf">func</span><span class="p">():</span> <span class="k">pass</span>
<span class="o">>>></span> <span class="n">func</span><span class="p">()</span>
<span class="n">calling</span> <span class="n">func</span> <span class="k">with</span> <span class="n">args</span> <span class="p">(),</span> <span class="p">{}</span>
</pre></div>
</div>
<p>If you are using an old Python version (Python 2.4) the
<tt class="docutils literal"><span class="pre">decorator</span></tt> module provides a poor man replacement for
<tt class="docutils literal"><span class="pre">functools.partial</span></tt>.</p>
<p>There is also an easy way to create one-parameter factories of
decorators, based on the following
<tt class="docutils literal"><span class="pre">decorator_factory</span></tt> utility:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="k">def</span> <span class="nf">decorator_factory</span><span class="p">(</span><span class="n">decfac</span><span class="p">):</span> <span class="c"># partial is functools.partial</span>
<span class="s">"decorator_factory(decfac) returns a one-parameter family of decorators"</span>
<span class="k">return</span> <span class="n">partial</span><span class="p">(</span><span class="k">lambda</span> <span class="n">df</span><span class="p">,</span> <span class="n">param</span><span class="p">:</span> <span class="n">decorator</span><span class="p">(</span><span class="n">partial</span><span class="p">(</span><span class="n">df</span><span class="p">,</span> <span class="n">param</span><span class="p">)),</span> <span class="n">decfac</span><span class="p">)</span>
</pre></div>
</div>
<p><tt class="docutils literal"><span class="pre">decorator_factory</span></tt> converts a function with signature
<tt class="docutils literal"><span class="pre">(param,</span> <span class="pre">func,</span> <span class="pre">*args,</span> <span class="pre">**kw)</span></tt> into a one-parameter family
of decorators. Suppose for instance you want to generated different
tracing generator, with different tracing messages.
Here is how to do it:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="nd">@decorator_factory</span>
<span class="o">...</span> <span class="k">def</span> <span class="nf">trace_factory</span><span class="p">(</span><span class="n">message_template</span><span class="p">,</span> <span class="n">f</span><span class="p">,</span> <span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">):</span>
<span class="o">...</span> <span class="n">name</span> <span class="o">=</span> <span class="n">f</span><span class="o">.</span><span class="n">func_name</span>
<span class="o">...</span> <span class="k">print</span> <span class="n">message_template</span> <span class="o">%</span> <span class="nb">locals</span><span class="p">()</span>
<span class="o">...</span> <span class="k">return</span> <span class="n">f</span><span class="p">(</span><span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">)</span>
</pre></div>
</div>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="n">trace_factory</span>
<span class="o"><</span><span class="n">functools</span><span class="o">.</span><span class="n">partial</span> <span class="nb">object</span> <span class="n">at</span> <span class="mf">0</span><span class="n">x</span><span class="o">...></span>
<span class="o">>>></span> <span class="n">trace</span> <span class="o">=</span> <span class="n">trace_factory</span><span class="p">(</span><span class="s">'Calling </span><span class="si">%(name)s</span><span class="s"> with args </span><span class="si">%(args)s</span><span class="s"> '</span>
<span class="o">...</span> <span class="s">'and keywords </span><span class="si">%(kw)s</span><span class="s">'</span><span class="p">)</span>
</pre></div>
</div>
<p>In this example the parameter (<tt class="docutils literal"><span class="pre">message_template</span></tt>) is
just a string, but in general it can be a tuple, a dictionary, or
a generic object, so there is no real restriction (for instance,
if you want to define a two-parameter family of decorators just
use a tuple with two arguments as parameter).
Here is an example of usage:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="nd">@trace</span>
<span class="o">...</span> <span class="k">def</span> <span class="nf">func</span><span class="p">():</span> <span class="k">pass</span>
<span class="o">>>></span> <span class="n">func</span><span class="p">()</span>
<span class="n">Calling</span> <span class="n">func</span> <span class="k">with</span> <span class="n">args</span> <span class="p">()</span> <span class="ow">and</span> <span class="n">keywords</span> <span class="p">{}</span>
</pre></div>
</div>
</div>
<div class="section" id="blocking">
<h1><a class="toc-backref" href="#id9"><tt class="docutils literal"><span class="pre">blocking</span></tt></a></h1>
<p>Sometimes one has to deal with blocking resources, such as <tt class="docutils literal"><span class="pre">stdin</span></tt>, and
sometimes it is best to have back a "busy" message than to block everything.
This behavior can be implemented with a suitable family of decorators,
where the parameter is the busy message:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="nd">@decorator_factory</span>
<span class="k">def</span> <span class="nf">blocking</span><span class="p">(</span><span class="n">not_avail</span><span class="p">,</span> <span class="n">f</span><span class="p">,</span> <span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">):</span>
<span class="k">if</span> <span class="ow">not</span> <span class="nb">hasattr</span><span class="p">(</span><span class="n">f</span><span class="p">,</span> <span class="s">"thread"</span><span class="p">):</span> <span class="c"># no thread running</span>
<span class="k">def</span> <span class="nf">set_result</span><span class="p">():</span> <span class="n">f</span><span class="o">.</span><span class="n">result</span> <span class="o">=</span> <span class="n">f</span><span class="p">(</span><span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">)</span>
<span class="n">f</span><span class="o">.</span><span class="n">thread</span> <span class="o">=</span> <span class="n">threading</span><span class="o">.</span><span class="n">Thread</span><span class="p">(</span><span class="bp">None</span><span class="p">,</span> <span class="n">set_result</span><span class="p">)</span>
<span class="n">f</span><span class="o">.</span><span class="n">thread</span><span class="o">.</span><span class="n">start</span><span class="p">()</span>
<span class="k">return</span> <span class="n">not_avail</span>
<span class="k">elif</span> <span class="n">f</span><span class="o">.</span><span class="n">thread</span><span class="o">.</span><span class="n">isAlive</span><span class="p">():</span>
<span class="k">return</span> <span class="n">not_avail</span>
<span class="k">else</span><span class="p">:</span> <span class="c"># the thread is ended, return the stored result</span>
<span class="k">del</span> <span class="n">f</span><span class="o">.</span><span class="n">thread</span>
<span class="k">return</span> <span class="n">f</span><span class="o">.</span><span class="n">result</span>
</pre></div>
</div>
<p>Functions decorated with <tt class="docutils literal"><span class="pre">blocking</span></tt> will return a busy message if
the resource is unavailable, and the intended result if the resource is
available. For instance:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="nd">@blocking</span><span class="p">(</span><span class="s">"Please wait ..."</span><span class="p">)</span>
<span class="o">...</span> <span class="k">def</span> <span class="nf">read_data</span><span class="p">():</span>
<span class="o">...</span> <span class="n">time</span><span class="o">.</span><span class="n">sleep</span><span class="p">(</span><span class="mf">3</span><span class="p">)</span> <span class="c"># simulate a blocking resource</span>
<span class="o">...</span> <span class="k">return</span> <span class="s">"some data"</span>
<span class="o">>>></span> <span class="k">print</span> <span class="n">read_data</span><span class="p">()</span> <span class="c"># data is not available yet</span>
<span class="n">Please</span> <span class="n">wait</span> <span class="o">...</span>
<span class="o">>>></span> <span class="n">time</span><span class="o">.</span><span class="n">sleep</span><span class="p">(</span><span class="mf">1</span><span class="p">)</span>
<span class="o">>>></span> <span class="k">print</span> <span class="n">read_data</span><span class="p">()</span> <span class="c"># data is not available yet</span>
<span class="n">Please</span> <span class="n">wait</span> <span class="o">...</span>
<span class="o">>>></span> <span class="n">time</span><span class="o">.</span><span class="n">sleep</span><span class="p">(</span><span class="mf">1</span><span class="p">)</span>
<span class="o">>>></span> <span class="k">print</span> <span class="n">read_data</span><span class="p">()</span> <span class="c"># data is not available yet</span>
<span class="n">Please</span> <span class="n">wait</span> <span class="o">...</span>
<span class="o">>>></span> <span class="n">time</span><span class="o">.</span><span class="n">sleep</span><span class="p">(</span><span class="mf">1.1</span><span class="p">)</span> <span class="c"># after 3.1 seconds, data is available</span>
<span class="o">>>></span> <span class="k">print</span> <span class="n">read_data</span><span class="p">()</span>
<span class="n">some</span> <span class="n">data</span>
</pre></div>
</div>
</div>
<div class="section" id="async">
<h1><a class="toc-backref" href="#id10"><tt class="docutils literal"><span class="pre">async</span></tt></a></h1>
<p>We have just seen an examples of a simple decorator factory,
implemented as a function returning a decorator.
For more complex situations, it is more
convenient to implement decorator factories as classes returning
callable objects that can be used as signature-preserving
decorators. The suggested pattern to do that is to introduce
a helper method <tt class="docutils literal"><span class="pre">call(self,</span> <span class="pre">func,</span> <span class="pre">*args,</span> <span class="pre">**kw)</span></tt> and to call
it in the <tt class="docutils literal"><span class="pre">__call__(self,</span> <span class="pre">func)</span></tt> method.</p>
<p>As an example, here I show a decorator
which is able to convert a blocking function into an asynchronous
function. The function, when called,
is executed in a separate thread. Moreover, it is possible to set
three callbacks <tt class="docutils literal"><span class="pre">on_success</span></tt>, <tt class="docutils literal"><span class="pre">on_failure</span></tt> and <tt class="docutils literal"><span class="pre">on_closing</span></tt>,
to specify how to manage the function call.
The implementation is the following:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="k">def</span> <span class="nf">on_success</span><span class="p">(</span><span class="n">result</span><span class="p">):</span> <span class="c"># default implementation</span>
<span class="s">"Called on the result of the function"</span>
<span class="k">return</span> <span class="n">result</span>
</pre></div>
</div>
<div class="codeblock python">
<div class="highlight"><pre><span class="k">def</span> <span class="nf">on_failure</span><span class="p">(</span><span class="n">exc_info</span><span class="p">):</span> <span class="c"># default implementation</span>
<span class="s">"Called if the function fails"</span>
<span class="k">pass</span>
</pre></div>
</div>
<div class="codeblock python">
<div class="highlight"><pre><span class="k">def</span> <span class="nf">on_closing</span><span class="p">():</span> <span class="c"># default implementation</span>
<span class="s">"Called at the end, both in case of success and failure"</span>
<span class="k">pass</span>
</pre></div>
</div>
<div class="codeblock python">
<div class="highlight"><pre><span class="k">class</span> <span class="nc">Async</span><span class="p">(</span><span class="nb">object</span><span class="p">):</span>
<span class="sd">"""</span>
<span class="sd"> A decorator converting blocking functions into asynchronous</span>
<span class="sd"> functions, by using threads or processes. Examples:</span>
<span class="sd"> async_with_threads = Async(threading.Thread)</span>
<span class="sd"> async_with_processes = Async(multiprocessing.Process)</span>
<span class="sd"> """</span>
<span class="k">def</span> <span class="nf">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">threadfactory</span><span class="p">):</span>
<span class="bp">self</span><span class="o">.</span><span class="n">threadfactory</span> <span class="o">=</span> <span class="n">threadfactory</span>
<span class="k">def</span> <span class="nf">__call__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">func</span><span class="p">,</span> <span class="n">on_success</span><span class="o">=</span><span class="n">on_success</span><span class="p">,</span>
<span class="n">on_failure</span><span class="o">=</span><span class="n">on_failure</span><span class="p">,</span> <span class="n">on_closing</span><span class="o">=</span><span class="n">on_closing</span><span class="p">):</span>
<span class="c"># every decorated function has its own independent thread counter</span>
<span class="n">func</span><span class="o">.</span><span class="n">counter</span> <span class="o">=</span> <span class="n">itertools</span><span class="o">.</span><span class="n">count</span><span class="p">(</span><span class="mf">1</span><span class="p">)</span>
<span class="n">func</span><span class="o">.</span><span class="n">on_success</span> <span class="o">=</span> <span class="n">on_success</span>
<span class="n">func</span><span class="o">.</span><span class="n">on_failure</span> <span class="o">=</span> <span class="n">on_failure</span>
<span class="n">func</span><span class="o">.</span><span class="n">on_closing</span> <span class="o">=</span> <span class="n">on_closing</span>
<span class="k">return</span> <span class="n">decorator</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">call</span><span class="p">,</span> <span class="n">func</span><span class="p">)</span>
<span class="k">def</span> <span class="nf">call</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">func</span><span class="p">,</span> <span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">):</span>
<span class="k">def</span> <span class="nf">func_wrapper</span><span class="p">():</span>
<span class="k">try</span><span class="p">:</span>
<span class="n">result</span> <span class="o">=</span> <span class="n">func</span><span class="p">(</span><span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">)</span>
<span class="k">except</span><span class="p">:</span>
<span class="n">func</span><span class="o">.</span><span class="n">on_failure</span><span class="p">(</span><span class="n">sys</span><span class="o">.</span><span class="n">exc_info</span><span class="p">())</span>
<span class="k">else</span><span class="p">:</span>
<span class="k">return</span> <span class="n">func</span><span class="o">.</span><span class="n">on_success</span><span class="p">(</span><span class="n">result</span><span class="p">)</span>
<span class="k">finally</span><span class="p">:</span>
<span class="n">func</span><span class="o">.</span><span class="n">on_closing</span><span class="p">()</span>
<span class="n">name</span> <span class="o">=</span> <span class="s">'</span><span class="si">%s</span><span class="s">-</span><span class="si">%s</span><span class="s">'</span> <span class="o">%</span> <span class="p">(</span><span class="n">func</span><span class="o">.</span><span class="n">__name__</span><span class="p">,</span> <span class="n">func</span><span class="o">.</span><span class="n">counter</span><span class="o">.</span><span class="n">next</span><span class="p">())</span>
<span class="n">thread</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">threadfactory</span><span class="p">(</span><span class="bp">None</span><span class="p">,</span> <span class="n">func_wrapper</span><span class="p">,</span> <span class="n">name</span><span class="p">)</span>
<span class="n">thread</span><span class="o">.</span><span class="n">start</span><span class="p">()</span>
<span class="k">return</span> <span class="n">thread</span>
</pre></div>
</div>
<p>The decorated function returns
the current execution thread, which can be stored and checked later, for
instance to verify that the thread <tt class="docutils literal"><span class="pre">.isAlive()</span></tt>.</p>
<p>Here is an example of usage. Suppose one wants to write some data to
an external resource which can be accessed by a single user at once
(for instance a printer). Then the access to the writing function must
be locked. Here is a minimalistic example:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="n">async</span> <span class="o">=</span> <span class="n">Async</span><span class="p">(</span><span class="n">threading</span><span class="o">.</span><span class="n">Thread</span><span class="p">)</span>
<span class="o">>>></span> <span class="n">datalist</span> <span class="o">=</span> <span class="p">[]</span> <span class="c"># for simplicity the written data are stored into a list.</span>
<span class="o">>>></span> <span class="nd">@async</span>
<span class="o">...</span> <span class="k">def</span> <span class="nf">write</span><span class="p">(</span><span class="n">data</span><span class="p">):</span>
<span class="o">...</span> <span class="c"># append data to the datalist by locking</span>
<span class="o">...</span> <span class="k">with</span> <span class="n">threading</span><span class="o">.</span><span class="n">Lock</span><span class="p">():</span>
<span class="o">...</span> <span class="n">time</span><span class="o">.</span><span class="n">sleep</span><span class="p">(</span><span class="mf">1</span><span class="p">)</span> <span class="c"># emulate some long running operation</span>
<span class="o">...</span> <span class="n">datalist</span><span class="o">.</span><span class="n">append</span><span class="p">(</span><span class="n">data</span><span class="p">)</span>
<span class="o">...</span> <span class="c"># other operations not requiring a lock here</span>
</pre></div>
</div>
<p>Each call to <tt class="docutils literal"><span class="pre">write</span></tt> will create a new writer thread, but there will
be no synchronization problems since <tt class="docutils literal"><span class="pre">write</span></tt> is locked.</p>
<pre class="doctest-block">
>>> write("data1")
<Thread(write-1, started...)>
</pre>
<pre class="doctest-block">
>>> time.sleep(.1) # wait a bit, so we are sure data2 is written after data1
</pre>
<pre class="doctest-block">
>>> write("data2")
<Thread(write-2, started...)>
</pre>
<pre class="doctest-block">
>>> time.sleep(2) # wait for the writers to complete
</pre>
<pre class="doctest-block">
>>> print datalist
['data1', 'data2']
</pre>
</div>
<div class="section" id="the-functionmaker-class">
<h1><a class="toc-backref" href="#id11">The <tt class="docutils literal"><span class="pre">FunctionMaker</span></tt> class</a></h1>
<p>You may wonder about how the functionality of the <tt class="docutils literal"><span class="pre">decorator</span></tt> module
is implemented. The basic building block is
a <tt class="docutils literal"><span class="pre">FunctionMaker</span></tt> class which is able to generate on the fly
functions with a given name and signature from a function template
passed as a string. Generally speaking, you should not need to
resort to <tt class="docutils literal"><span class="pre">FunctionMaker</span></tt> when writing ordinary decorators, but
it is handy in some circumstances. You will see an example shortly, in
the implementation of a cool decorator utility (<tt class="docutils literal"><span class="pre">decorator_apply</span></tt>).</p>
<p><tt class="docutils literal"><span class="pre">FunctionMaker</span></tt> provides a <tt class="docutils literal"><span class="pre">.create</span></tt> classmethod which
takes as input the name, signature, and body of the function
we want to generate as well as the execution environment
were the function is generated by <tt class="docutils literal"><span class="pre">exec</span></tt>. Here is an example:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="k">def</span> <span class="nf">f</span><span class="p">(</span><span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">):</span> <span class="c"># a function with a generic signature</span>
<span class="o">...</span> <span class="k">print</span> <span class="n">args</span><span class="p">,</span> <span class="n">kw</span>
<span class="o">>>></span> <span class="n">f1</span> <span class="o">=</span> <span class="n">FunctionMaker</span><span class="o">.</span><span class="n">create</span><span class="p">(</span><span class="s">'f1(a, b)'</span><span class="p">,</span> <span class="s">'f(a, b)'</span><span class="p">,</span> <span class="nb">dict</span><span class="p">(</span><span class="n">f</span><span class="o">=</span><span class="n">f</span><span class="p">))</span>
<span class="o">>>></span> <span class="n">f1</span><span class="p">(</span><span class="mf">1</span><span class="p">,</span><span class="mf">2</span><span class="p">)</span>
<span class="p">(</span><span class="mf">1</span><span class="p">,</span> <span class="mf">2</span><span class="p">)</span> <span class="p">{}</span>
</pre></div>
</div>
<p>It is important to notice that the function body is interpolated
before being executed, so be careful with the <tt class="docutils literal"><span class="pre">%</span></tt> sign!</p>
<p><tt class="docutils literal"><span class="pre">FunctionMaker.create</span></tt> also accepts keyword arguments and such
arguments are attached to the resulting function. This is useful
if you want to set some function attributes, for instance the
docstring <tt class="docutils literal"><span class="pre">__doc__</span></tt>.</p>
<p>For debugging/introspection purposes it may be useful to see
the source code of the generated function; to do that, just
pass the flag <tt class="docutils literal"><span class="pre">addsource=True</span></tt> and a <tt class="docutils literal"><span class="pre">__source__</span></tt> attribute will
be added to the generated function:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="n">f1</span> <span class="o">=</span> <span class="n">FunctionMaker</span><span class="o">.</span><span class="n">create</span><span class="p">(</span>
<span class="o">...</span> <span class="s">'f1(a, b)'</span><span class="p">,</span> <span class="s">'f(a, b)'</span><span class="p">,</span> <span class="nb">dict</span><span class="p">(</span><span class="n">f</span><span class="o">=</span><span class="n">f</span><span class="p">),</span> <span class="n">addsource</span><span class="o">=</span><span class="bp">True</span><span class="p">)</span>
<span class="o">>>></span> <span class="k">print</span> <span class="n">f1</span><span class="o">.</span><span class="n">__source__</span>
<span class="k">def</span> <span class="nf">f1</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">):</span>
<span class="n">f</span><span class="p">(</span><span class="n">a</span><span class="p">,</span> <span class="n">b</span><span class="p">)</span>
<span class="o"><</span><span class="n">BLANKLINE</span><span class="o">></span>
</pre></div>
</div>
<p><tt class="docutils literal"><span class="pre">FunctionMaker.create</span></tt> can take as first argument a string,
as in the examples before, or a function. This is the most common
usage, since typically you want to decorate a pre-existing
function. A framework author may want to use directly <tt class="docutils literal"><span class="pre">FunctionMaker.create</span></tt>
instead of <tt class="docutils literal"><span class="pre">decorator</span></tt>, since it gives you direct access to the body
of the generated function. For instance, suppose you want to instrument
the <tt class="docutils literal"><span class="pre">__init__</span></tt> methods of a set of classes, by preserving their
signature (such use case is not made up; this is done in SQAlchemy
and in other frameworks). When the first argument of <tt class="docutils literal"><span class="pre">FunctionMaker.create</span></tt>
is a function, a <tt class="docutils literal"><span class="pre">FunctionMaker</span></tt> object is instantiated internally,
with attributes <tt class="docutils literal"><span class="pre">args</span></tt>, <tt class="docutils literal"><span class="pre">varargs</span></tt>,
<tt class="docutils literal"><span class="pre">keywords</span></tt> and <tt class="docutils literal"><span class="pre">defaults</span></tt> which are the
the return values of the standard library function <tt class="docutils literal"><span class="pre">inspect.getargspec</span></tt>.
For each argument in the <tt class="docutils literal"><span class="pre">args</span></tt> (which is a list of strings containing
the names of the mandatory arguments) an attribute <tt class="docutils literal"><span class="pre">arg0</span></tt>, <tt class="docutils literal"><span class="pre">arg1</span></tt>,
..., <tt class="docutils literal"><span class="pre">argN</span></tt> is also generated. Finally, there is a <tt class="docutils literal"><span class="pre">signature</span></tt>
attribute, a string with the signature of the original function.</p>
<p>Notice that while I do not have plans
to change or remove the functionality provided in the
<tt class="docutils literal"><span class="pre">FunctionMaker</span></tt> class, I do not guarantee that it will stay
unchanged forever. For instance, right now I am using the traditional
string interpolation syntax for function templates, but Python 2.6
and Python 3.0 provide a newer interpolation syntax and I may use
the new syntax in the future.
On the other hand, the functionality provided by
<tt class="docutils literal"><span class="pre">decorator</span></tt> has been there from version 0.1 and it is guaranteed to
stay there forever.</p>
</div>
<div class="section" id="getting-the-source-code">
<h1><a class="toc-backref" href="#id12">Getting the source code</a></h1>
<p>Internally <tt class="docutils literal"><span class="pre">FunctionMaker.create</span></tt> uses <tt class="docutils literal"><span class="pre">exec</span></tt> to generate the
decorated function. Therefore
<tt class="docutils literal"><span class="pre">inspect.getsource</span></tt> will not work for decorated functions. That
means that the usual '??' trick in IPython will give you the (right on
the spot) message <tt class="docutils literal"><span class="pre">Dynamically</span> <span class="pre">generated</span> <span class="pre">function.</span> <span class="pre">No</span> <span class="pre">source</span> <span class="pre">code</span>
<span class="pre">available</span></tt>. In the past I have considered this acceptable, since
<tt class="docutils literal"><span class="pre">inspect.getsource</span></tt> does not really work even with regular
decorators. In that case <tt class="docutils literal"><span class="pre">inspect.getsource</span></tt> gives you the wrapper
source code which is probably not what you want:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="k">def</span> <span class="nf">identity_dec</span><span class="p">(</span><span class="n">func</span><span class="p">):</span>
<span class="k">def</span> <span class="nf">wrapper</span><span class="p">(</span><span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">):</span>
<span class="k">return</span> <span class="n">func</span><span class="p">(</span><span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">)</span>
<span class="k">return</span> <span class="n">wrapper</span>
</pre></div>
</div>
<div class="codeblock python">
<div class="highlight"><pre><span class="nd">@identity_dec</span>
<span class="k">def</span> <span class="nf">example</span><span class="p">():</span> <span class="k">pass</span>
<span class="o">>>></span> <span class="k">print</span> <span class="n">inspect</span><span class="o">.</span><span class="n">getsource</span><span class="p">(</span><span class="n">example</span><span class="p">)</span>
<span class="k">def</span> <span class="nf">wrapper</span><span class="p">(</span><span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">):</span>
<span class="k">return</span> <span class="n">func</span><span class="p">(</span><span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">)</span>
<span class="o"><</span><span class="n">BLANKLINE</span><span class="o">></span>
</pre></div>
</div>
<p>(see bug report <a class="reference external" href="http://bugs.python.org/issue1764286">1764286</a> for an explanation of what is happening).
Unfortunately the bug is still there, even in Python 2.6 and 3.0.
There is however a workaround. The decorator module adds an
attribute <tt class="docutils literal"><span class="pre">.undecorated</span></tt> to the decorated function, containing
a reference to the original function. The easy way to get
the source code is to call <tt class="docutils literal"><span class="pre">inspect.getsource</span></tt> on the
undecorated function:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="k">print</span> <span class="n">inspect</span><span class="o">.</span><span class="n">getsource</span><span class="p">(</span><span class="n">factorial</span><span class="o">.</span><span class="n">undecorated</span><span class="p">)</span>
<span class="nd">@tail_recursive</span>
<span class="k">def</span> <span class="nf">factorial</span><span class="p">(</span><span class="n">n</span><span class="p">,</span> <span class="n">acc</span><span class="o">=</span><span class="mf">1</span><span class="p">):</span>
<span class="s">"The good old factorial"</span>
<span class="k">if</span> <span class="n">n</span> <span class="o">==</span> <span class="mf">0</span><span class="p">:</span> <span class="k">return</span> <span class="n">acc</span>
<span class="k">return</span> <span class="n">factorial</span><span class="p">(</span><span class="n">n</span><span class="o">-</span><span class="mf">1</span><span class="p">,</span> <span class="n">n</span><span class="o">*</span><span class="n">acc</span><span class="p">)</span>
<span class="o"><</span><span class="n">BLANKLINE</span><span class="o">></span>
</pre></div>
</div>
</div>
<div class="section" id="dealing-with-third-party-decorators">
<h1><a class="toc-backref" href="#id13">Dealing with third party decorators</a></h1>
<p>Sometimes you find on the net some cool decorator that you would
like to include in your code. However, more often than not the cool
decorator is not signature-preserving. Therefore you may want an easy way to
upgrade third party decorators to signature-preserving decorators without
having to rewrite them in terms of <tt class="docutils literal"><span class="pre">decorator</span></tt>. You can use a
<tt class="docutils literal"><span class="pre">FunctionMaker</span></tt> to implement that functionality as follows:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="k">def</span> <span class="nf">decorator_apply</span><span class="p">(</span><span class="n">dec</span><span class="p">,</span> <span class="n">func</span><span class="p">):</span>
<span class="sd">"""</span>
<span class="sd"> Decorate a function by preserving the signature even if dec</span>
<span class="sd"> is not a signature-preserving decorator.</span>
<span class="sd"> """</span>
<span class="k">return</span> <span class="n">FunctionMaker</span><span class="o">.</span><span class="n">create</span><span class="p">(</span>
<span class="n">func</span><span class="p">,</span> <span class="s">'return decorated(</span><span class="si">%(signature)s</span><span class="s">)'</span><span class="p">,</span>
<span class="nb">dict</span><span class="p">(</span><span class="n">decorated</span><span class="o">=</span><span class="n">dec</span><span class="p">(</span><span class="n">func</span><span class="p">)),</span> <span class="n">undecorated</span><span class="o">=</span><span class="n">func</span><span class="p">)</span>
</pre></div>
</div>
<p><tt class="docutils literal"><span class="pre">decorator_apply</span></tt> sets the attribute <tt class="docutils literal"><span class="pre">.undecorated</span></tt> of the generated
function to the original function, so that you can get the right
source code.</p>
<p>Notice that I am not providing this functionality in the <tt class="docutils literal"><span class="pre">decorator</span></tt>
module directly since I think it is best to rewrite the decorator rather
than adding an additional level of indirection. However, practicality
beats purity, so you can add <tt class="docutils literal"><span class="pre">decorator_apply</span></tt> to your toolbox and
use it if you need to.</p>
<p>In order to give an example of usage of <tt class="docutils literal"><span class="pre">decorator_apply</span></tt>, I will show a
pretty slick decorator that converts a tail-recursive function in an iterative
function. I have shamelessly stolen the basic idea from Kay Schluehr's recipe
in the Python Cookbook,
<a class="reference external" href="http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/496691">http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/496691</a>.</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="k">class</span> <span class="nc">TailRecursive</span><span class="p">(</span><span class="nb">object</span><span class="p">):</span>
<span class="sd">"""</span>
<span class="sd"> tail_recursive decorator based on Kay Schluehr's recipe</span>
<span class="sd"> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/496691</span>
<span class="sd"> with improvements by me and George Sakkis.</span>
<span class="sd"> """</span>
<span class="k">def</span> <span class="nf">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">func</span><span class="p">):</span>
<span class="bp">self</span><span class="o">.</span><span class="n">func</span> <span class="o">=</span> <span class="n">func</span>
<span class="bp">self</span><span class="o">.</span><span class="n">firstcall</span> <span class="o">=</span> <span class="bp">True</span>
<span class="bp">self</span><span class="o">.</span><span class="n">CONTINUE</span> <span class="o">=</span> <span class="nb">object</span><span class="p">()</span> <span class="c"># sentinel</span>
<span class="k">def</span> <span class="nf">__call__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kwd</span><span class="p">):</span>
<span class="n">CONTINUE</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">CONTINUE</span>
<span class="k">if</span> <span class="bp">self</span><span class="o">.</span><span class="n">firstcall</span><span class="p">:</span>
<span class="n">func</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">func</span>
<span class="bp">self</span><span class="o">.</span><span class="n">firstcall</span> <span class="o">=</span> <span class="bp">False</span>
<span class="k">try</span><span class="p">:</span>
<span class="k">while</span> <span class="bp">True</span><span class="p">:</span>
<span class="n">result</span> <span class="o">=</span> <span class="n">func</span><span class="p">(</span><span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kwd</span><span class="p">)</span>
<span class="k">if</span> <span class="n">result</span> <span class="ow">is</span> <span class="n">CONTINUE</span><span class="p">:</span> <span class="c"># update arguments</span>
<span class="n">args</span><span class="p">,</span> <span class="n">kwd</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">argskwd</span>
<span class="k">else</span><span class="p">:</span> <span class="c"># last call</span>
<span class="k">return</span> <span class="n">result</span>
<span class="k">finally</span><span class="p">:</span>
<span class="bp">self</span><span class="o">.</span><span class="n">firstcall</span> <span class="o">=</span> <span class="bp">True</span>
<span class="k">else</span><span class="p">:</span> <span class="c"># return the arguments of the tail call</span>
<span class="bp">self</span><span class="o">.</span><span class="n">argskwd</span> <span class="o">=</span> <span class="n">args</span><span class="p">,</span> <span class="n">kwd</span>
<span class="k">return</span> <span class="n">CONTINUE</span>
</pre></div>
</div>
<p>Here the decorator is implemented as a class returning callable
objects.</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="k">def</span> <span class="nf">tail_recursive</span><span class="p">(</span><span class="n">func</span><span class="p">):</span>
<span class="k">return</span> <span class="n">decorator_apply</span><span class="p">(</span><span class="n">TailRecursive</span><span class="p">,</span> <span class="n">func</span><span class="p">)</span>
</pre></div>
</div>
<p>Here is how you apply the upgraded decorator to the good old factorial:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="nd">@tail_recursive</span>
<span class="k">def</span> <span class="nf">factorial</span><span class="p">(</span><span class="n">n</span><span class="p">,</span> <span class="n">acc</span><span class="o">=</span><span class="mf">1</span><span class="p">):</span>
<span class="s">"The good old factorial"</span>
<span class="k">if</span> <span class="n">n</span> <span class="o">==</span> <span class="mf">0</span><span class="p">:</span> <span class="k">return</span> <span class="n">acc</span>
<span class="k">return</span> <span class="n">factorial</span><span class="p">(</span><span class="n">n</span><span class="o">-</span><span class="mf">1</span><span class="p">,</span> <span class="n">n</span><span class="o">*</span><span class="n">acc</span><span class="p">)</span>
</pre></div>
</div>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="k">print</span> <span class="n">factorial</span><span class="p">(</span><span class="mf">4</span><span class="p">)</span>
<span class="mf">24</span>
</pre></div>
</div>
<p>This decorator is pretty impressive, and should give you some food for
your mind ;) Notice that there is no recursion limit now, and you can
easily compute <tt class="docutils literal"><span class="pre">factorial(1001)</span></tt> or larger without filling the stack
frame. Notice also that the decorator will not work on functions which
are not tail recursive, such as the following</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="k">def</span> <span class="nf">fact</span><span class="p">(</span><span class="n">n</span><span class="p">):</span> <span class="c"># this is not tail-recursive</span>
<span class="k">if</span> <span class="n">n</span> <span class="o">==</span> <span class="mf">0</span><span class="p">:</span> <span class="k">return</span> <span class="mf">1</span>
<span class="k">return</span> <span class="n">n</span> <span class="o">*</span> <span class="n">fact</span><span class="p">(</span><span class="n">n</span><span class="o">-</span><span class="mf">1</span><span class="p">)</span>
</pre></div>
</div>
<p>(reminder: a function is tail recursive if it either returns a value without
making a recursive call, or returns directly the result of a recursive
call).</p>
</div>
<div class="section" id="caveats-and-limitations">
<h1><a class="toc-backref" href="#id14">Caveats and limitations</a></h1>
<p>The first thing you should be aware of, it the fact that decorators
have a performance penalty.
The worse case is shown by the following example:</p>
<pre class="literal-block">
$ cat performance.sh
python -m timeit -s "
from decorator import decorator
@decorator
def do_nothing(func, *args, **kw):
return func(*args, **kw)
@do_nothing
def f():
pass
" "f()"
python -m timeit -s "
def f():
pass
" "f()"
</pre>
<p>On my MacBook, using the <tt class="docutils literal"><span class="pre">do_nothing</span></tt> decorator instead of the
plain function is more than three times slower:</p>
<pre class="literal-block">
$ bash performance.sh
1000000 loops, best of 3: 0.995 usec per loop
1000000 loops, best of 3: 0.273 usec per loop
</pre>
<p>It should be noted that a real life function would probably do
something more useful than <tt class="docutils literal"><span class="pre">f</span></tt> here, and therefore in real life the
performance penalty could be completely negligible. As always, the
only way to know if there is
a penalty in your specific use case is to measure it.</p>
<p>You should be aware that decorators will make your tracebacks
longer and more difficult to understand. Consider this example:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="nd">@trace</span>
<span class="o">...</span> <span class="k">def</span> <span class="nf">f</span><span class="p">():</span>
<span class="o">...</span> <span class="mf">1</span><span class="o">/</span><span class="mf">0</span>
</pre></div>
</div>
<p>Calling <tt class="docutils literal"><span class="pre">f()</span></tt> will give you a <tt class="docutils literal"><span class="pre">ZeroDivisionError</span></tt>, but since the
function is decorated the traceback will be longer:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="n">f</span><span class="p">()</span>
<span class="n">Traceback</span> <span class="p">(</span><span class="n">most</span> <span class="n">recent</span> <span class="n">call</span> <span class="n">last</span><span class="p">):</span>
<span class="o">...</span>
<span class="n">File</span> <span class="s">"<string>"</span><span class="p">,</span> <span class="n">line</span> <span class="mf">2</span><span class="p">,</span> <span class="ow">in</span> <span class="n">f</span>
<span class="n">File</span> <span class="s">"<doctest __main__[18]>"</span><span class="p">,</span> <span class="n">line</span> <span class="mf">4</span><span class="p">,</span> <span class="ow">in</span> <span class="n">trace</span>
<span class="k">return</span> <span class="n">f</span><span class="p">(</span><span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">)</span>
<span class="n">File</span> <span class="s">"<doctest __main__[47]>"</span><span class="p">,</span> <span class="n">line</span> <span class="mf">3</span><span class="p">,</span> <span class="ow">in</span> <span class="n">f</span>
<span class="mf">1</span><span class="o">/</span><span class="mf">0</span>
<span class="ne">ZeroDivisionError</span><span class="p">:</span> <span class="n">integer</span> <span class="n">division</span> <span class="ow">or</span> <span class="n">modulo</span> <span class="n">by</span> <span class="n">zero</span>
</pre></div>
</div>
<p>You see here the inner call to the decorator <tt class="docutils literal"><span class="pre">trace</span></tt>, which calls
<tt class="docutils literal"><span class="pre">f(*args,</span> <span class="pre">**kw)</span></tt>, and a reference to <tt class="docutils literal"><span class="pre">File</span> <span class="pre">"<string>",</span> <span class="pre">line</span> <span class="pre">2,</span> <span class="pre">in</span> <span class="pre">f</span></tt>.
This latter reference is due to the fact that internally the decorator
module uses <tt class="docutils literal"><span class="pre">exec</span></tt> to generate the decorated function. Notice that
<tt class="docutils literal"><span class="pre">exec</span></tt> is <em>not</em> responsibile for the performance penalty, since is the
called <em>only once</em> at function decoration time, and not every time
the decorated function is called.</p>
<p>At present, there is no clean way to avoid <tt class="docutils literal"><span class="pre">exec</span></tt>. A clean solution
would require to change the CPython implementation of functions and
add an hook to make it possible to change their signature directly.
That could happen in future versions of Python (see PEP <a class="reference external" href="http://www.python.org/dev/peps/pep-0362">362</a>) and
then the decorator module would become obsolete. However, at present,
even in Python 3.1 it is impossible to change the function signature
directly, therefore the <tt class="docutils literal"><span class="pre">decorator</span></tt> module is still useful.
Actually, this is one of the main reasons why I keep maintaining
the module and releasing new versions.</p>
<p>In the present implementation, decorators generated by <tt class="docutils literal"><span class="pre">decorator</span></tt>
can only be used on user-defined Python functions or methods, not on generic
callable objects, nor on built-in functions, due to limitations of the
<tt class="docutils literal"><span class="pre">inspect</span></tt> module in the standard library. Moreover, notice
that you can decorate a method, but only before if becomes a bound or unbound
method, i.e. inside the class.
Here is an example of valid decoration:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="k">class</span> <span class="nc">C</span><span class="p">(</span><span class="nb">object</span><span class="p">):</span>
<span class="o">...</span> <span class="nd">@trace</span>
<span class="o">...</span> <span class="k">def</span> <span class="nf">meth</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="o">...</span> <span class="k">pass</span>
</pre></div>
</div>
<p>Here is an example of invalid decoration, when the decorator in
called too late:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="k">class</span> <span class="nc">C</span><span class="p">(</span><span class="nb">object</span><span class="p">):</span>
<span class="o">...</span> <span class="k">def</span> <span class="nf">meth</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="o">...</span> <span class="k">pass</span>
<span class="o">...</span>
<span class="o">>>></span> <span class="n">trace</span><span class="p">(</span><span class="n">C</span><span class="o">.</span><span class="n">meth</span><span class="p">)</span>
<span class="n">Traceback</span> <span class="p">(</span><span class="n">most</span> <span class="n">recent</span> <span class="n">call</span> <span class="n">last</span><span class="p">):</span>
<span class="o">...</span>
<span class="ne">TypeError</span><span class="p">:</span> <span class="n">You</span> <span class="n">are</span> <span class="n">decorating</span> <span class="n">a</span> <span class="n">non</span> <span class="n">function</span><span class="p">:</span> <span class="o"><</span><span class="n">unbound</span> <span class="n">method</span> <span class="n">C</span><span class="o">.</span><span class="n">meth</span><span class="o">></span>
</pre></div>
</div>
<p>The solution is to extract the inner function from the unbound method:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="n">trace</span><span class="p">(</span><span class="n">C</span><span class="o">.</span><span class="n">meth</span><span class="o">.</span><span class="n">im_func</span><span class="p">)</span>
<span class="o"><</span><span class="n">function</span> <span class="n">meth</span> <span class="n">at</span> <span class="mf">0</span><span class="n">x</span><span class="o">...></span>
</pre></div>
</div>
<p>There is a restriction on the names of the arguments: for instance,
if try to call an argument <tt class="docutils literal"><span class="pre">_call_</span></tt> or <tt class="docutils literal"><span class="pre">_func_</span></tt>
you will get a <tt class="docutils literal"><span class="pre">NameError</span></tt>:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="nd">@trace</span>
<span class="o">...</span> <span class="k">def</span> <span class="nf">f</span><span class="p">(</span><span class="n">_func_</span><span class="p">):</span> <span class="k">print</span> <span class="n">f</span>
<span class="o">...</span>
<span class="n">Traceback</span> <span class="p">(</span><span class="n">most</span> <span class="n">recent</span> <span class="n">call</span> <span class="n">last</span><span class="p">):</span>
<span class="o">...</span>
<span class="ne">NameError</span><span class="p">:</span> <span class="n">_func_</span> <span class="ow">is</span> <span class="n">overridden</span> <span class="ow">in</span>
<span class="k">def</span> <span class="nf">f</span><span class="p">(</span><span class="n">_func_</span><span class="p">):</span>
<span class="k">return</span> <span class="n">_call_</span><span class="p">(</span><span class="n">_func_</span><span class="p">,</span> <span class="n">_func_</span><span class="p">)</span>
</pre></div>
</div>
<p>Finally, the implementation is such that the decorated function contains
a <em>copy</em> of the original function dictionary
(<tt class="docutils literal"><span class="pre">vars(decorated_f)</span> <span class="pre">is</span> <span class="pre">not</span> <span class="pre">vars(f)</span></tt>):</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="o">>>></span> <span class="k">def</span> <span class="nf">f</span><span class="p">():</span> <span class="k">pass</span> <span class="c"># the original function</span>
<span class="o">>>></span> <span class="n">f</span><span class="o">.</span><span class="n">attr1</span> <span class="o">=</span> <span class="s">"something"</span> <span class="c"># setting an attribute</span>
<span class="o">>>></span> <span class="n">f</span><span class="o">.</span><span class="n">attr2</span> <span class="o">=</span> <span class="s">"something else"</span> <span class="c"># setting another attribute</span>
<span class="o">>>></span> <span class="n">traced_f</span> <span class="o">=</span> <span class="n">trace</span><span class="p">(</span><span class="n">f</span><span class="p">)</span> <span class="c"># the decorated function</span>
<span class="o">>>></span> <span class="n">traced_f</span><span class="o">.</span><span class="n">attr1</span>
<span class="s">'something'</span>
<span class="o">>>></span> <span class="n">traced_f</span><span class="o">.</span><span class="n">attr2</span> <span class="o">=</span> <span class="s">"something different"</span> <span class="c"># setting attr</span>
<span class="o">>>></span> <span class="n">f</span><span class="o">.</span><span class="n">attr2</span> <span class="c"># the original attribute did not change</span>
<span class="s">'something else'</span>
</pre></div>
</div>
</div>
<div class="section" id="compatibility-notes">
<h1><a class="toc-backref" href="#id15">Compatibility notes</a></h1>
<p>Version 3.0 is a complete rewrite of the original implementation.
It is mostly compatible with the past, a part for a few differences.</p>
<p>First of all, the utilites <tt class="docutils literal"><span class="pre">get_info</span></tt> and <tt class="docutils literal"><span class="pre">new_wrapper</span></tt>, available
in the 2.X versions, have been deprecated and they will be removed
in the future. For the moment, using them raises a <tt class="docutils literal"><span class="pre">DeprecationWarning</span></tt>.
Incidentally, the functionality has been implemented through a
decorator which makes a good example for this documentation:</p>
<div class="codeblock python">
<div class="highlight"><pre><span class="nd">@decorator</span>
<span class="k">def</span> <span class="nf">deprecated</span><span class="p">(</span><span class="n">func</span><span class="p">,</span> <span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">):</span>
<span class="s">"A decorator for deprecated functions"</span>
<span class="n">warnings</span><span class="o">.</span><span class="n">warn</span><span class="p">(</span>
<span class="p">(</span><span class="s">'Calling the deprecated function </span><span class="si">%r</span><span class="se">\n</span><span class="s">'</span>
<span class="s">'Downgrade to decorator 2.3 if you want to use this functionality'</span><span class="p">)</span>
<span class="o">%</span> <span class="n">func</span><span class="o">.</span><span class="n">__name__</span><span class="p">,</span> <span class="ne">DeprecationWarning</span><span class="p">,</span> <span class="n">stacklevel</span><span class="o">=</span><span class="mf">3</span><span class="p">)</span>
<span class="k">return</span> <span class="n">func</span><span class="p">(</span><span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kw</span><span class="p">)</span>
</pre></div>
</div>
<p><tt class="docutils literal"><span class="pre">get_info</span></tt> has been removed since it was little used and since it had
to be changed anyway to work with Python 3.0; <tt class="docutils literal"><span class="pre">new_wrapper</span></tt> has been
removed since it was useless: its major use case (converting
signature changing decorators to signature preserving decorators)
has been subsumed by <tt class="docutils literal"><span class="pre">decorator_apply</span></tt>
and the other use case can be managed with the <tt class="docutils literal"><span class="pre">FunctionMaker</span></tt>.</p>
<p>Finally <tt class="docutils literal"><span class="pre">decorator</span></tt> cannot be used as a class decorator and the
<a class="reference external" href="http://www.phyast.pitt.edu/~micheles/python/documentation.html#class-decorators-and-decorator-factories">functionality introduced in version 2.3</a> has been removed. That
means that in order to define decorator factories with classes you
need to define the <tt class="docutils literal"><span class="pre">__call__</span></tt> method explicitly (no magic anymore).
Since there is
an easy way to define decorator factories by using <tt class="docutils literal"><span class="pre">decorator_factory</span></tt>,
there is less need to use classes to implement decorator factories.</p>
<p>All these changes should not cause any trouble, since they were
all rarely used features. Should you have any trouble, you can always
downgrade to the 2.3 version.</p>
<p>The examples shown here have been tested with Python 2.5. Python 2.4
is also supported - of course the examples requiring the <tt class="docutils literal"><span class="pre">with</span></tt>
statement will not work there. Python 2.6 works fine, but if you
run the examples here in the interactive interpreter
you will notice a few differences since
<tt class="docutils literal"><span class="pre">getargspec</span></tt> returns an <tt class="docutils literal"><span class="pre">ArgSpec</span></tt> namedtuple instead of a regular
tuple. That means that running the file
<tt class="docutils literal"><span class="pre">documentation.py</span></tt> under Python 2.5 will a few errors, but
they are not serious. Python 3.0 is kind of supported too.
Simply run the script <tt class="docutils literal"><span class="pre">2to3</span></tt> on the module
<tt class="docutils literal"><span class="pre">decorator.py</span></tt> and you will get a version of the code running
with Python 3.0 (at least, I did some simple checks and it seemed
to work). However there is no support for <a class="reference external" href="http://www.python.org/dev/peps/pep-3107/">function annotations</a> yet
since it seems premature at this moment (most people are
still using Python 2.5).</p>
</div>
<div class="section" id="licence">
<h1><a class="toc-backref" href="#id16">LICENCE</a></h1>
<p>Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:</p>
<pre class="literal-block">
Copyright (c) 2005, Michele Simionato
All rights reserved.
Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
Redistributions in bytecode form must reproduce the above copyright
notice, this list of conditions and the following disclaimer in
the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
HOLDERS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS
OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR
TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH
DAMAGE.
</pre>
<p>If you use this software and you are happy with it, consider sending me a
note, just to gratify my ego. On the other hand, if you use this software and
you are unhappy with it, send me a patch!</p>
</div>
</div>
<div class="footer">
<hr class="footer" />
<a class="reference external" href="documentation.rst">View document source</a>.
Generated on: 2009-08-25 12:37 UTC.
Generated by <a class="reference external" href="http://docutils.sourceforge.net/">Docutils</a> from <a class="reference external" href="http://docutils.sourceforge.net/rst.html">reStructuredText</a> source.
</div>
</body>
</html>
distrib
>
Mandriva
>
2010.0
>
i586
>
media
>
contrib-release
>
by-pkgid
>
1fed1d33b2e620178c1ca30f32e02d10
>
files
>
7
