Admin Control Panel v2.0

✎ Editing: extending.html ← Back

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

<html xmlns="http://www.w3.org/1999/xhtml">
 <head>
 <meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
 
 <title>1. Extending Python with C or C++ &mdash; Python 2.7.5 documentation</title>
 
 <link rel="stylesheet" href="../_static/default.css" type="text/css" />
 <link rel="stylesheet" href="../_static/pygments.css" type="text/css" />
 
 <script type="text/javascript">
 var DOCUMENTATION_OPTIONS = {
 URL_ROOT: '../',
 VERSION: '2.7.5',
 COLLAPSE_INDEX: false,
 FILE_SUFFIX: '.html',
 HAS_SOURCE: true
 };
 </script>
 <script type="text/javascript" src="../_static/jquery.js"></script>
 <script type="text/javascript" src="../_static/underscore.js"></script>
 <script type="text/javascript" src="../_static/doctools.js"></script>
 <script type="text/javascript" src="../_static/sidebar.js"></script>
 <link rel="search" type="application/opensearchdescription+xml"
 title="Search within Python 2.7.5 documentation"
 href="../_static/opensearch.xml"/>
 <link rel="author" title="About these documents" href="../about.html" />
 <link rel="copyright" title="Copyright" href="../copyright.html" />
 <link rel="top" title="Python 2.7.5 documentation" href="../index.html" />
 <link rel="up" title="Extending and Embedding the Python Interpreter" href="index.html" />
 <link rel="next" title="2. Defining New Types" href="newtypes.html" />
 <link rel="prev" title="Extending and Embedding the Python Interpreter" href="index.html" />
 <link rel="shortcut icon" type="image/png" href="../_static/py.png" />
 <script type="text/javascript" src="../_static/copybutton.js"></script>

</head>
 <body>
 <div class="related">
 <h3>Navigation</h3>
 <ul>
 <li class="right" style="margin-right: 10px">
 <a href="../genindex.html" title="General Index"
 accesskey="I">index</a></li>
 <li class="right" >
 <a href="../py-modindex.html" title="Python Module Index"
 >modules</a> |</li>
 <li class="right" >
 <a href="newtypes.html" title="2. Defining New Types"
 accesskey="N">next</a> |</li>
 <li class="right" >
 <a href="index.html" title="Extending and Embedding the Python Interpreter"
 accesskey="P">previous</a> |</li>
 <li><img src="../_static/py.png" alt=""
 style="vertical-align: middle; margin-top: -1px"/></li>
 <li><a href="http://www.python.org/">Python</a> &raquo;</li>
 <li>
 <a href="../index.html">Python 2.7.5 documentation</a> &raquo;
 </li>

<li><a href="index.html" accesskey="U">Extending and Embedding the Python Interpreter</a> &raquo;</li> 
 </ul>
 </div>

<div class="document">
 <div class="documentwrapper">
 <div class="bodywrapper">
 <div class="body">
 
 <div class="section" id="extending-python-with-c-or-c">
<h1>1. Extending Python with C or C++<a class="headerlink" href="#extending-python-with-c-or-c" title="Permalink to this headline">¶</a></h1>
It is quite easy to add new built-in modules to Python, if you know how to
program in C. Such extension modules can do two things that can&#8217;t be
done directly in Python: they can implement new built-in object types, and they
can call C library functions and system calls.
To support extensions, the Python API (Application Programmers Interface)
defines a set of functions, macros and variables that provide access to most
aspects of the Python run-time system. The Python API is incorporated in a C
source file by including the header <tt class="docutils literal">&quot;Python.h&quot;</tt>.
The compilation of an extension module depends on its intended use as well as on
your system setup; details are given in later chapters.
Do note that if your use case is calling C library functions or system calls,
you should consider using the <a class="reference internal" href="../library/ctypes.html#module-ctypes" title="ctypes: A foreign function library for Python."><tt class="xref py py-mod docutils literal">ctypes</tt></a> module rather than writing custom
C code. Not only does <a class="reference internal" href="../library/ctypes.html#module-ctypes" title="ctypes: A foreign function library for Python."><tt class="xref py py-mod docutils literal">ctypes</tt></a> let you write Python code to interface
with C code, but it is more portable between implementations of Python than
writing and compiling an extension module which typically ties you to CPython.
<div class="section" id="a-simple-example">
<h2>1.1. A Simple Example<a class="headerlink" href="#a-simple-example" title="Permalink to this headline">¶</a></h2>
Let&#8217;s create an extension module called <tt class="docutils literal">spam</tt> (the favorite food of Monty
Python fans...) and let&#8217;s say we want to create a Python interface to the C
library function <tt class="xref c c-func docutils literal">system()</tt>. <a class="footnote-reference" href="#id5" id="id1">[1]</a> This function takes a null-terminated
character string as argument and returns an integer. We want this function to
be callable from Python as follows:
<div class="highlight-c"><div class="highlight"><pre>&gt;&gt;&gt; import spam
&gt;&gt;&gt; status = spam.system(&quot;ls -l&quot;)
</pre></div>
</div>
Begin by creating a file <tt class="file docutils literal">spammodule.c</tt>. (Historically, if a module is
called <tt class="docutils literal">spam</tt>, the C file containing its implementation is called
<tt class="file docutils literal">spammodule.c</tt>; if the module name is very long, like <tt class="docutils literal">spammify</tt>, the
module name can be just <tt class="file docutils literal">spammify.c</tt>.)
The first line of our file can be:
<div class="highlight-c"><div class="highlight"><pre>#include &lt;Python.h&gt;
</pre></div>
</div>
which pulls in the Python API (you can add a comment describing the purpose of
the module and a copyright notice if you like).
<div class="admonition note">
Note
Since Python may define some pre-processor definitions which affect the standard
headers on some systems, you must include <tt class="file docutils literal">Python.h</tt> before any standard
headers are included.
</div>
All user-visible symbols defined by <tt class="file docutils literal">Python.h</tt> have a prefix of <tt class="docutils literal">Py</tt> or
<tt class="docutils literal">PY</tt>, except those defined in standard header files. For convenience, and
since they are used extensively by the Python interpreter, <tt class="docutils literal">&quot;Python.h&quot;</tt>
includes a few standard header files: <tt class="docutils literal">&lt;stdio.h&gt;</tt>, <tt class="docutils literal">&lt;string.h&gt;</tt>,
<tt class="docutils literal">&lt;errno.h&gt;</tt>, and <tt class="docutils literal">&lt;stdlib.h&gt;</tt>. If the latter header file does not exist on
your system, it declares the functions <tt class="xref c c-func docutils literal">malloc()</tt>, <tt class="xref c c-func docutils literal">free()</tt> and
<tt class="xref c c-func docutils literal">realloc()</tt> directly.
The next thing we add to our module file is the C function that will be called
when the Python expression <tt class="docutils literal">spam.system(string)</tt> is evaluated (we&#8217;ll see
shortly how it ends up being called):
<div class="highlight-c"><div class="highlight"><pre>static PyObject *
spam_system(PyObject *self, PyObject *args)
{
 const char *command;
 int sts;

if (!PyArg_ParseTuple(args, &quot;s&quot;, &amp;command))
 return NULL;
 sts = system(command);
 return Py_BuildValue(&quot;i&quot;, sts);
}
</pre></div>
</div>
There is a straightforward translation from the argument list in Python (for
example, the single expression <tt class="docutils literal">&quot;ls -l&quot;</tt>) to the arguments passed to the C
function. The C function always has two arguments, conventionally named self
and args.
The self argument points to the module object for module-level functions;
for a method it would point to the object instance.
The args argument will be a pointer to a Python tuple object containing the
arguments. Each item of the tuple corresponds to an argument in the call&#8217;s
argument list. The arguments are Python objects &#8212; in order to do anything
with them in our C function we have to convert them to C values. The function
<a class="reference internal" href="../c-api/arg.html#PyArg_ParseTuple" title="PyArg_ParseTuple"><tt class="xref c c-func docutils literal">PyArg_ParseTuple()</tt></a> in the Python API checks the argument types and
converts them to C values. It uses a template string to determine the required
types of the arguments as well as the types of the C variables into which to
store the converted values. More about this later.
<a class="reference internal" href="../c-api/arg.html#PyArg_ParseTuple" title="PyArg_ParseTuple"><tt class="xref c c-func docutils literal">PyArg_ParseTuple()</tt></a> returns true (nonzero) if all arguments have the right
type and its components have been stored in the variables whose addresses are
passed. It returns false (zero) if an invalid argument list was passed. In the
latter case it also raises an appropriate exception so the calling function can
return NULL immediately (as we saw in the example).
</div>
<div class="section" id="intermezzo-errors-and-exceptions">
<h2>1.2. Intermezzo: Errors and Exceptions<a class="headerlink" href="#intermezzo-errors-and-exceptions" title="Permalink to this headline">¶</a></h2>
An important convention throughout the Python interpreter is the following: when
a function fails, it should set an exception condition and return an error value
(usually a NULL pointer). Exceptions are stored in a static global variable
inside the interpreter; if this variable is NULL no exception has occurred. A
second global variable stores the &#8220;associated value&#8221; of the exception (the
second argument to <a class="reference internal" href="../reference/simple_stmts.html#raise"><tt class="xref std std-keyword docutils literal">raise</tt></a>). A third variable contains the stack
traceback in case the error originated in Python code. These three variables
are the C equivalents of the Python variables <tt class="docutils literal">sys.exc_type</tt>,
<tt class="docutils literal">sys.exc_value</tt> and <tt class="docutils literal">sys.exc_traceback</tt> (see the section on module
<a class="reference internal" href="../library/sys.html#module-sys" title="sys: Access system-specific parameters and functions."><tt class="xref py py-mod docutils literal">sys</tt></a> in the Python Library Reference). It is important to know about them
to understand how errors are passed around.
The Python API defines a number of functions to set various types of exceptions.
The most common one is <a class="reference internal" href="../c-api/exceptions.html#PyErr_SetString" title="PyErr_SetString"><tt class="xref c c-func docutils literal">PyErr_SetString()</tt></a>. Its arguments are an exception
object and a C string. The exception object is usually a predefined object like
<tt class="xref c c-data docutils literal">PyExc_ZeroDivisionError</tt>. The C string indicates the cause of the error
and is converted to a Python string object and stored as the &#8220;associated value&#8221;
of the exception.
Another useful function is <a class="reference internal" href="../c-api/exceptions.html#PyErr_SetFromErrno" title="PyErr_SetFromErrno"><tt class="xref c c-func docutils literal">PyErr_SetFromErrno()</tt></a>, which only takes an
exception argument and constructs the associated value by inspection of the
global variable <tt class="xref c c-data docutils literal">errno</tt>. The most general function is
<a class="reference internal" href="../c-api/exceptions.html#PyErr_SetObject" title="PyErr_SetObject"><tt class="xref c c-func docutils literal">PyErr_SetObject()</tt></a>, which takes two object arguments, the exception and
its associated value. You don&#8217;t need to <a class="reference internal" href="../c-api/refcounting.html#Py_INCREF" title="Py_INCREF"><tt class="xref c c-func docutils literal">Py_INCREF()</tt></a> the objects passed
to any of these functions.
You can test non-destructively whether an exception has been set with
<a class="reference internal" href="../c-api/exceptions.html#PyErr_Occurred" title="PyErr_Occurred"><tt class="xref c c-func docutils literal">PyErr_Occurred()</tt></a>. This returns the current exception object, or NULL
if no exception has occurred. You normally don&#8217;t need to call
<a class="reference internal" href="../c-api/exceptions.html#PyErr_Occurred" title="PyErr_Occurred"><tt class="xref c c-func docutils literal">PyErr_Occurred()</tt></a> to see whether an error occurred in a function call,
since you should be able to tell from the return value.
When a function f that calls another function g detects that the latter
fails, f should itself return an error value (usually NULL or <tt class="docutils literal">-1</tt>). It
should not call one of the <tt class="xref c c-func docutils literal">PyErr_*()</tt> functions &#8212; one has already
been called by g. f&#8216;s caller is then supposed to also return an error
indication to its caller, again without calling <tt class="xref c c-func docutils literal">PyErr_*()</tt>, and so on
&#8212; the most detailed cause of the error was already reported by the function
that first detected it. Once the error reaches the Python interpreter&#8217;s main
loop, this aborts the currently executing Python code and tries to find an
exception handler specified by the Python programmer.
(There are situations where a module can actually give a more detailed error
message by calling another <tt class="xref c c-func docutils literal">PyErr_*()</tt> function, and in such cases it is
fine to do so. As a general rule, however, this is not necessary, and can cause
information about the cause of the error to be lost: most operations can fail
for a variety of reasons.)
To ignore an exception set by a function call that failed, the exception
condition must be cleared explicitly by calling <a class="reference internal" href="../c-api/exceptions.html#PyErr_Clear" title="PyErr_Clear"><tt class="xref c c-func docutils literal">PyErr_Clear()</tt></a>. The only
time C code should call <a class="reference internal" href="../c-api/exceptions.html#PyErr_Clear" title="PyErr_Clear"><tt class="xref c c-func docutils literal">PyErr_Clear()</tt></a> is if it doesn&#8217;t want to pass the
error on to the interpreter but wants to handle it completely by itself
(possibly by trying something else, or pretending nothing went wrong).
Every failing <tt class="xref c c-func docutils literal">malloc()</tt> call must be turned into an exception &#8212; the
direct caller of <tt class="xref c c-func docutils literal">malloc()</tt> (or <tt class="xref c c-func docutils literal">realloc()</tt>) must call
<a class="reference internal" href="../c-api/exceptions.html#PyErr_NoMemory" title="PyErr_NoMemory"><tt class="xref c c-func docutils literal">PyErr_NoMemory()</tt></a> and return a failure indicator itself. All the
object-creating functions (for example, <a class="reference internal" href="../c-api/int.html#PyInt_FromLong" title="PyInt_FromLong"><tt class="xref c c-func docutils literal">PyInt_FromLong()</tt></a>) already do
this, so this note is only relevant to those who call <tt class="xref c c-func docutils literal">malloc()</tt> directly.
Also note that, with the important exception of <a class="reference internal" href="../c-api/arg.html#PyArg_ParseTuple" title="PyArg_ParseTuple"><tt class="xref c c-func docutils literal">PyArg_ParseTuple()</tt></a> and
friends, functions that return an integer status usually return a positive value
or zero for success and <tt class="docutils literal">-1</tt> for failure, like Unix system calls.
Finally, be careful to clean up garbage (by making <a class="reference internal" href="../c-api/refcounting.html#Py_XDECREF" title="Py_XDECREF"><tt class="xref c c-func docutils literal">Py_XDECREF()</tt></a> or
<a class="reference internal" href="../c-api/refcounting.html#Py_DECREF" title="Py_DECREF"><tt class="xref c c-func docutils literal">Py_DECREF()</tt></a> calls for objects you have already created) when you return
an error indicator!
The choice of which exception to raise is entirely yours. There are predeclared
C objects corresponding to all built-in Python exceptions, such as
<tt class="xref c c-data docutils literal">PyExc_ZeroDivisionError</tt>, which you can use directly. Of course, you
should choose exceptions wisely &#8212; don&#8217;t use <tt class="xref c c-data docutils literal">PyExc_TypeError</tt> to mean
that a file couldn&#8217;t be opened (that should probably be <tt class="xref c c-data docutils literal">PyExc_IOError</tt>).
If something&#8217;s wrong with the argument list, the <a class="reference internal" href="../c-api/arg.html#PyArg_ParseTuple" title="PyArg_ParseTuple"><tt class="xref c c-func docutils literal">PyArg_ParseTuple()</tt></a>
function usually raises <tt class="xref c c-data docutils literal">PyExc_TypeError</tt>. If you have an argument whose
value must be in a particular range or must satisfy other conditions,
<tt class="xref c c-data docutils literal">PyExc_ValueError</tt> is appropriate.
You can also define a new exception that is unique to your module. For this, you
usually declare a static object variable at the beginning of your file:
<div class="highlight-c"><div class="highlight"><pre>static PyObject *SpamError;
</pre></div>
</div>
and initialize it in your module&#8217;s initialization function (<tt class="xref c c-func docutils literal">initspam()</tt>)
with an exception object (leaving out the error checking for now):
<div class="highlight-c"><div class="highlight"><pre>PyMODINIT_FUNC
initspam(void)
{
 PyObject *m;

m = Py_InitModule(&quot;spam&quot;, SpamMethods);
 if (m == NULL)
 return;

SpamError = PyErr_NewException(&quot;spam.error&quot;, NULL, NULL);
 Py_INCREF(SpamError);
 PyModule_AddObject(m, &quot;error&quot;, SpamError);
}
</pre></div>
</div>
Note that the Python name for the exception object is <tt class="xref py py-exc docutils literal">spam.error</tt>. The
<a class="reference internal" href="../c-api/exceptions.html#PyErr_NewException" title="PyErr_NewException"><tt class="xref c c-func docutils literal">PyErr_NewException()</tt></a> function may create a class with the base class
being <a class="reference internal" href="../library/exceptions.html#exceptions.Exception" title="exceptions.Exception"><tt class="xref py py-exc docutils literal">Exception</tt></a> (unless another class is passed in instead of NULL),
described in <a class="reference internal" href="../library/exceptions.html#bltin-exceptions">Built-in Exceptions</a>.
Note also that the <tt class="xref c c-data docutils literal">SpamError</tt> variable retains a reference to the newly
created exception class; this is intentional! Since the exception could be
removed from the module by external code, an owned reference to the class is
needed to ensure that it will not be discarded, causing <tt class="xref c c-data docutils literal">SpamError</tt> to
become a dangling pointer. Should it become a dangling pointer, C code which
raises the exception could cause a core dump or other unintended side effects.
We discuss the use of <tt class="docutils literal">PyMODINIT_FUNC</tt> as a function return type later in this
sample.
The <tt class="xref py py-exc docutils literal">spam.error</tt> exception can be raised in your extension module using a
call to <a class="reference internal" href="../c-api/exceptions.html#PyErr_SetString" title="PyErr_SetString"><tt class="xref c c-func docutils literal">PyErr_SetString()</tt></a> as shown below:
<div class="highlight-c"><div class="highlight"><pre>static PyObject *
spam_system(PyObject *self, PyObject *args)
{
 const char *command;
 int sts;

if (!PyArg_ParseTuple(args, &quot;s&quot;, &amp;command))
 return NULL;
 sts = system(command);
 if (sts &lt; 0) {
 PyErr_SetString(SpamError, &quot;System command failed&quot;);
 return NULL;
 }
 return PyLong_FromLong(sts);
}
</pre></div>
</div>
</div>
<div class="section" id="back-to-the-example">
<h2>1.3. Back to the Example<a class="headerlink" href="#back-to-the-example" title="Permalink to this headline">¶</a></h2>
Going back to our example function, you should now be able to understand this
statement:
<div class="highlight-c"><div class="highlight"><pre>if (!PyArg_ParseTuple(args, &quot;s&quot;, &amp;command))
 return NULL;
</pre></div>
</div>
It returns NULL (the error indicator for functions returning object pointers)
if an error is detected in the argument list, relying on the exception set by
<a class="reference internal" href="../c-api/arg.html#PyArg_ParseTuple" title="PyArg_ParseTuple"><tt class="xref c c-func docutils literal">PyArg_ParseTuple()</tt></a>. Otherwise the string value of the argument has been
copied to the local variable <tt class="xref c c-data docutils literal">command</tt>. This is a pointer assignment and
you are not supposed to modify the string to which it points (so in Standard C,
the variable <tt class="xref c c-data docutils literal">command</tt> should properly be declared as <tt class="docutils literal">const char
*command</tt>).
The next statement is a call to the Unix function <tt class="xref c c-func docutils literal">system()</tt>, passing it
the string we just got from <a class="reference internal" href="../c-api/arg.html#PyArg_ParseTuple" title="PyArg_ParseTuple"><tt class="xref c c-func docutils literal">PyArg_ParseTuple()</tt></a>:
<div class="highlight-c"><div class="highlight"><pre>sts = system(command);
</pre></div>
</div>
Our <tt class="xref py py-func docutils literal">spam.system()</tt> function must return the value of <tt class="xref c c-data docutils literal">sts</tt> as a
Python object. This is done using the function <a class="reference internal" href="../c-api/arg.html#Py_BuildValue" title="Py_BuildValue"><tt class="xref c c-func docutils literal">Py_BuildValue()</tt></a>, which is
something like the inverse of <a class="reference internal" href="../c-api/arg.html#PyArg_ParseTuple" title="PyArg_ParseTuple"><tt class="xref c c-func docutils literal">PyArg_ParseTuple()</tt></a>: it takes a format
string and an arbitrary number of C values, and returns a new Python object.
More info on <a class="reference internal" href="../c-api/arg.html#Py_BuildValue" title="Py_BuildValue"><tt class="xref c c-func docutils literal">Py_BuildValue()</tt></a> is given later.
<div class="highlight-c"><div class="highlight"><pre>return Py_BuildValue(&quot;i&quot;, sts);
</pre></div>
</div>
In this case, it will return an integer object. (Yes, even integers are objects
on the heap in Python!)
If you have a C function that returns no useful argument (a function returning
<tt class="xref c c-type docutils literal">void</tt>), the corresponding Python function must return <tt class="docutils literal">None</tt>. You
need this idiom to do so (which is implemented by the <a class="reference internal" href="../c-api/none.html#Py_RETURN_NONE" title="Py_RETURN_NONE"><tt class="xref c c-macro docutils literal">Py_RETURN_NONE</tt></a>
macro):
<div class="highlight-c"><div class="highlight"><pre>Py_INCREF(Py_None);
return Py_None;
</pre></div>
</div>
<a class="reference internal" href="../c-api/none.html#Py_None" title="Py_None"><tt class="xref c c-data docutils literal">Py_None</tt></a> is the C name for the special Python object <tt class="docutils literal">None</tt>. It is a
genuine Python object rather than a NULL pointer, which means &#8220;error&#8221; in most
contexts, as we have seen.
</div>
<div class="section" id="the-module-s-method-table-and-initialization-function">
<h2>1.4. The Module&#8217;s Method Table and Initialization Function<a class="headerlink" href="#the-module-s-method-table-and-initialization-function" title="Permalink to this headline">¶</a></h2>
I promised to show how <tt class="xref c c-func docutils literal">spam_system()</tt> is called from Python programs.
First, we need to list its name and address in a &#8220;method table&#8221;:
<div class="highlight-c"><div class="highlight"><pre>static PyMethodDef SpamMethods[] = {
 ...
 {&quot;system&quot;, spam_system, METH_VARARGS,
 &quot;Execute a shell command.&quot;},
 ...
 {NULL, NULL, 0, NULL} /* Sentinel */
};
</pre></div>
</div>
Note the third entry (<tt class="docutils literal">METH_VARARGS</tt>). This is a flag telling the interpreter
the calling convention to be used for the C function. It should normally always
be <tt class="docutils literal">METH_VARARGS</tt> or <tt class="docutils literal">METH_VARARGS | METH_KEYWORDS</tt>; a value of <tt class="docutils literal">0</tt> means
that an obsolete variant of <a class="reference internal" href="../c-api/arg.html#PyArg_ParseTuple" title="PyArg_ParseTuple"><tt class="xref c c-func docutils literal">PyArg_ParseTuple()</tt></a> is used.
When using only <tt class="docutils literal">METH_VARARGS</tt>, the function should expect the Python-level
parameters to be passed in as a tuple acceptable for parsing via
<a class="reference internal" href="../c-api/arg.html#PyArg_ParseTuple" title="PyArg_ParseTuple"><tt class="xref c c-func docutils literal">PyArg_ParseTuple()</tt></a>; more information on this function is provided below.
The <a class="reference internal" href="../c-api/structures.html#METH_KEYWORDS" title="METH_KEYWORDS"><tt class="xref py py-const docutils literal">METH_KEYWORDS</tt></a> bit may be set in the third field if keyword
arguments should be passed to the function. In this case, the C function should
accept a third <tt class="docutils literal">PyObject *</tt> parameter which will be a dictionary of keywords.
Use <a class="reference internal" href="../c-api/arg.html#PyArg_ParseTupleAndKeywords" title="PyArg_ParseTupleAndKeywords"><tt class="xref c c-func docutils literal">PyArg_ParseTupleAndKeywords()</tt></a> to parse the arguments to such a
function.
The method table must be passed to the interpreter in the module&#8217;s
initialization function. The initialization function must be named
<tt class="xref c c-func docutils literal">initname()</tt>, where name is the name of the module, and should be the
only non-<tt class="docutils literal">static</tt> item defined in the module file:
<div class="highlight-c"><div class="highlight"><pre>PyMODINIT_FUNC
initspam(void)
{
 (void) Py_InitModule(&quot;spam&quot;, SpamMethods);
}
</pre></div>
</div>
Note that PyMODINIT_FUNC declares the function as <tt class="docutils literal">void</tt> return type,
declares any special linkage declarations required by the platform, and for C++
declares the function as <tt class="docutils literal">extern &quot;C&quot;</tt>.
When the Python program imports module <tt class="xref py py-mod docutils literal">spam</tt> for the first time,
<tt class="xref c c-func docutils literal">initspam()</tt> is called. (See below for comments about embedding Python.)
It calls <a class="reference internal" href="../c-api/allocation.html#Py_InitModule" title="Py_InitModule"><tt class="xref c c-func docutils literal">Py_InitModule()</tt></a>, which creates a &#8220;module object&#8221; (which is
inserted in the dictionary <tt class="docutils literal">sys.modules</tt> under the key <tt class="docutils literal">&quot;spam&quot;</tt>), and
inserts built-in function objects into the newly created module based upon the
table (an array of <a class="reference internal" href="../c-api/structures.html#PyMethodDef" title="PyMethodDef"><tt class="xref c c-type docutils literal">PyMethodDef</tt></a> structures) that was passed as its
second argument. <a class="reference internal" href="../c-api/allocation.html#Py_InitModule" title="Py_InitModule"><tt class="xref c c-func docutils literal">Py_InitModule()</tt></a> returns a pointer to the module object
that it creates (which is unused here). It may abort with a fatal error for
certain errors, or return NULL if the module could not be initialized
satisfactorily.
When embedding Python, the <tt class="xref c c-func docutils literal">initspam()</tt> function is not called
automatically unless there&#8217;s an entry in the <tt class="xref c c-data docutils literal">_PyImport_Inittab</tt> table.
The easiest way to handle this is to statically initialize your
statically-linked modules by directly calling <tt class="xref c c-func docutils literal">initspam()</tt> after the call
to <a class="reference internal" href="../c-api/init.html#Py_Initialize" title="Py_Initialize"><tt class="xref c c-func docutils literal">Py_Initialize()</tt></a>:
<div class="highlight-c"><div class="highlight"><pre>int
main(int argc, char *argv[])
{
 /* Pass argv[0] to the Python interpreter */
 Py_SetProgramName(argv[0]);

/* Initialize the Python interpreter. Required. */
 Py_Initialize();

/* Add a static module */
 initspam();
</pre></div>
</div>
An example may be found in the file <tt class="file docutils literal">Demo/embed/demo.c</tt> in the Python
source distribution.
<div class="admonition note">
Note
Removing entries from <tt class="docutils literal">sys.modules</tt> or importing compiled modules into
multiple interpreters within a process (or following a <tt class="xref c c-func docutils literal">fork()</tt> without an
intervening <tt class="xref c c-func docutils literal">exec()</tt>) can create problems for some extension modules.
Extension module authors should exercise caution when initializing internal data
structures. Note also that the <a class="reference internal" href="../library/functions.html#reload" title="reload"><tt class="xref py py-func docutils literal">reload()</tt></a> function can be used with
extension modules, and will call the module initialization function
(<tt class="xref c c-func docutils literal">initspam()</tt> in the example), but will not load the module again if it was
loaded from a dynamically loadable object file (<tt class="file docutils literal">.so</tt> on Unix,
<tt class="file docutils literal">.dll</tt> on Windows).
</div>
A more substantial example module is included in the Python source distribution
as <tt class="file docutils literal">Modules/xxmodule.c</tt>. This file may be used as a template or simply
read as an example.
</div>
<div class="section" id="compilation-and-linkage">
<h2>1.5. Compilation and Linkage<a class="headerlink" href="#compilation-and-linkage" title="Permalink to this headline">¶</a></h2>
There are two more things to do before you can use your new extension: compiling
and linking it with the Python system. If you use dynamic loading, the details
may depend on the style of dynamic loading your system uses; see the chapters
about building extension modules (chapter <a class="reference internal" href="building.html#building">Building C and C++ Extensions with distutils</a>) and additional
information that pertains only to building on Windows (chapter
<a class="reference internal" href="windows.html#building-on-windows">Building C and C++ Extensions on Windows</a>) for more information about this.
If you can&#8217;t use dynamic loading, or if you want to make your module a permanent
part of the Python interpreter, you will have to change the configuration setup
and rebuild the interpreter. Luckily, this is very simple on Unix: just place
your file (<tt class="file docutils literal">spammodule.c</tt> for example) in the <tt class="file docutils literal">Modules/</tt> directory
of an unpacked source distribution, add a line to the file
<tt class="file docutils literal">Modules/Setup.local</tt> describing your file:
<div class="highlight-c"><div class="highlight"><pre>spam spammodule.o
</pre></div>
</div>
and rebuild the interpreter by running make in the toplevel
directory. You can also run make in the <tt class="file docutils literal">Modules/</tt>
subdirectory, but then you must first rebuild <tt class="file docutils literal">Makefile</tt> there by running
&#8216;make Makefile&#8217;. (This is necessary each time you change the
<tt class="file docutils literal">Setup</tt> file.)
If your module requires additional libraries to link with, these can be listed
on the line in the configuration file as well, for instance:
<div class="highlight-c"><div class="highlight"><pre>spam spammodule.o -lX11
</pre></div>
</div>
</div>
<div class="section" id="calling-python-functions-from-c">
<h2>1.6. Calling Python Functions from C<a class="headerlink" href="#calling-python-functions-from-c" title="Permalink to this headline">¶</a></h2>
So far we have concentrated on making C functions callable from Python. The
reverse is also useful: calling Python functions from C. This is especially the
case for libraries that support so-called &#8220;callback&#8221; functions. If a C
interface makes use of callbacks, the equivalent Python often needs to provide a
callback mechanism to the Python programmer; the implementation will require
calling the Python callback functions from a C callback. Other uses are also
imaginable.
Fortunately, the Python interpreter is easily called recursively, and there is a
standard interface to call a Python function. (I won&#8217;t dwell on how to call the
Python parser with a particular string as input &#8212; if you&#8217;re interested, have a
look at the implementation of the <a class="reference internal" href="../using/cmdline.html#cmdoption-c">-c</a> command line option in
<tt class="file docutils literal">Modules/main.c</tt> from the Python source code.)
Calling a Python function is easy. First, the Python program must somehow pass
you the Python function object. You should provide a function (or some other
interface) to do this. When this function is called, save a pointer to the
Python function object (be careful to <a class="reference internal" href="../c-api/refcounting.html#Py_INCREF" title="Py_INCREF"><tt class="xref c c-func docutils literal">Py_INCREF()</tt></a> it!) in a global
variable &#8212; or wherever you see fit. For example, the following function might
be part of a module definition:
<div class="highlight-c"><div class="highlight"><pre>static PyObject *my_callback = NULL;

static PyObject *
my_set_callback(PyObject *dummy, PyObject *args)
{
 PyObject *result = NULL;
 PyObject *temp;

if (PyArg_ParseTuple(args, &quot;O:set_callback&quot;, &amp;temp)) {
 if (!PyCallable_Check(temp)) {
 PyErr_SetString(PyExc_TypeError, &quot;parameter must be callable&quot;);
 return NULL;
 }
 Py_XINCREF(temp); /* Add a reference to new callback */
 Py_XDECREF(my_callback); /* Dispose of previous callback */
 my_callback = temp; /* Remember new callback */
 /* Boilerplate to return &quot;None&quot; */
 Py_INCREF(Py_None);
 result = Py_None;
 }
 return result;
}
</pre></div>
</div>
This function must be registered with the interpreter using the
<a class="reference internal" href="../c-api/structures.html#METH_VARARGS" title="METH_VARARGS"><tt class="xref py py-const docutils literal">METH_VARARGS</tt></a> flag; this is described in section <a class="reference internal" href="#methodtable">The Module&#8217;s Method Table and Initialization Function</a>. The
<a class="reference internal" href="../c-api/arg.html#PyArg_ParseTuple" title="PyArg_ParseTuple"><tt class="xref c c-func docutils literal">PyArg_ParseTuple()</tt></a> function and its arguments are documented in section
<a class="reference internal" href="#parsetuple">Extracting Parameters in Extension Functions</a>.
The macros <a class="reference internal" href="../c-api/refcounting.html#Py_XINCREF" title="Py_XINCREF"><tt class="xref c c-func docutils literal">Py_XINCREF()</tt></a> and <a class="reference internal" href="../c-api/refcounting.html#Py_XDECREF" title="Py_XDECREF"><tt class="xref c c-func docutils literal">Py_XDECREF()</tt></a> increment/decrement the
reference count of an object and are safe in the presence of NULL pointers
(but note that temp will not be NULL in this context). More info on them
in section <a class="reference internal" href="#refcounts">Reference Counts</a>.
Later, when it is time to call the function, you call the C function
<a class="reference internal" href="../c-api/object.html#PyObject_CallObject" title="PyObject_CallObject"><tt class="xref c c-func docutils literal">PyObject_CallObject()</tt></a>. This function has two arguments, both pointers to
arbitrary Python objects: the Python function, and the argument list. The
argument list must always be a tuple object, whose length is the number of
arguments. To call the Python function with no arguments, pass in NULL, or
an empty tuple; to call it with one argument, pass a singleton tuple.
<a class="reference internal" href="../c-api/arg.html#Py_BuildValue" title="Py_BuildValue"><tt class="xref c c-func docutils literal">Py_BuildValue()</tt></a> returns a tuple when its format string consists of zero
or more format codes between parentheses. For example:
<div class="highlight-c"><div class="highlight"><pre>int arg;
PyObject *arglist;
PyObject *result;
...
arg = 123;
...
/* Time to call the callback */
arglist = Py_BuildValue(&quot;(i)&quot;, arg);
result = PyObject_CallObject(my_callback, arglist);
Py_DECREF(arglist);
</pre></div>
</div>
<a class="reference internal" href="../c-api/object.html#PyObject_CallObject" title="PyObject_CallObject"><tt class="xref c c-func docutils literal">PyObject_CallObject()</tt></a> returns a Python object pointer: this is the return
value of the Python function. <a class="reference internal" href="../c-api/object.html#PyObject_CallObject" title="PyObject_CallObject"><tt class="xref c c-func docutils literal">PyObject_CallObject()</tt></a> is
&#8220;reference-count-neutral&#8221; with respect to its arguments. In the example a new
tuple was created to serve as the argument list, which is <a class="reference internal" href="../c-api/refcounting.html#Py_DECREF" title="Py_DECREF"><tt class="xref c c-func docutils literal">Py_DECREF()</tt></a>-ed immediately after the call.
The return value of <a class="reference internal" href="../c-api/object.html#PyObject_CallObject" title="PyObject_CallObject"><tt class="xref c c-func docutils literal">PyObject_CallObject()</tt></a> is &#8220;new&#8221;: either it is a brand
new object, or it is an existing object whose reference count has been
incremented. So, unless you want to save it in a global variable, you should
somehow <a class="reference internal" href="../c-api/refcounting.html#Py_DECREF" title="Py_DECREF"><tt class="xref c c-func docutils literal">Py_DECREF()</tt></a> the result, even (especially!) if you are not
interested in its value.
Before you do this, however, it is important to check that the return value
isn&#8217;t NULL. If it is, the Python function terminated by raising an exception.
If the C code that called <a class="reference internal" href="../c-api/object.html#PyObject_CallObject" title="PyObject_CallObject"><tt class="xref c c-func docutils literal">PyObject_CallObject()</tt></a> is called from Python, it
should now return an error indication to its Python caller, so the interpreter
can print a stack trace, or the calling Python code can handle the exception.
If this is not possible or desirable, the exception should be cleared by calling
<a class="reference internal" href="../c-api/exceptions.html#PyErr_Clear" title="PyErr_Clear"><tt class="xref c c-func docutils literal">PyErr_Clear()</tt></a>. For example:
<div class="highlight-c"><div class="highlight"><pre>if (result == NULL)
 return NULL; /* Pass error back */
...use result...
Py_DECREF(result);
</pre></div>
</div>
Depending on the desired interface to the Python callback function, you may also
have to provide an argument list to <a class="reference internal" href="../c-api/object.html#PyObject_CallObject" title="PyObject_CallObject"><tt class="xref c c-func docutils literal">PyObject_CallObject()</tt></a>. In some cases
the argument list is also provided by the Python program, through the same
interface that specified the callback function. It can then be saved and used
in the same manner as the function object. In other cases, you may have to
construct a new tuple to pass as the argument list. The simplest way to do this
is to call <a class="reference internal" href="../c-api/arg.html#Py_BuildValue" title="Py_BuildValue"><tt class="xref c c-func docutils literal">Py_BuildValue()</tt></a>. For example, if you want to pass an integral
event code, you might use the following code:
<div class="highlight-c"><div class="highlight"><pre>PyObject *arglist;
...
arglist = Py_BuildValue(&quot;(l)&quot;, eventcode);
result = PyObject_CallObject(my_callback, arglist);
Py_DECREF(arglist);
if (result == NULL)
 return NULL; /* Pass error back */
/* Here maybe use the result */
Py_DECREF(result);
</pre></div>
</div>
Note the placement of <tt class="docutils literal">Py_DECREF(arglist)</tt> immediately after the call, before
the error check! Also note that strictly speaking this code is not complete:
<a class="reference internal" href="../c-api/arg.html#Py_BuildValue" title="Py_BuildValue"><tt class="xref c c-func docutils literal">Py_BuildValue()</tt></a> may run out of memory, and this should be checked.
You may also call a function with keyword arguments by using
<a class="reference internal" href="../c-api/object.html#PyObject_Call" title="PyObject_Call"><tt class="xref c c-func docutils literal">PyObject_Call()</tt></a>, which supports arguments and keyword arguments. As in
the above example, we use <a class="reference internal" href="../c-api/arg.html#Py_BuildValue" title="Py_BuildValue"><tt class="xref c c-func docutils literal">Py_BuildValue()</tt></a> to construct the dictionary.
<div class="highlight-c"><div class="highlight"><pre>PyObject *dict;
...
dict = Py_BuildValue(&quot;{s:i}&quot;, &quot;name&quot;, val);
result = PyObject_Call(my_callback, NULL, dict);
Py_DECREF(dict);
if (result == NULL)
 return NULL; /* Pass error back */
/* Here maybe use the result */
Py_DECREF(result);
</pre></div>
</div>
</div>
<div class="section" id="extracting-parameters-in-extension-functions">
<h2>1.7. Extracting Parameters in Extension Functions<a class="headerlink" href="#extracting-parameters-in-extension-functions" title="Permalink to this headline">¶</a></h2>
The <a class="reference internal" href="../c-api/arg.html#PyArg_ParseTuple" title="PyArg_ParseTuple"><tt class="xref c c-func docutils literal">PyArg_ParseTuple()</tt></a> function is declared as follows:
<div class="highlight-c"><div class="highlight"><pre>int PyArg_ParseTuple(PyObject *arg, char *format, ...);
</pre></div>
</div>
The arg argument must be a tuple object containing an argument list passed
from Python to a C function. The format argument must be a format string,
whose syntax is explained in <a class="reference internal" href="../c-api/arg.html#arg-parsing">Parsing arguments and building values</a> in the Python/C API Reference
Manual. The remaining arguments must be addresses of variables whose type is
determined by the format string.
Note that while <a class="reference internal" href="../c-api/arg.html#PyArg_ParseTuple" title="PyArg_ParseTuple"><tt class="xref c c-func docutils literal">PyArg_ParseTuple()</tt></a> checks that the Python arguments have
the required types, it cannot check the validity of the addresses of C variables
passed to the call: if you make mistakes there, your code will probably crash or
at least overwrite random bits in memory. So be careful!
Note that any Python object references which are provided to the caller are
borrowed references; do not decrement their reference count!
Some example calls:
<div class="highlight-c"><div class="highlight"><pre>int ok;
int i, j;
long k, l;
const char *s;
int size;

ok = PyArg_ParseTuple(args, &quot;&quot;); /* No arguments */
 /* Python call: f() */
</pre></div>
</div>
<div class="highlight-c"><div class="highlight"><pre>ok = PyArg_ParseTuple(args, &quot;s&quot;, &amp;s); /* A string */
 /* Possible Python call: f(&#39;whoops!&#39;) */
</pre></div>
</div>
<div class="highlight-c"><div class="highlight"><pre>ok = PyArg_ParseTuple(args, &quot;lls&quot;, &amp;k, &amp;l, &amp;s); /* Two longs and a string */
 /* Possible Python call: f(1, 2, &#39;three&#39;) */
</pre></div>
</div>
<div class="highlight-c"><div class="highlight"><pre>ok = PyArg_ParseTuple(args, &quot;(ii)s#&quot;, &amp;i, &amp;j, &amp;s, &amp;size);
 /* A pair of ints and a string, whose size is also returned */
 /* Possible Python call: f((1, 2), &#39;three&#39;) */
</pre></div>
</div>
<div class="highlight-c"><div class="highlight"><pre>{
 const char *file;
 const char *mode = &quot;r&quot;;
 int bufsize = 0;
 ok = PyArg_ParseTuple(args, &quot;s|si&quot;, &amp;file, &amp;mode, &amp;bufsize);
 /* A string, and optionally another string and an integer */
 /* Possible Python calls:
 f(&#39;spam&#39;)
 f(&#39;spam&#39;, &#39;w&#39;)
 f(&#39;spam&#39;, &#39;wb&#39;, 100000) */
}
</pre></div>
</div>
<div class="highlight-c"><div class="highlight"><pre>{
 int left, top, right, bottom, h, v;
 ok = PyArg_ParseTuple(args, &quot;((ii)(ii))(ii)&quot;,
 &amp;left, &amp;top, &amp;right, &amp;bottom, &amp;h, &amp;v);
 /* A rectangle and a point */
 /* Possible Python call:
 f(((0, 0), (400, 300)), (10, 10)) */
}
</pre></div>
</div>
<div class="highlight-c"><div class="highlight"><pre>{
 Py_complex c;
 ok = PyArg_ParseTuple(args, &quot;D:myfunction&quot;, &amp;c);
 /* a complex, also providing a function name for errors */
 /* Possible Python call: myfunction(1+2j) */
}
</pre></div>
</div>
</div>
<div class="section" id="keyword-parameters-for-extension-functions">
<h2>1.8. Keyword Parameters for Extension Functions<a class="headerlink" href="#keyword-parameters-for-extension-functions" title="Permalink to this headline">¶</a></h2>
The <a class="reference internal" href="../c-api/arg.html#PyArg_ParseTupleAndKeywords" title="PyArg_ParseTupleAndKeywords"><tt class="xref c c-func docutils literal">PyArg_ParseTupleAndKeywords()</tt></a> function is declared as follows:
<div class="highlight-c"><div class="highlight"><pre>int PyArg_ParseTupleAndKeywords(PyObject *arg, PyObject *kwdict,
 char *format, char *kwlist[], ...);
</pre></div>
</div>
The arg and format parameters are identical to those of the
<a class="reference internal" href="../c-api/arg.html#PyArg_ParseTuple" title="PyArg_ParseTuple"><tt class="xref c c-func docutils literal">PyArg_ParseTuple()</tt></a> function. The kwdict parameter is the dictionary of
keywords received as the third parameter from the Python runtime. The kwlist
parameter is a NULL-terminated list of strings which identify the parameters;
the names are matched with the type information from format from left to
right. On success, <a class="reference internal" href="../c-api/arg.html#PyArg_ParseTupleAndKeywords" title="PyArg_ParseTupleAndKeywords"><tt class="xref c c-func docutils literal">PyArg_ParseTupleAndKeywords()</tt></a> returns true, otherwise
it returns false and raises an appropriate exception.
<div class="admonition note">
Note
Nested tuples cannot be parsed when using keyword arguments! Keyword parameters
passed in which are not present in the kwlist will cause <a class="reference internal" href="../library/exceptions.html#exceptions.TypeError" title="exceptions.TypeError"><tt class="xref py py-exc docutils literal">TypeError</tt></a> to
be raised.
</div>
Here is an example module which uses keywords, based on an example by Geoff
Philbrick (<a class="reference external" href="mailto:philbrick&#37;&#52;&#48;hks&#46;com">philbrick&#64;hks&#46;com</a>):
<div class="highlight-c"><div class="highlight"><pre>#include &quot;Python.h&quot;

static PyObject *
keywdarg_parrot(PyObject *self, PyObject *args, PyObject *keywds)
{
 int voltage;
 char *state = &quot;a stiff&quot;;
 char *action = &quot;voom&quot;;
 char *type = &quot;Norwegian Blue&quot;;

static char *kwlist[] = {&quot;voltage&quot;, &quot;state&quot;, &quot;action&quot;, &quot;type&quot;, NULL};

if (!PyArg_ParseTupleAndKeywords(args, keywds, &quot;i|sss&quot;, kwlist,
 &amp;voltage, &amp;state, &amp;action, &amp;type))
 return NULL;

printf(&quot;-- This parrot wouldn&#39;t %s if you put %i Volts through it.\n&quot;,
 action, voltage);
 printf(&quot;-- Lovely plumage, the %s -- It&#39;s %s!\n&quot;, type, state);

Py_INCREF(Py_None);

return Py_None;
}

static PyMethodDef keywdarg_methods[] = {
 /* The cast of the function is necessary since PyCFunction values
 * only take two PyObject* parameters, and keywdarg_parrot() takes
 * three.
 */
 {&quot;parrot&quot;, (PyCFunction)keywdarg_parrot, METH_VARARGS | METH_KEYWORDS,
 &quot;Print a lovely skit to standard output.&quot;},
 {NULL, NULL, 0, NULL} /* sentinel */
};
</pre></div>
</div>
<div class="highlight-c"><div class="highlight"><pre>void
initkeywdarg(void)
{
 /* Create the module and add the functions */
 Py_InitModule(&quot;keywdarg&quot;, keywdarg_methods);
}
</pre></div>
</div>
</div>
<div class="section" id="building-arbitrary-values">
<h2>1.9. Building Arbitrary Values<a class="headerlink" href="#building-arbitrary-values" title="Permalink to this headline">¶</a></h2>
This function is the counterpart to <a class="reference internal" href="../c-api/arg.html#PyArg_ParseTuple" title="PyArg_ParseTuple"><tt class="xref c c-func docutils literal">PyArg_ParseTuple()</tt></a>. It is declared
as follows:
<div class="highlight-c"><div class="highlight"><pre>PyObject *Py_BuildValue(char *format, ...);
</pre></div>
</div>
It recognizes a set of format units similar to the ones recognized by
<a class="reference internal" href="../c-api/arg.html#PyArg_ParseTuple" title="PyArg_ParseTuple"><tt class="xref c c-func docutils literal">PyArg_ParseTuple()</tt></a>, but the arguments (which are input to the function,
not output) must not be pointers, just values. It returns a new Python object,
suitable for returning from a C function called from Python.
One difference with <a class="reference internal" href="../c-api/arg.html#PyArg_ParseTuple" title="PyArg_ParseTuple"><tt class="xref c c-func docutils literal">PyArg_ParseTuple()</tt></a>: while the latter requires its
first argument to be a tuple (since Python argument lists are always represented
as tuples internally), <a class="reference internal" href="../c-api/arg.html#Py_BuildValue" title="Py_BuildValue"><tt class="xref c c-func docutils literal">Py_BuildValue()</tt></a> does not always build a tuple. It
builds a tuple only if its format string contains two or more format units. If
the format string is empty, it returns <tt class="docutils literal">None</tt>; if it contains exactly one
format unit, it returns whatever object is described by that format unit. To
force it to return a tuple of size 0 or one, parenthesize the format string.
Examples (to the left the call, to the right the resulting Python value):
<div class="highlight-c"><pre>Py_BuildValue("") None
Py_BuildValue("i", 123) 123
Py_BuildValue("iii", 123, 456, 789) (123, 456, 789)
Py_BuildValue("s", "hello") 'hello'
Py_BuildValue("ss", "hello", "world") ('hello', 'world')
Py_BuildValue("s#", "hello", 4) 'hell'
Py_BuildValue("()") ()
Py_BuildValue("(i)", 123) (123,)
Py_BuildValue("(ii)", 123, 456) (123, 456)
Py_BuildValue("(i,i)", 123, 456) (123, 456)
Py_BuildValue("[i,i]", 123, 456) [123, 456]
Py_BuildValue("{s:i,s:i}",
 "abc", 123, "def", 456) {'abc': 123, 'def': 456}
Py_BuildValue("((ii)(ii)) (ii)",
 1, 2, 3, 4, 5, 6) (((1, 2), (3, 4)), (5, 6))</pre>
</div>
</div>
<div class="section" id="reference-counts">
<h2>1.10. Reference Counts<a class="headerlink" href="#reference-counts" title="Permalink to this headline">¶</a></h2>
In languages like C or C++, the programmer is responsible for dynamic allocation
and deallocation of memory on the heap. In C, this is done using the functions
<tt class="xref c c-func docutils literal">malloc()</tt> and <tt class="xref c c-func docutils literal">free()</tt>. In C++, the operators <tt class="docutils literal">new</tt> and
<tt class="docutils literal">delete</tt> are used with essentially the same meaning and we&#8217;ll restrict
the following discussion to the C case.
Every block of memory allocated with <tt class="xref c c-func docutils literal">malloc()</tt> should eventually be
returned to the pool of available memory by exactly one call to <tt class="xref c c-func docutils literal">free()</tt>.
It is important to call <tt class="xref c c-func docutils literal">free()</tt> at the right time. If a block&#8217;s address
is forgotten but <tt class="xref c c-func docutils literal">free()</tt> is not called for it, the memory it occupies
cannot be reused until the program terminates. This is called a memory
leak. On the other hand, if a program calls <tt class="xref c c-func docutils literal">free()</tt> for a block and then
continues to use the block, it creates a conflict with re-use of the block
through another <tt class="xref c c-func docutils literal">malloc()</tt> call. This is called using freed memory.
It has the same bad consequences as referencing uninitialized data &#8212; core
dumps, wrong results, mysterious crashes.
Common causes of memory leaks are unusual paths through the code. For instance,
a function may allocate a block of memory, do some calculation, and then free
the block again. Now a change in the requirements for the function may add a
test to the calculation that detects an error condition and can return
prematurely from the function. It&#8217;s easy to forget to free the allocated memory
block when taking this premature exit, especially when it is added later to the
code. Such leaks, once introduced, often go undetected for a long time: the
error exit is taken only in a small fraction of all calls, and most modern
machines have plenty of virtual memory, so the leak only becomes apparent in a
long-running process that uses the leaking function frequently. Therefore, it&#8217;s
important to prevent leaks from happening by having a coding convention or
strategy that minimizes this kind of errors.
Since Python makes heavy use of <tt class="xref c c-func docutils literal">malloc()</tt> and <tt class="xref c c-func docutils literal">free()</tt>, it needs a
strategy to avoid memory leaks as well as the use of freed memory. The chosen
method is called reference counting. The principle is simple: every
object contains a counter, which is incremented when a reference to the object
is stored somewhere, and which is decremented when a reference to it is deleted.
When the counter reaches zero, the last reference to the object has been deleted
and the object is freed.
An alternative strategy is called automatic garbage collection.
(Sometimes, reference counting is also referred to as a garbage collection
strategy, hence my use of &#8220;automatic&#8221; to distinguish the two.) The big
advantage of automatic garbage collection is that the user doesn&#8217;t need to call
<tt class="xref c c-func docutils literal">free()</tt> explicitly. (Another claimed advantage is an improvement in speed
or memory usage &#8212; this is no hard fact however.) The disadvantage is that for
C, there is no truly portable automatic garbage collector, while reference
counting can be implemented portably (as long as the functions <tt class="xref c c-func docutils literal">malloc()</tt>
and <tt class="xref c c-func docutils literal">free()</tt> are available &#8212; which the C Standard guarantees). Maybe some
day a sufficiently portable automatic garbage collector will be available for C.
Until then, we&#8217;ll have to live with reference counts.
While Python uses the traditional reference counting implementation, it also
offers a cycle detector that works to detect reference cycles. This allows
applications to not worry about creating direct or indirect circular references;
these are the weakness of garbage collection implemented using only reference
counting. Reference cycles consist of objects which contain (possibly indirect)
references to themselves, so that each object in the cycle has a reference count
which is non-zero. Typical reference counting implementations are not able to
reclaim the memory belonging to any objects in a reference cycle, or referenced
from the objects in the cycle, even though there are no further references to
the cycle itself.
The cycle detector is able to detect garbage cycles and can reclaim them so long
as there are no finalizers implemented in Python (<a class="reference internal" href="../reference/datamodel.html#object.__del__" title="object.__del__"><tt class="xref py py-meth docutils literal">__del__()</tt></a> methods).
When there are such finalizers, the detector exposes the cycles through the
<a class="reference internal" href="../library/gc.html#module-gc" title="gc: Interface to the cycle-detecting garbage collector."><tt class="xref py py-mod docutils literal">gc</tt></a> module (specifically, the
<tt class="docutils literal">garbage</tt> variable in that module). The <a class="reference internal" href="../library/gc.html#module-gc" title="gc: Interface to the cycle-detecting garbage collector."><tt class="xref py py-mod docutils literal">gc</tt></a> module also exposes a way
to run the detector (the <tt class="xref py py-func docutils literal">collect()</tt> function), as well as configuration
interfaces and the ability to disable the detector at runtime. The cycle
detector is considered an optional component; though it is included by default,
it can be disabled at build time using the --without-cycle-gc option
to the configure script on Unix platforms (including Mac OS X) or by
removing the definition of <tt class="docutils literal">WITH_CYCLE_GC</tt> in the <tt class="file docutils literal">pyconfig.h</tt> header on
other platforms. If the cycle detector is disabled in this way, the <a class="reference internal" href="../library/gc.html#module-gc" title="gc: Interface to the cycle-detecting garbage collector."><tt class="xref py py-mod docutils literal">gc</tt></a>
module will not be available.
<div class="section" id="reference-counting-in-python">
<h3>1.10.1. Reference Counting in Python<a class="headerlink" href="#reference-counting-in-python" title="Permalink to this headline">¶</a></h3>
There are two macros, <tt class="docutils literal">Py_INCREF(x)</tt> and <tt class="docutils literal">Py_DECREF(x)</tt>, which handle the
incrementing and decrementing of the reference count. <a class="reference internal" href="../c-api/refcounting.html#Py_DECREF" title="Py_DECREF"><tt class="xref c c-func docutils literal">Py_DECREF()</tt></a> also
frees the object when the count reaches zero. For flexibility, it doesn&#8217;t call
<tt class="xref c c-func docutils literal">free()</tt> directly &#8212; rather, it makes a call through a function pointer in
the object&#8217;s type object. For this purpose (and others), every object
also contains a pointer to its type object.
The big question now remains: when to use <tt class="docutils literal">Py_INCREF(x)</tt> and <tt class="docutils literal">Py_DECREF(x)</tt>?
Let&#8217;s first introduce some terms. Nobody &#8220;owns&#8221; an object; however, you can
own a reference to an object. An object&#8217;s reference count is now defined
as the number of owned references to it. The owner of a reference is
responsible for calling <a class="reference internal" href="../c-api/refcounting.html#Py_DECREF" title="Py_DECREF"><tt class="xref c c-func docutils literal">Py_DECREF()</tt></a> when the reference is no longer
needed. Ownership of a reference can be transferred. There are three ways to
dispose of an owned reference: pass it on, store it, or call <a class="reference internal" href="../c-api/refcounting.html#Py_DECREF" title="Py_DECREF"><tt class="xref c c-func docutils literal">Py_DECREF()</tt></a>.
Forgetting to dispose of an owned reference creates a memory leak.
It is also possible to borrow <a class="footnote-reference" href="#id6" id="id2">[2]</a> a reference to an object. The
borrower of a reference should not call <a class="reference internal" href="../c-api/refcounting.html#Py_DECREF" title="Py_DECREF"><tt class="xref c c-func docutils literal">Py_DECREF()</tt></a>. The borrower must
not hold on to the object longer than the owner from which it was borrowed.
Using a borrowed reference after the owner has disposed of it risks using freed
memory and should be avoided completely. <a class="footnote-reference" href="#id7" id="id3">[3]</a>
The advantage of borrowing over owning a reference is that you don&#8217;t need to
take care of disposing of the reference on all possible paths through the code
&#8212; in other words, with a borrowed reference you don&#8217;t run the risk of leaking
when a premature exit is taken. The disadvantage of borrowing over owning is
that there are some subtle situations where in seemingly correct code a borrowed
reference can be used after the owner from which it was borrowed has in fact
disposed of it.
A borrowed reference can be changed into an owned reference by calling
<a class="reference internal" href="../c-api/refcounting.html#Py_INCREF" title="Py_INCREF"><tt class="xref c c-func docutils literal">Py_INCREF()</tt></a>. This does not affect the status of the owner from which the
reference was borrowed &#8212; it creates a new owned reference, and gives full
owner responsibilities (the new owner must dispose of the reference properly, as
well as the previous owner).
</div>
<div class="section" id="ownership-rules">
<h3>1.10.2. Ownership Rules<a class="headerlink" href="#ownership-rules" title="Permalink to this headline">¶</a></h3>
Whenever an object reference is passed into or out of a function, it is part of
the function&#8217;s interface specification whether ownership is transferred with the
reference or not.
Most functions that return a reference to an object pass on ownership with the
reference. In particular, all functions whose function it is to create a new
object, such as <a class="reference internal" href="../c-api/int.html#PyInt_FromLong" title="PyInt_FromLong"><tt class="xref c c-func docutils literal">PyInt_FromLong()</tt></a> and <a class="reference internal" href="../c-api/arg.html#Py_BuildValue" title="Py_BuildValue"><tt class="xref c c-func docutils literal">Py_BuildValue()</tt></a>, pass
ownership to the receiver. Even if the object is not actually new, you still
receive ownership of a new reference to that object. For instance,
<a class="reference internal" href="../c-api/int.html#PyInt_FromLong" title="PyInt_FromLong"><tt class="xref c c-func docutils literal">PyInt_FromLong()</tt></a> maintains a cache of popular values and can return a
reference to a cached item.
Many functions that extract objects from other objects also transfer ownership
with the reference, for instance <a class="reference internal" href="../c-api/object.html#PyObject_GetAttrString" title="PyObject_GetAttrString"><tt class="xref c c-func docutils literal">PyObject_GetAttrString()</tt></a>. The picture
is less clear, here, however, since a few common routines are exceptions:
<a class="reference internal" href="../c-api/tuple.html#PyTuple_GetItem" title="PyTuple_GetItem"><tt class="xref c c-func docutils literal">PyTuple_GetItem()</tt></a>, <a class="reference internal" href="../c-api/list.html#PyList_GetItem" title="PyList_GetItem"><tt class="xref c c-func docutils literal">PyList_GetItem()</tt></a>, <a class="reference internal" href="../c-api/dict.html#PyDict_GetItem" title="PyDict_GetItem"><tt class="xref c c-func docutils literal">PyDict_GetItem()</tt></a>, and
<a class="reference internal" href="../c-api/dict.html#PyDict_GetItemString" title="PyDict_GetItemString"><tt class="xref c c-func docutils literal">PyDict_GetItemString()</tt></a> all return references that you borrow from the
tuple, list or dictionary.
The function <a class="reference internal" href="../c-api/import.html#PyImport_AddModule" title="PyImport_AddModule"><tt class="xref c c-func docutils literal">PyImport_AddModule()</tt></a> also returns a borrowed reference, even
though it may actually create the object it returns: this is possible because an
owned reference to the object is stored in <tt class="docutils literal">sys.modules</tt>.
When you pass an object reference into another function, in general, the
function borrows the reference from you &#8212; if it needs to store it, it will use
<a class="reference internal" href="../c-api/refcounting.html#Py_INCREF" title="Py_INCREF"><tt class="xref c c-func docutils literal">Py_INCREF()</tt></a> to become an independent owner. There are exactly two
important exceptions to this rule: <a class="reference internal" href="../c-api/tuple.html#PyTuple_SetItem" title="PyTuple_SetItem"><tt class="xref c c-func docutils literal">PyTuple_SetItem()</tt></a> and
<a class="reference internal" href="../c-api/list.html#PyList_SetItem" title="PyList_SetItem"><tt class="xref c c-func docutils literal">PyList_SetItem()</tt></a>. These functions take over ownership of the item passed
to them &#8212; even if they fail! (Note that <a class="reference internal" href="../c-api/dict.html#PyDict_SetItem" title="PyDict_SetItem"><tt class="xref c c-func docutils literal">PyDict_SetItem()</tt></a> and friends
don&#8217;t take over ownership &#8212; they are &#8220;normal.&#8221;)
When a C function is called from Python, it borrows references to its arguments
from the caller. The caller owns a reference to the object, so the borrowed
reference&#8217;s lifetime is guaranteed until the function returns. Only when such a
borrowed reference must be stored or passed on, it must be turned into an owned
reference by calling <a class="reference internal" href="../c-api/refcounting.html#Py_INCREF" title="Py_INCREF"><tt class="xref c c-func docutils literal">Py_INCREF()</tt></a>.
The object reference returned from a C function that is called from Python must
be an owned reference &#8212; ownership is transferred from the function to its
caller.
</div>
<div class="section" id="thin-ice">
<h3>1.10.3. Thin Ice<a class="headerlink" href="#thin-ice" title="Permalink to this headline">¶</a></h3>
There are a few situations where seemingly harmless use of a borrowed reference
can lead to problems. These all have to do with implicit invocations of the
interpreter, which can cause the owner of a reference to dispose of it.
The first and most important case to know about is using <a class="reference internal" href="../c-api/refcounting.html#Py_DECREF" title="Py_DECREF"><tt class="xref c c-func docutils literal">Py_DECREF()</tt></a> on
an unrelated object while borrowing a reference to a list item. For instance:
<div class="highlight-c"><div class="highlight"><pre>void
bug(PyObject *list)
{
 PyObject *item = PyList_GetItem(list, 0);

PyList_SetItem(list, 1, PyInt_FromLong(0L));
 PyObject_Print(item, stdout, 0); /* BUG! */
}
</pre></div>
</div>
This function first borrows a reference to <tt class="docutils literal">list[0]</tt>, then replaces
<tt class="docutils literal">list[1]</tt> with the value <tt class="docutils literal">0</tt>, and finally prints the borrowed reference.
Looks harmless, right? But it&#8217;s not!
Let&#8217;s follow the control flow into <a class="reference internal" href="../c-api/list.html#PyList_SetItem" title="PyList_SetItem"><tt class="xref c c-func docutils literal">PyList_SetItem()</tt></a>. The list owns
references to all its items, so when item 1 is replaced, it has to dispose of
the original item 1. Now let&#8217;s suppose the original item 1 was an instance of a
user-defined class, and let&#8217;s further suppose that the class defined a
<a class="reference internal" href="../reference/datamodel.html#object.__del__" title="object.__del__"><tt class="xref py py-meth docutils literal">__del__()</tt></a> method. If this class instance has a reference count of 1,
disposing of it will call its <a class="reference internal" href="../reference/datamodel.html#object.__del__" title="object.__del__"><tt class="xref py py-meth docutils literal">__del__()</tt></a> method.
Since it is written in Python, the <a class="reference internal" href="../reference/datamodel.html#object.__del__" title="object.__del__"><tt class="xref py py-meth docutils literal">__del__()</tt></a> method can execute arbitrary
Python code. Could it perhaps do something to invalidate the reference to
<tt class="docutils literal">item</tt> in <tt class="xref c c-func docutils literal">bug()</tt>? You bet! Assuming that the list passed into
<tt class="xref c c-func docutils literal">bug()</tt> is accessible to the <a class="reference internal" href="../reference/datamodel.html#object.__del__" title="object.__del__"><tt class="xref py py-meth docutils literal">__del__()</tt></a> method, it could execute a
statement to the effect of <tt class="docutils literal">del list[0]</tt>, and assuming this was the last
reference to that object, it would free the memory associated with it, thereby
invalidating <tt class="docutils literal">item</tt>.
The solution, once you know the source of the problem, is easy: temporarily
increment the reference count. The correct version of the function reads:
<div class="highlight-c"><div class="highlight"><pre>void
no_bug(PyObject *list)
{
 PyObject *item = PyList_GetItem(list, 0);

Py_INCREF(item);
 PyList_SetItem(list, 1, PyInt_FromLong(0L));
 PyObject_Print(item, stdout, 0);
 Py_DECREF(item);
}
</pre></div>
</div>
This is a true story. An older version of Python contained variants of this bug
and someone spent a considerable amount of time in a C debugger to figure out
why his <a class="reference internal" href="../reference/datamodel.html#object.__del__" title="object.__del__"><tt class="xref py py-meth docutils literal">__del__()</tt></a> methods would fail...
The second case of problems with a borrowed reference is a variant involving
threads. Normally, multiple threads in the Python interpreter can&#8217;t get in each
other&#8217;s way, because there is a global lock protecting Python&#8217;s entire object
space. However, it is possible to temporarily release this lock using the macro
<a class="reference internal" href="../c-api/init.html#Py_BEGIN_ALLOW_THREADS" title="Py_BEGIN_ALLOW_THREADS"><tt class="xref c c-macro docutils literal">Py_BEGIN_ALLOW_THREADS</tt></a>, and to re-acquire it using
<a class="reference internal" href="../c-api/init.html#Py_END_ALLOW_THREADS" title="Py_END_ALLOW_THREADS"><tt class="xref c c-macro docutils literal">Py_END_ALLOW_THREADS</tt></a>. This is common around blocking I/O calls, to
let other threads use the processor while waiting for the I/O to complete.
Obviously, the following function has the same problem as the previous one:
<div class="highlight-c"><div class="highlight"><pre>void
bug(PyObject *list)
{
 PyObject *item = PyList_GetItem(list, 0);
 Py_BEGIN_ALLOW_THREADS
 ...some blocking I/O call...
 Py_END_ALLOW_THREADS
 PyObject_Print(item, stdout, 0); /* BUG! */
}
</pre></div>
</div>
</div>
<div class="section" id="null-pointers">
<h3>1.10.4. NULL Pointers<a class="headerlink" href="#null-pointers" title="Permalink to this headline">¶</a></h3>
In general, functions that take object references as arguments do not expect you
to pass them NULL pointers, and will dump core (or cause later core dumps) if
you do so. Functions that return object references generally return NULL only
to indicate that an exception occurred. The reason for not testing for NULL
arguments is that functions often pass the objects they receive on to other
function &#8212; if each function were to test for NULL, there would be a lot of
redundant tests and the code would run more slowly.
It is better to test for NULL only at the &#8220;source:&#8221; when a pointer that may be
NULL is received, for example, from <tt class="xref c c-func docutils literal">malloc()</tt> or from a function that
may raise an exception.
The macros <a class="reference internal" href="../c-api/refcounting.html#Py_INCREF" title="Py_INCREF"><tt class="xref c c-func docutils literal">Py_INCREF()</tt></a> and <a class="reference internal" href="../c-api/refcounting.html#Py_DECREF" title="Py_DECREF"><tt class="xref c c-func docutils literal">Py_DECREF()</tt></a> do not check for NULL
pointers &#8212; however, their variants <a class="reference internal" href="../c-api/refcounting.html#Py_XINCREF" title="Py_XINCREF"><tt class="xref c c-func docutils literal">Py_XINCREF()</tt></a> and <a class="reference internal" href="../c-api/refcounting.html#Py_XDECREF" title="Py_XDECREF"><tt class="xref c c-func docutils literal">Py_XDECREF()</tt></a>
do.
The macros for checking for a particular object type (<tt class="docutils literal">Pytype_Check()</tt>) don&#8217;t
check for NULL pointers &#8212; again, there is much code that calls several of
these in a row to test an object against various different expected types, and
this would generate redundant tests. There are no variants with NULL
checking.
The C function calling mechanism guarantees that the argument list passed to C
functions (<tt class="docutils literal">args</tt> in the examples) is never NULL &#8212; in fact it guarantees
that it is always a tuple. <a class="footnote-reference" href="#id8" id="id4">[4]</a>
It is a severe error to ever let a NULL pointer &#8220;escape&#8221; to the Python user.
</div>
</div>
<div class="section" id="writing-extensions-in-c">
<h2>1.11. Writing Extensions in C++<a class="headerlink" href="#writing-extensions-in-c" title="Permalink to this headline">¶</a></h2>
It is possible to write extension modules in C++. Some restrictions apply. If
the main program (the Python interpreter) is compiled and linked by the C
compiler, global or static objects with constructors cannot be used. This is
not a problem if the main program is linked by the C++ compiler. Functions that
will be called by the Python interpreter (in particular, module initialization
functions) have to be declared using <tt class="docutils literal">extern &quot;C&quot;</tt>. It is unnecessary to
enclose the Python header files in <tt class="docutils literal">extern &quot;C&quot; {...}</tt> &#8212; they use this form
already if the symbol <tt class="docutils literal">__cplusplus</tt> is defined (all recent C++ compilers
define this symbol).
</div>
<div class="section" id="providing-a-c-api-for-an-extension-module">
<h2>1.12. Providing a C API for an Extension Module<a class="headerlink" href="#providing-a-c-api-for-an-extension-module" title="Permalink to this headline">¶</a></h2>
Many extension modules just provide new functions and types to be used from
Python, but sometimes the code in an extension module can be useful for other
extension modules. For example, an extension module could implement a type
&#8220;collection&#8221; which works like lists without order. Just like the standard Python
list type has a C API which permits extension modules to create and manipulate
lists, this new collection type should have a set of C functions for direct
manipulation from other extension modules.
At first sight this seems easy: just write the functions (without declaring them
<tt class="docutils literal">static</tt>, of course), provide an appropriate header file, and document
the C API. And in fact this would work if all extension modules were always
linked statically with the Python interpreter. When modules are used as shared
libraries, however, the symbols defined in one module may not be visible to
another module. The details of visibility depend on the operating system; some
systems use one global namespace for the Python interpreter and all extension
modules (Windows, for example), whereas others require an explicit list of
imported symbols at module link time (AIX is one example), or offer a choice of
different strategies (most Unices). And even if symbols are globally visible,
the module whose functions one wishes to call might not have been loaded yet!
Portability therefore requires not to make any assumptions about symbol
visibility. This means that all symbols in extension modules should be declared
<tt class="docutils literal">static</tt>, except for the module&#8217;s initialization function, in order to
avoid name clashes with other extension modules (as discussed in section
<a class="reference internal" href="#methodtable">The Module&#8217;s Method Table and Initialization Function</a>). And it means that symbols that should be accessible from
other extension modules must be exported in a different way.
Python provides a special mechanism to pass C-level information (pointers) from
one extension module to another one: Capsules. A Capsule is a Python data type
which stores a pointer (<tt class="xref c c-type docutils literal">void *</tt>). Capsules can only be created and
accessed via their C API, but they can be passed around like any other Python
object. In particular, they can be assigned to a name in an extension module&#8217;s
namespace. Other extension modules can then import this module, retrieve the
value of this name, and then retrieve the pointer from the Capsule.
There are many ways in which Capsules can be used to export the C API of an
extension module. Each function could get its own Capsule, or all C API pointers
could be stored in an array whose address is published in a Capsule. And the
various tasks of storing and retrieving the pointers can be distributed in
different ways between the module providing the code and the client modules.
Whichever method you choose, it&#8217;s important to name your Capsules properly.
The function <a class="reference internal" href="../c-api/capsule.html#PyCapsule_New" title="PyCapsule_New"><tt class="xref c c-func docutils literal">PyCapsule_New()</tt></a> takes a name parameter
(<tt class="xref c c-type docutils literal">const char *</tt>); you&#8217;re permitted to pass in a NULL name, but
we strongly encourage you to specify a name. Properly named Capsules provide
a degree of runtime type-safety; there is no feasible way to tell one unnamed
Capsule from another.
In particular, Capsules used to expose C APIs should be given a name following
this convention:
<div class="highlight-c"><div class="highlight"><pre>modulename.attributename
</pre></div>
</div>
The convenience function <a class="reference internal" href="../c-api/capsule.html#PyCapsule_Import" title="PyCapsule_Import"><tt class="xref c c-func docutils literal">PyCapsule_Import()</tt></a> makes it easy to
load a C API provided via a Capsule, but only if the Capsule&#8217;s name
matches this convention. This behavior gives C API users a high degree
of certainty that the Capsule they load contains the correct C API.
The following example demonstrates an approach that puts most of the burden on
the writer of the exporting module, which is appropriate for commonly used
library modules. It stores all C API pointers (just one in the example!) in an
array of <tt class="xref c c-type docutils literal">void</tt> pointers which becomes the value of a Capsule. The header
file corresponding to the module provides a macro that takes care of importing
the module and retrieving its C API pointers; client modules only have to call
this macro before accessing the C API.
The exporting module is a modification of the <tt class="xref py py-mod docutils literal">spam</tt> module from section
<a class="reference internal" href="#extending-simpleexample">A Simple Example</a>. The function <tt class="xref py py-func docutils literal">spam.system()</tt> does not call
the C library function <tt class="xref c c-func docutils literal">system()</tt> directly, but a function
<tt class="xref c c-func docutils literal">PySpam_System()</tt>, which would of course do something more complicated in
reality (such as adding &#8220;spam&#8221; to every command). This function
<tt class="xref c c-func docutils literal">PySpam_System()</tt> is also exported to other extension modules.
The function <tt class="xref c c-func docutils literal">PySpam_System()</tt> is a plain C function, declared
<tt class="docutils literal">static</tt> like everything else:
<div class="highlight-c"><div class="highlight"><pre>static int
PySpam_System(const char *command)
{
 return system(command);
}
</pre></div>
</div>
The function <tt class="xref c c-func docutils literal">spam_system()</tt> is modified in a trivial way:
<div class="highlight-c"><div class="highlight"><pre>static PyObject *
spam_system(PyObject *self, PyObject *args)
{
 const char *command;
 int sts;

/* Initialize the C API pointer array */
 PySpam_API[PySpam_System_NUM] = (void *)PySpam_System;

/* Create a Capsule containing the API pointer array&#39;s address */
 c_api_object = PyCapsule_New((void *)PySpam_API, &quot;spam._C_API&quot;, NULL);

if (c_api_object != NULL)
 PyModule_AddObject(m, &quot;_C_API&quot;, c_api_object);
}
</pre></div>
</div>
Note that <tt class="docutils literal">PySpam_API</tt> is declared <tt class="docutils literal">static</tt>; otherwise the pointer
array would disappear when <tt class="xref py py-func docutils literal">initspam()</tt> terminates!
The bulk of the work is in the header file <tt class="file docutils literal">spammodule.h</tt>, which looks
like this:
<div class="highlight-c"><div class="highlight"><pre>#ifndef Py_SPAMMODULE_H
#define Py_SPAMMODULE_H
#ifdef __cplusplus
extern &quot;C&quot; {
#endif

/* Header file for spammodule */

/* C API functions */
#define PySpam_System_NUM 0
#define PySpam_System_RETURN int
#define PySpam_System_PROTO (const char *command)

/* Total number of C API pointers */
#define PySpam_API_pointers 1

#ifdef SPAM_MODULE
/* This section is used when compiling spammodule.c */

static PySpam_System_RETURN PySpam_System PySpam_System_PROTO;

#else
/* This section is used in modules that use spammodule&#39;s API */

static void **PySpam_API;

#define PySpam_System \
 (*(PySpam_System_RETURN (*)PySpam_System_PROTO) PySpam_API[PySpam_System_NUM])

/* Return -1 on error, 0 on success.
 * PyCapsule_Import will set an exception if there&#39;s an error.
 */
static int
import_spam(void)
{
 PySpam_API = (void **)PyCapsule_Import(&quot;spam._C_API&quot;, 0);
 return (PySpam_API != NULL) ? 0 : -1;
}

#endif

#ifdef __cplusplus
}
#endif

#endif /* !defined(Py_SPAMMODULE_H) */
</pre></div>
</div>
All that a client module must do in order to have access to the function
<tt class="xref c c-func docutils literal">PySpam_System()</tt> is to call the function (or rather macro)
<tt class="xref c c-func docutils literal">import_spam()</tt> in its initialization function:
<div class="highlight-c"><div class="highlight"><pre>PyMODINIT_FUNC
initclient(void)
{
 PyObject *m;

m = Py_InitModule(&quot;client&quot;, ClientMethods);
 if (m == NULL)
 return;
 if (import_spam() &lt; 0)
 return;
 /* additional initialization can happen here */
}
</pre></div>
</div>
The main disadvantage of this approach is that the file <tt class="file docutils literal">spammodule.h</tt> is
rather complicated. However, the basic structure is the same for each function
that is exported, so it has to be learned only once.
Finally it should be mentioned that Capsules offer additional functionality,
which is especially useful for memory allocation and deallocation of the pointer
stored in a Capsule. The details are described in the Python/C API Reference
Manual in the section <a class="reference internal" href="../c-api/capsule.html#capsules">Capsules</a> and in the implementation of Capsules (files
<tt class="file docutils literal">Include/pycapsule.h</tt> and <tt class="file docutils literal">Objects/pycapsule.c</tt> in the Python source
code distribution).
Footnotes
<table class="docutils footnote" frame="void" id="id5" rules="none">
<colgroup><col class="label" /><col /></colgroup>
<tbody valign="top">
<tr><td class="label"><a class="fn-backref" href="#id1">[1]</a></td><td>An interface for this function already exists in the standard module <a class="reference internal" href="../library/os.html#module-os" title="os: Miscellaneous operating system interfaces."><tt class="xref py py-mod docutils literal">os</tt></a>
&#8212; it was chosen as a simple and straightforward example.</td></tr>
</tbody>
</table>
<table class="docutils footnote" frame="void" id="id6" rules="none">
<colgroup><col class="label" /><col /></colgroup>
<tbody valign="top">
<tr><td class="label"><a class="fn-backref" href="#id2">[2]</a></td><td>The metaphor of &#8220;borrowing&#8221; a reference is not completely correct: the owner
still has a copy of the reference.</td></tr>
</tbody>
</table>
<table class="docutils footnote" frame="void" id="id7" rules="none">
<colgroup><col class="label" /><col /></colgroup>
<tbody valign="top">
<tr><td class="label"><a class="fn-backref" href="#id3">[3]</a></td><td>Checking that the reference count is at least 1 does not work &#8212; the
reference count itself could be in freed memory and may thus be reused for
another object!</td></tr>
</tbody>
</table>
<table class="docutils footnote" frame="void" id="id8" rules="none">
<colgroup><col class="label" /><col /></colgroup>
<tbody valign="top">
<tr><td class="label"><a class="fn-backref" href="#id4">[4]</a></td><td>These guarantees don&#8217;t hold when you use the &#8220;old&#8221; style calling convention &#8212;
this is still found in much existing code.</td></tr>
</tbody>
</table>
</div>
</div>

</div>
 </div>
 </div>
 <div class="sphinxsidebar">
 <div class="sphinxsidebarwrapper">
 <h3><a href="../contents.html">Table Of Contents</a></h3>
 <ul>
<li><a class="reference internal" href="#">1. Extending Python with C or C++</a><ul>
<li><a class="reference internal" href="#a-simple-example">1.1. A Simple Example</a></li>
<li><a class="reference internal" href="#intermezzo-errors-and-exceptions">1.2. Intermezzo: Errors and Exceptions</a></li>
<li><a class="reference internal" href="#back-to-the-example">1.3. Back to the Example</a></li>
<li><a class="reference internal" href="#the-module-s-method-table-and-initialization-function">1.4. The Module&#8217;s Method Table and Initialization Function</a></li>
<li><a class="reference internal" href="#compilation-and-linkage">1.5. Compilation and Linkage</a></li>
<li><a class="reference internal" href="#calling-python-functions-from-c">1.6. Calling Python Functions from C</a></li>
<li><a class="reference internal" href="#extracting-parameters-in-extension-functions">1.7. Extracting Parameters in Extension Functions</a></li>
<li><a class="reference internal" href="#keyword-parameters-for-extension-functions">1.8. Keyword Parameters for Extension Functions</a></li>
<li><a class="reference internal" href="#building-arbitrary-values">1.9. Building Arbitrary Values</a></li>
<li><a class="reference internal" href="#reference-counts">1.10. Reference Counts</a><ul>
<li><a class="reference internal" href="#reference-counting-in-python">1.10.1. Reference Counting in Python</a></li>
<li><a class="reference internal" href="#ownership-rules">1.10.2. Ownership Rules</a></li>
<li><a class="reference internal" href="#thin-ice">1.10.3. Thin Ice</a></li>
<li><a class="reference internal" href="#null-pointers">1.10.4. NULL Pointers</a></li>
</ul>
</li>
<li><a class="reference internal" href="#writing-extensions-in-c">1.11. Writing Extensions in C++</a></li>
<li><a class="reference internal" href="#providing-a-c-api-for-an-extension-module">1.12. Providing a C API for an Extension Module</a></li>
</ul>
</li>
</ul>

<h4>Previous topic</h4>
 <a href="index.html"
 title="previous chapter">Extending and Embedding the Python Interpreter</a>
 <h4>Next topic</h4>
 <a href="newtypes.html"
 title="next chapter">2. Defining New Types</a>
<h3>This Page</h3>
<ul class="this-page-menu">
 <li><a href="../bugs.html">Report a Bug</a></li>
 <li><a href="../_sources/extending/extending.txt"
 rel="nofollow">Show Source</a></li>
</ul>

<div id="searchbox" style="display: none">
 <h3>Quick search</h3>
 <form class="search" action="../search.html" method="get">
 <input type="text" name="q" />
 <input type="submit" value="Go" />
 <input type="hidden" name="check_keywords" value="yes" />
 <input type="hidden" name="area" value="default" />
 </form>
 
 Enter search terms or a module, class or function name.
 
</div>
<script type="text/javascript">$('#searchbox').show(0);</script>
 </div>
 </div>
 <div class="clearer"></div>
 </div>
 <div class="related">
 <h3>Navigation</h3>
 <ul>
 <li class="right" style="margin-right: 10px">
 <a href="../genindex.html" title="General Index"
 >index</a></li>
 <li class="right" >
 <a href="../py-modindex.html" title="Python Module Index"
 >modules</a> |</li>
 <li class="right" >
 <a href="newtypes.html" title="2. Defining New Types"
 >next</a> |</li>
 <li class="right" >
 <a href="index.html" title="Extending and Embedding the Python Interpreter"
 >previous</a> |</li>
 <li><img src="../_static/py.png" alt=""
 style="vertical-align: middle; margin-top: -1px"/></li>
 <li><a href="http://www.python.org/">Python</a> &raquo;</li>
 <li>
 <a href="../index.html">Python 2.7.5 documentation</a> &raquo;
 </li>

<li><a href="index.html" >Extending and Embedding the Python Interpreter</a> &raquo;</li> 
 </ul>
 </div>
 <div class="footer">
 &copy; <a href="../copyright.html">Copyright</a> 1990-2019, Python Software Foundation.
 
 The Python Software Foundation is a non-profit corporation.
 <a href="http://www.python.org/psf/donations/">Please donate.</a>
 
 Last updated on Jul 03, 2019.
 <a href="../bugs.html">Found a bug</a>?
 
 Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.1.3.
 </div>

</body>
</html>

Cancel