\documentclass{manual}

\title{Python/C API Reference Manual}

\input{boilerplate}

\makeindex % tell \index to actually write the .idx file

\begin{document}

\maketitle

\ifhtml

\chapter*{Front Matter\label{front}}

\fi

\input{copyright}

\begin{abstract}

\noindent

This manual documents the API used by C and \Cpp{} programmers who

want to write extension modules or embed Python. It is a companion to

\citetitle[../ext/ext.html]{Extending and Embedding the Python

Interpreter}, which describes the general principles of extension

writing but does not document the API functions in detail.

\strong{Warning:} The current version of this document is incomplete.

I hope that it is nevertheless useful. I will continue to work on it,

and release new versions from time to time, independent from Python

source code releases.

\end{abstract}

\tableofcontents

% XXX Consider moving all this back to ext.tex and giving api.tex

% XXX a *really* short intro only.

\chapter{Introduction \label{intro}}

The Application Programmer's Interface to Python gives C and

\Cpp{} programmers access to the Python interpreter at a variety of

levels. The API is equally usable from \Cpp{}, but for brevity it is

generally referred to as the Python/C API. There are two

fundamentally different reasons for using the Python/C API. The first

reason is to write \emph{extension modules} for specific purposes;

these are C modules that extend the Python interpreter. This is

probably the most common use. The second reason is to use Python as a

component in a larger application; this technique is generally

referred to as \dfn{embedding} Python in an application.

Writing an extension module is a relatively well-understood process,

where a ``cookbook'' approach works well. There are several tools

that automate the process to some extent. While people have embedded

Python in other applications since its early existence, the process of

embedding Python is less straightforward that writing an extension.

Many API functions are useful independent of whether you're embedding

or extending Python; moreover, most applications that embed Python

will need to provide a custom extension as well, so it's probably a

good idea to become familiar with writing an extension before

attempting to embed Python in a real application.

\section{Include Files \label{includes}}

All function, type and macro definitions needed to use the Python/C

API are included in your code by the following line:

\begin{verbatim}

#include "Python.h"

\end{verbatim}

This implies inclusion of the following standard headers:

\code{<stdio.h>}, \code{<string.h>}, \code{<errno.h>},

\code{<limits.h>}, and \code{<stdlib.h>} (if available).

All user visible names defined by Python.h (except those defined by

the included standard headers) have one of the prefixes \samp{Py} or

\samp{_Py}. Names beginning with \samp{_Py} are for internal use by

the Python implementation and should not be used by extension writers.

Structure member names do not have a reserved prefix.

\strong{Important:} user code should never define names that begin

with \samp{Py} or \samp{_Py}. This confuses the reader, and

jeopardizes the portability of the user code to future Python

versions, which may define additional names beginning with one of

these prefixes.

The header files are typically installed with Python. On \UNIX, these

are located in the directories

\file{\envvar{prefix}/include/python\var{version}/} and

\file{\envvar{exec_prefix}/include/python\var{version}/}, where

\envvar{prefix} and \envvar{exec_prefix} are defined by the

corresponding parameters to Python's \program{configure} script and

\var{version} is \code{sys.version[:3]}. On Windows, the headers are

installed in \file{\envvar{prefix}/include}, where \envvar{prefix} is

the installation directory specified to the installer.

To include the headers, place both directories (if different) on your

compiler's search path for includes. Do \emph{not} place the parent

directories on the search path and then use

\samp{\#include <python\shortversion/Python.h>}; this will break on

multi-platform builds since the platform independent headers under

\envvar{prefix} include the platform specific headers from

\envvar{exec_prefix}.

\section{Objects, Types and Reference Counts \label{objects}}

Most Python/C API functions have one or more arguments as well as a

return value of type \ctype{PyObject*}. This type is a pointer

to an opaque data type representing an arbitrary Python

object. Since all Python object types are treated the same way by the

Python language in most situations (e.g., assignments, scope rules,

and argument passing), it is only fitting that they should be

represented by a single C type. Almost all Python objects live on the

heap: you never declare an automatic or static variable of type

\ctype{PyObject}, only pointer variables of type \ctype{PyObject*} can

be declared. The sole exception are the type objects\obindex{type};

since these must never be deallocated, they are typically static

\ctype{PyTypeObject} objects.

All Python objects (even Python integers) have a \dfn{type} and a

\dfn{reference count}. An object's type determines what kind of object

it is (e.g., an integer, a list, or a user-defined function; there are

many more as explained in the \citetitle[../ref/ref.html]{Python

Reference Manual}). For each of the well-known types there is a macro

to check whether an object is of that type; for instance,

\samp{PyList_Check(\var{a})} is true if (and only if) the object

pointed to by \var{a} is a Python list.

\subsection{Reference Counts \label{refcounts}}

The reference count is important because today's computers have a

finite (and often severely limited) memory size; it counts how many

different places there are that have a reference to an object. Such a

place could be another object, or a global (or static) C variable, or

a local variable in some C function. When an object's reference count

becomes zero, the object is deallocated. If it contains references to

other objects, their reference count is decremented. Those other

objects may be deallocated in turn, if this decrement makes their

reference count become zero, and so on. (There's an obvious problem

with objects that reference each other here; for now, the solution is

``don't do that.'')

Reference counts are always manipulated explicitly. The normal way is

to use the macro \cfunction{Py_INCREF()}\ttindex{Py_INCREF()} to

increment an object's reference count by one, and

\cfunction{Py_DECREF()}\ttindex{Py_DECREF()} to decrement it by

one. The \cfunction{Py_DECREF()} macro is considerably more complex

than the incref one, since it must check whether the reference count

becomes zero and then cause the object's deallocator to be called.

The deallocator is a function pointer contained in the object's type

structure. The type-specific deallocator takes care of decrementing

the reference counts for other objects contained in the object if this

is a compound object type, such as a list, as well as performing any

additional finalization that's needed. There's no chance that the

reference count can overflow; at least as many bits are used to hold

the reference count as there are distinct memory locations in virtual

memory (assuming \code{sizeof(long) >= sizeof(char*)}). Thus, the

reference count increment is a simple operation.

It is not necessary to increment an object's reference count for every

local variable that contains a pointer to an object. In theory, the

object's reference count goes up by one when the variable is made to

point to it and it goes down by one when the variable goes out of

scope. However, these two cancel each other out, so at the end the

reference count hasn't changed. The only real reason to use the

reference count is to prevent the object from being deallocated as

long as our variable is pointing to it. If we know that there is at

least one other reference to the object that lives at least as long as

our variable, there is no need to increment the reference count

temporarily. An important situation where this arises is in objects

that are passed as arguments to C functions in an extension module

that are called from Python; the call mechanism guarantees to hold a

reference to every argument for the duration of the call.

However, a common pitfall is to extract an object from a list and

hold on to it for a while without incrementing its reference count.

Some other operation might conceivably remove the object from the

list, decrementing its reference count and possible deallocating it.

The real danger is that innocent-looking operations may invoke

arbitrary Python code which could do this; there is a code path which

allows control to flow back to the user from a \cfunction{Py_DECREF()},

so almost any operation is potentially dangerous.

A safe approach is to always use the generic operations (functions

whose name begins with \samp{PyObject_}, \samp{PyNumber_},

\samp{PySequence_} or \samp{PyMapping_}). These operations always

increment the reference count of the object they return. This leaves

the caller with the responsibility to call

\cfunction{Py_DECREF()} when they are done with the result; this soon

becomes second nature.

\subsubsection{Reference Count Details \label{refcountDetails}}

The reference count behavior of functions in the Python/C API is best

explained in terms of \emph{ownership of references}. Note that we

talk of owning references, never of owning objects; objects are always

shared! When a function owns a reference, it has to dispose of it

properly --- either by passing ownership on (usually to its caller) or

by calling \cfunction{Py_DECREF()} or \cfunction{Py_XDECREF()}. When

a function passes ownership of a reference on to its caller, the

caller is said to receive a \emph{new} reference. When no ownership

is transferred, the caller is said to \emph{borrow} the reference.

Nothing needs to be done for a borrowed reference.

Conversely, when a calling function passes it a reference to an

object, there are two possibilities: the function \emph{steals} a

reference to the object, or it does not. Few functions steal

references; the two notable exceptions are

\cfunction{PyList_SetItem()}\ttindex{PyList_SetItem()} and

\cfunction{PyTuple_SetItem()}\ttindex{PyTuple_SetItem()}, which

steal a reference to the item (but not to the tuple or list into which

the item is put!). These functions were designed to steal a reference

because of a common idiom for populating a tuple or list with newly

created objects; for example, the code to create the tuple \code{(1,

2, "three")} could look like this (forgetting about error handling for

the moment; a better way to code this is shown below):

\begin{verbatim}

PyObject *t;

t = PyTuple_New(3);

PyTuple_SetItem(t, 0, PyInt_FromLong(1L));

PyTuple_SetItem(t, 1, PyInt_FromLong(2L));

PyTuple_SetItem(t, 2, PyString_FromString("three"));

\end{verbatim}

Incidentally, \cfunction{PyTuple_SetItem()} is the \emph{only} way to

set tuple items; \cfunction{PySequence_SetItem()} and

\cfunction{PyObject_SetItem()} refuse to do this since tuples are an

immutable data type. You should only use

\cfunction{PyTuple_SetItem()} for tuples that you are creating

yourself.

Equivalent code for populating a list can be written using

\cfunction{PyList_New()} and \cfunction{PyList_SetItem()}. Such code

can also use \cfunction{PySequence_SetItem()}; this illustrates the

difference between the two (the extra \cfunction{Py_DECREF()} calls):

\begin{verbatim}

PyObject *l, *x;

l = PyList_New(3);

x = PyInt_FromLong(1L);

PySequence_SetItem(l, 0, x); Py_DECREF(x);

x = PyInt_FromLong(2L);

PySequence_SetItem(l, 1, x); Py_DECREF(x);

x = PyString_FromString("three");

PySequence_SetItem(l, 2, x); Py_DECREF(x);

\end{verbatim}

You might find it strange that the ``recommended'' approach takes more

code. However, in practice, you will rarely use these ways of

creating and populating a tuple or list. There's a generic function,

\cfunction{Py_BuildValue()}, that can create most common objects from

C values, directed by a \dfn{format string}. For example, the

above two blocks of code could be replaced by the following (which

also takes care of the error checking):

\begin{verbatim}

PyObject *t, *l;

t = Py_BuildValue("(iis)", 1, 2, "three");

l = Py_BuildValue("[iis]", 1, 2, "three");

\end{verbatim}

It is much more common to use \cfunction{PyObject_SetItem()} and

friends with items whose references you are only borrowing, like

arguments that were passed in to the function you are writing. In

that case, their behaviour regarding reference counts is much saner,

since you don't have to increment a reference count so you can give a

reference away (``have it be stolen''). For example, this function

sets all items of a list (actually, any mutable sequence) to a given

item:

\begin{verbatim}

int set_all(PyObject *target, PyObject *item)

{

int i, n;

n = PyObject_Length(target);

if (n < 0)

return -1;

for (i = 0; i < n; i++) {

if (PyObject_SetItem(target, i, item) < 0)

return -1;

}

return 0;

}

\end{verbatim}

\ttindex{set_all()}

The situation is slightly different for function return values.

While passing a reference to most functions does not change your

ownership responsibilities for that reference, many functions that

return a referece to an object give you ownership of the reference.

The reason is simple: in many cases, the returned object is created

on the fly, and the reference you get is the only reference to the

object. Therefore, the generic functions that return object

references, like \cfunction{PyObject_GetItem()} and

\cfunction{PySequence_GetItem()}, always return a new reference (i.e.,

the caller becomes the owner of the reference).

It is important to realize that whether you own a reference returned

by a function depends on which function you call only --- \emph{the

plumage} (i.e., the type of the type of the object passed as an

argument to the function) \emph{doesn't enter into it!} Thus, if you

extract an item from a list using \cfunction{PyList_GetItem()}, you

don't own the reference --- but if you obtain the same item from the

same list using \cfunction{PySequence_GetItem()} (which happens to

take exactly the same arguments), you do own a reference to the

returned object.

Here is an example of how you could write a function that computes the

sum of the items in a list of integers; once using

\cfunction{PyList_GetItem()}\ttindex{PyList_GetItem()}, and once using

\cfunction{PySequence_GetItem()}\ttindex{PySequence_GetItem()}.

\begin{verbatim}

long sum_list(PyObject *list)

{

int i, n;

long total = 0;

PyObject *item;

n = PyList_Size(list);

if (n < 0)

return -1; /* Not a list */

for (i = 0; i < n; i++) {

item = PyList_GetItem(list, i); /* Can't fail */

if (!PyInt_Check(item)) continue; /* Skip non-integers */

total += PyInt_AsLong(item);

}

return total;

}

\end{verbatim}

\ttindex{sum_list()}

\begin{verbatim}

long sum_sequence(PyObject *sequence)

{

int i, n;

long total = 0;

PyObject *item;

n = PySequence_Length(sequence);

if (n < 0)

return -1; /* Has no length */

for (i = 0; i < n; i++) {

item = PySequence_GetItem(sequence, i);

if (item == NULL)

return -1; /* Not a sequence, or other failure */

if (PyInt_Check(item))

total += PyInt_AsLong(item);

Py_DECREF(item); /* Discard reference ownership */

}

return total;

}

\end{verbatim}

\ttindex{sum_sequence()}

\subsection{Types \label{types}}

There are few other data types that play a significant role in

the Python/C API; most are simple C types such as \ctype{int},

\ctype{long}, \ctype{double} and \ctype{char*}. A few structure types

are used to describe static tables used to list the functions exported

by a module or the data attributes of a new object type, and another

is used to describe the value of a complex number. These will

be discussed together with the functions that use them.

\section{Exceptions \label{exceptions}}

The Python programmer only needs to deal with exceptions if specific

error handling is required; unhandled exceptions are automatically

propagated to the caller, then to the caller's caller, and so on, until

they reach the top-level interpreter, where they are reported to the

user accompanied by a stack traceback.

For C programmers, however, error checking always has to be explicit.

All functions in the Python/C API can raise exceptions, unless an

explicit claim is made otherwise in a function's documentation. In

general, when a function encounters an error, it sets an exception,

discards any object references that it owns, and returns an

error indicator --- usually \NULL{} or \code{-1}. A few functions

return a Boolean true/false result, with false indicating an error.

Very few functions return no explicit error indicator or have an

ambiguous return value, and require explicit testing for errors with

\cfunction{PyErr_Occurred()}\ttindex{PyErr_Occurred()}.

Exception state is maintained in per-thread storage (this is

equivalent to using global storage in an unthreaded application). A

thread can be in one of two states: an exception has occurred, or not.

The function \cfunction{PyErr_Occurred()} can be used to check for

this: it returns a borrowed reference to the exception type object

when an exception has occurred, and \NULL{} otherwise. There are a

number of functions to set the exception state:

\cfunction{PyErr_SetString()}\ttindex{PyErr_SetString()} is the most

common (though not the most general) function to set the exception

state, and \cfunction{PyErr_Clear()}\ttindex{PyErr_Clear()} clears the

exception state.

The full exception state consists of three objects (all of which can

be \NULL{}): the exception type, the corresponding exception

value, and the traceback. These have the same meanings as the Python

\withsubitem{(in module sys)}{

\ttindex{exc_type}\ttindex{exc_value}\ttindex{exc_traceback}}

objects \code{sys.exc_type}, \code{sys.exc_value}, and

\code{sys.exc_traceback}; however, they are not the same: the Python

objects represent the last exception being handled by a Python

\keyword{try} \ldots\ \keyword{except} statement, while the C level

exception state only exists while an exception is being passed on

between C functions until it reaches the Python bytecode interpreter's

main loop, which takes care of transferring it to \code{sys.exc_type}

and friends.

Note that starting with Python 1.5, the preferred, thread-safe way to

access the exception state from Python code is to call the function

\withsubitem{(in module sys)}{\ttindex{exc_info()}}

\function{sys.exc_info()}, which returns the per-thread exception state

for Python code. Also, the semantics of both ways to access the

exception state have changed so that a function which catches an

exception will save and restore its thread's exception state so as to

preserve the exception state of its caller. This prevents common bugs

in exception handling code caused by an innocent-looking function

overwriting the exception being handled; it also reduces the often

unwanted lifetime extension for objects that are referenced by the

stack frames in the traceback.

As a general principle, a function that calls another function to

perform some task should check whether the called function raised an

exception, and if so, pass the exception state on to its caller. It

should discard any object references that it owns, and return an

error indicator, but it should \emph{not} set another exception ---

that would overwrite the exception that was just raised, and lose

important information about the exact cause of the error.

A simple example of detecting exceptions and passing them on is shown

in the \cfunction{sum_sequence()}\ttindex{sum_sequence()} example

above. It so happens that that example doesn't need to clean up any

owned references when it detects an error. The following example

function shows some error cleanup. First, to remind you why you like

Python, we show the equivalent Python code:

\begin{verbatim}

def incr_item(dict, key):

try:

item = dict[key]

except KeyError:

item = 0

return item + 1

\end{verbatim}

\ttindex{incr_item()}

Here is the corresponding C code, in all its glory:

\begin{verbatim}

int incr_item(PyObject *dict, PyObject *key)

{

/* Objects all initialized to NULL for Py_XDECREF */

PyObject *item = NULL, *const_one = NULL, *incremented_item = NULL;

int rv = -1; /* Return value initialized to -1 (failure) */

item = PyObject_GetItem(dict, key);

if (item == NULL) {

/* Handle KeyError only: */

if (!PyErr_ExceptionMatches(PyExc_KeyError)) goto error;

/* Clear the error and use zero: */

PyErr_Clear();

item = PyInt_FromLong(0L);

if (item == NULL) goto error;

}

const_one = PyInt_FromLong(1L);

if (const_one == NULL) goto error;

incremented_item = PyNumber_Add(item, const_one);

if (incremented_item == NULL) goto error;

if (PyObject_SetItem(dict, key, incremented_item) < 0) goto error;

rv = 0; /* Success */

/* Continue with cleanup code */

error:

/* Cleanup code, shared by success and failure path */

/* Use Py_XDECREF() to ignore NULL references */

Py_XDECREF(item);

Py_XDECREF(const_one);

Py_XDECREF(incremented_item);

return rv; /* -1 for error, 0 for success */

}

\end{verbatim}

\ttindex{incr_item()}

This example represents an endorsed use of the \keyword{goto} statement

in C! It illustrates the use of

\cfunction{PyErr_ExceptionMatches()}\ttindex{PyErr_ExceptionMatches()} and

\cfunction{PyErr_Clear()}\ttindex{PyErr_Clear()} to

handle specific exceptions, and the use of

\cfunction{Py_XDECREF()}\ttindex{Py_XDECREF()} to

dispose of owned references that may be \NULL{} (note the

\character{X} in the name; \cfunction{Py_DECREF()} would crash when

confronted with a \NULL{} reference). It is important that the

variables used to hold owned references are initialized to \NULL{} for

this to work; likewise, the proposed return value is initialized to

\code{-1} (failure) and only set to success after the final call made

is successful.

\section{Embedding Python \label{embedding}}

The one important task that only embedders (as opposed to extension

writers) of the Python interpreter have to worry about is the

initialization, and possibly the finalization, of the Python

interpreter. Most functionality of the interpreter can only be used

after the interpreter has been initialized.

The basic initialization function is

\cfunction{Py_Initialize()}\ttindex{Py_Initialize()}.

This initializes the table of loaded modules, and creates the

fundamental modules \module{__builtin__}\refbimodindex{__builtin__},

\module{__main__}\refbimodindex{__main__} and

\module{sys}\refbimodindex{sys}. It also initializes the module

search path (\code{sys.path}).%

\indexiii{module}{search}{path}

\withsubitem{(in module sys)}{\ttindex{path}}

\cfunction{Py_Initialize()} does not set the ``script argument list''

(\code{sys.argv}). If this variable is needed by Python code that

will be executed later, it must be set explicitly with a call to

\code{PySys_SetArgv(\var{argc},

\var{argv})}\ttindex{PySys_SetArgv()} subsequent to the call to

\cfunction{Py_Initialize()}.

On most systems (in particular, on \UNIX{} and Windows, although the

details are slightly different),

\cfunction{Py_Initialize()} calculates the module search path based

upon its best guess for the location of the standard Python

interpreter executable, assuming that the Python library is found in a

fixed location relative to the Python interpreter executable. In

particular, it looks for a directory named

\file{lib/python\shortversion} relative to the parent directory where

the executable named \file{python} is found on the shell command

search path (the environment variable \envvar{PATH}).

For instance, if the Python executable is found in

\file{/usr/local/bin/python}, it will assume that the libraries are in

\file{/usr/local/lib/python\shortversion}. (In fact, this particular path

is also the ``fallback'' location, used when no executable file named

\file{python} is found along \envvar{PATH}.) The user can override

this behavior by setting the environment variable \envvar{PYTHONHOME},

or insert additional directories in front of the standard path by

setting \envvar{PYTHONPATH}.

The embedding application can steer the search by calling

\code{Py_SetProgramName(\var{file})}\ttindex{Py_SetProgramName()} \emph{before} calling

\cfunction{Py_Initialize()}. Note that \envvar{PYTHONHOME} still

overrides this and \envvar{PYTHONPATH} is still inserted in front of

the standard path. An application that requires total control has to

provide its own implementation of

\cfunction{Py_GetPath()}\ttindex{Py_GetPath()},

\cfunction{Py_GetPrefix()}\ttindex{Py_GetPrefix()},

\cfunction{Py_GetExecPrefix()}\ttindex{Py_GetExecPrefix()}, and

\cfunction{Py_GetProgramFullPath()}\ttindex{Py_GetProgramFullPath()} (all

defined in \file{Modules/getpath.c}).

Sometimes, it is desirable to ``uninitialize'' Python. For instance,

the application may want to start over (make another call to

\cfunction{Py_Initialize()}) or the application is simply done with its

use of Python and wants to free all memory allocated by Python. This

can be accomplished by calling \cfunction{Py_Finalize()}. The function

\cfunction{Py_IsInitialized()}\ttindex{Py_IsInitialized()} returns

true if Python is currently in the initialized state. More

information about these functions is given in a later chapter.

\chapter{The Very High Level Layer \label{veryhigh}}

The functions in this chapter will let you execute Python source code

given in a file or a buffer, but they will not let you interact in a

more detailed way with the interpreter.

Several of these functions accept a start symbol from the grammar as a

parameter. The available start symbols are \constant{Py_eval_input},

\constant{Py_file_input}, and \constant{Py_single_input}. These are

described following the functions which accept them as parameters.

Note also that several of these functions take \ctype{FILE*}

parameters. On particular issue which needs to be handled carefully

is that the \ctype{FILE} structure for different C libraries can be

different and incompatible. Under Windows (at least), it is possible

for dynamically linked extensions to actually use different libraries,

so care should be taken that \ctype{FILE*} parameters are only passed

to these functions if it is certain that they were created by the same

library that the Python runtime is using.

\begin{cfuncdesc}{int}{PyRun_AnyFile}{FILE *fp, char *filename}

If \var{fp} refers to a file associated with an interactive device

(console or terminal input or \UNIX{} pseudo-terminal), return the

value of \cfunction{PyRun_InteractiveLoop()}, otherwise return the

result of \cfunction{PyRun_SimpleFile()}. If \var{filename} is

\NULL{}, this function uses \code{"???"} as the filename.

\end{cfuncdesc}

\begin{cfuncdesc}{int}{PyRun_SimpleString}{char *command}

Executes the Python source code from \var{command} in the

\module{__main__} module. If \module{__main__} does not already

exist, it is created. Returns \code{0} on success or \code{-1} if

an exception was raised. If there was an error, there is no way to

get the exception information.

\end{cfuncdesc}

\begin{cfuncdesc}{int}{PyRun_SimpleFile}{FILE *fp, char *filename}

Similar to \cfunction{PyRun_SimpleString()}, but the Python source

code is read from \var{fp} instead of an in-memory string.

\var{filename} should be the name of the file.

\end{cfuncdesc}

\begin{cfuncdesc}{int}{PyRun_InteractiveOne}{FILE *fp, char *filename}

Read and execute a single statement from a file associated with an

interactive device. If \var{filename} is \NULL, \code{"???"} is

used instead. The user will be prompted using \code{sys.ps1} and

\code{sys.ps2}. Returns \code{0} when the input was executed

successfully, \code{-1} if there was an exception, or an error code

from the \file{errcode.h} include file distributed as part of Python

in case of a parse error. (Note that \file{errcode.h} is not

included by \file{Python.h}, so must be included specifically if

needed.)

\end{cfuncdesc}

\begin{cfuncdesc}{int}{PyRun_InteractiveLoop}{FILE *fp, char *filename}

Read and execute statements from a file associated with an

interactive device until \EOF{} is reached. If \var{filename} is

\NULL, \code{"???"} is used instead. The user will be prompted

using \code{sys.ps1} and \code{sys.ps2}. Returns \code{0} at \EOF.

\end{cfuncdesc}

\begin{cfuncdesc}{struct _node*}{PyParser_SimpleParseString}{char *str,

int start}

Parse Python source code from \var{str} using the start token

\var{start}. The result can be used to create a code object which

can be evaluated efficiently. This is useful if a code fragment

must be evaluated many times.

\end{cfuncdesc}

\begin{cfuncdesc}{struct _node*}{PyParser_SimpleParseFile}{FILE *fp,

char *filename, int start}

Similar to \cfunction{PyParser_SimpleParseString()}, but the Python

source code is read from \var{fp} instead of an in-memory string.

\var{filename} should be the name of the file.

\end{cfuncdesc}

\begin{cfuncdesc}{PyObject*}{PyRun_String}{char *str, int start,

PyObject *globals,

PyObject *locals}

Execute Python source code from \var{str} in the context specified

by the dictionaries \var{globals} and \var{locals}. The parameter

\var{start} specifies the start token that should be used to parse

the source code.

Returns the result of executing the code as a Python object, or

\NULL{} if an exception was raised.

\end{cfuncdesc}

\begin{cfuncdesc}{PyObject*}{PyRun_File}{FILE *fp, char *filename,

int start, PyObject *globals,

PyObject *locals}

Similar to \cfunction{PyRun_String()}, but the Python source code is

read from \var{fp} instead of an in-memory string.

\var{filename} should be the name of the file.

\end{cfuncdesc}

\begin{cfuncdesc}{PyObject*}{Py_CompileString}{char *str, char *filename,

int start}

Parse and compile the Python source code in \var{str}, returning the

resulting code object. The start token is given by \var{start};

this can be used to constrain the code which can be compiled and should

be \constant{Py_eval_input}, \constant{Py_file_input}, or

\constant{Py_single_input}. The filename specified by

\var{filename} is used to construct the code object and may appear

in tracebacks or \exception{SyntaxError} exception messages. This

returns \NULL{} if the code cannot be parsed or compiled.

\end{cfuncdesc}

\begin{cvardesc}{int}{Py_eval_input}

The start symbol from the Python grammar for isolated expressions;

for use with \cfunction{Py_CompileString()}\ttindex{Py_CompileString()}.

\end{cvardesc}

\begin{cvardesc}{int}{Py_file_input}

The start symbol from the Python grammar for sequences of statements

as read from a file or other source; for use with

\cfunction{Py_CompileString()}\ttindex{Py_CompileString()}. This is

the symbol to use when compiling arbitrarily long Python source code.

\end{cvardesc}

\begin{cvardesc}{int}{Py_single_input}

The start symbol from the Python grammar for a single statement; for

use with \cfunction{Py_CompileString()}\ttindex{Py_CompileString()}.

This is the symbol used for the interactive interpreter loop.

\end{cvardesc}

\chapter{Reference Counting \label{countingRefs}}

The macros in this section are used for managing reference counts

of Python objects.

\begin{cfuncdesc}{void}{Py_INCREF}{PyObject *o}

Increment the reference count for object \var{o}. The object must

not be \NULL{}; if you aren't sure that it isn't \NULL{}, use

\cfunction{Py_XINCREF()}.

\end{cfuncdesc}

\begin{cfuncdesc}{void}{Py_XINCREF}{PyObject *o}

Increment the reference count for object \var{o}. The object may be

\NULL{}, in which case the macro has no effect.

\end{cfuncdesc}

\begin{cfuncdesc}{void}{Py_DECREF}{PyObject *o}

Decrement the reference count for object \var{o}. The object must

not be \NULL{}; if you aren't sure that it isn't \NULL{}, use

\cfunction{Py_XDECREF()}. If the reference count reaches zero, the

object's type's deallocation function (which must not be \NULL{}) is

invoked.

\strong{Warning:} The deallocation function can cause arbitrary Python

code to be invoked (e.g. when a class instance with a

\method{__del__()} method is deallocated). While exceptions in such

code are not propagated, the executed code has free access to all

Python global variables. This means that any object that is reachable

from a global variable should be in a consistent state before

\cfunction{Py_DECREF()} is invoked. For example, code to delete an

object from a list should copy a reference to the deleted object in a

temporary variable, update the list data structure, and then call

\cfunction{Py_DECREF()} for the temporary variable.

\end{cfuncdesc}

\begin{cfuncdesc}{void}{Py_XDECREF}{PyObject *o}

Decrement the reference count for object \var{o}. The object may be

\NULL{}, in which case the macro has no effect; otherwise the effect

is the same as for \cfunction{Py_DECREF()}, and the same warning

applies.

\end{cfuncdesc}

The following functions or macros are only for use within the

interpreter core: \cfunction{_Py_Dealloc()},

\cfunction{_Py_ForgetReference()}, \cfunction{_Py_NewReference()}, as

well as the global variable \cdata{_Py_RefTotal}.

\chapter{Exception Handling \label{exceptionHandling}}

The functions described in this chapter will let you handle and raise Python

exceptions. It is important to understand some of the basics of

Python exception handling. It works somewhat like the

\UNIX{} \cdata{errno} variable: there is a global indicator (per

thread) of the last error that occurred. Most functions don't clear

this on success, but will set it to indicate the cause of the error on

failure. Most functions also return an error indicator, usually

\NULL{} if they are supposed to return a pointer, or \code{-1} if they

return an integer (exception: the \cfunction{PyArg_Parse*()} functions

return \code{1} for success and \code{0} for failure). When a

function must fail because some function it called failed, it

generally doesn't set the error indicator; the function it called

already set it.

The error indicator consists of three Python objects corresponding to

\withsubitem{(in module sys)}{

\ttindex{exc_type}\ttindex{exc_value}\ttindex{exc_traceback}}

the Python variables \code{sys.exc_type}, \code{sys.exc_value} and

\code{sys.exc_traceback}. API functions exist to interact with the

error indicator in various ways. There is a separate error indicator

for each thread.

% XXX Order of these should be more thoughtful.

% Either alphabetical or some kind of structure.

\begin{cfuncdesc}{void}{PyErr_Print}{}

Print a standard traceback to \code{sys.stderr} and clear the error

indicator. Call this function only when the error indicator is set.

(Otherwise it will cause a fatal error!)

\end{cfuncdesc}

\begin{cfuncdesc}{PyObject*}{PyErr_Occurred}{}

Test whether the error indicator is set. If set, return the exception

\emph{type} (the first argument to the last call to one of the

\cfunction{PyErr_Set*()} functions or to \cfunction{PyErr_Restore()}). If

not set, return \NULL{}. You do not own a reference to the return

value, so you do not need to \cfunction{Py_DECREF()} it.

\strong{Note:} Do not compare the return value to a specific

exception; use \cfunction{PyErr_ExceptionMatches()} instead, shown

below. (The comparison could easily fail since the exception may be

an instance instead of a class, in the case of a class exception, or

it may the a subclass of the expected exception.)

\end{cfuncdesc}

\begin{cfuncdesc}{int}{PyErr_ExceptionMatches}{PyObject *exc}

Equivalent to

\samp{PyErr_GivenExceptionMatches(PyErr_Occurred(), \var{exc})}.

This should only be called when an exception is actually set; a memory

access violation will occur if no exception has been raised.

\end{cfuncdesc}

\begin{cfuncdesc}{int}{PyErr_GivenExceptionMatches}{PyObject *given, PyObject *exc}

Return true if the \var{given} exception matches the exception in

\var{exc}. If \var{exc} is a class object, this also returns true

when \var{given} is an instance of a subclass. If \var{exc} is a tuple, all

exceptions in the tuple (and recursively in subtuples) are searched

for a match. If \var{given} is \NULL, a memory access violation will

occur.

\end{cfuncdesc}

\begin{cfuncdesc}{void}{PyErr_NormalizeException}{PyObject**exc, PyObject**val, PyObject**tb}

Under certain circumstances, the values returned by

\cfunction{PyErr_Fetch()} below can be ``unnormalized'', meaning that

\code{*\var{exc}} is a class object but \code{*\var{val}} is not an

instance of the same class. This function can be used to instantiate

the class in that case. If the values are already normalized, nothing

happens. The delayed normalization is implemented to improve

performance.

\end{cfuncdesc}

\begin{cfuncdesc}{void}{PyErr_Clear}{}

Clear the error indicator. If the error indicator is not set, there

is no effect.

\end{cfuncdesc}

\begin{cfuncdesc}{void}{PyErr_Fetch}{PyObject **ptype, PyObject **pvalue,

PyObject **ptraceback}

Retrieve the error indicator into three variables whose addresses are

passed. If the error indicator is not set, set all three variables to

\NULL{}. If it is set, it will be cleared and you own a reference to

each object retrieved. The value and traceback object may be

\NULL{} even when the type object is not. \strong{Note:} This

function is normally only used by code that needs to handle exceptions

or by code that needs to save and restore the error indicator

temporarily.

\end{cfuncdesc}

\begin{cfuncdesc}{void}{PyErr_Restore}{PyObject *type, PyObject *value,

PyObject *traceback}

Set the error indicator from the three objects. If the error

indicator is already set, it is cleared first. If the objects are

\NULL{}, the error indicator is cleared. Do not pass a \NULL{} type

and non-\NULL{} value or traceback. The exception type should be a

string or class; if it is a class, the value should be an instance of

that class. Do not pass an invalid exception type or value.

(Violating these rules will cause subtle problems later.) This call

takes away a reference to each object, i.e.\ you must own a reference

to each object before the call and after the call you no longer own

these references. (If you don't understand this, don't use this

function. I warned you.) \strong{Note:} This function is normally

only used by code that needs to save and restore the error indicator

temporarily.

\end{cfuncdesc}

\begin{cfuncdesc}{void}{PyErr_SetString}{PyObject *type, char *message}

This is the most common way to set the error indicator. The first

argument specifies the exception type; it is normally one of the

standard exceptions, e.g. \cdata{PyExc_RuntimeError}. You need not

increment its reference count. The second argument is an error

message; it is converted to a string object.

\end{cfuncdesc}

\begin{cfuncdesc}{void}{PyErr_SetObject}{PyObject *type, PyObject *value}

This function is similar to \cfunction{PyErr_SetString()} but lets you

specify an arbitrary Python object for the ``value'' of the exception.

You need not increment its reference count.

\end{cfuncdesc}

\begin{cfuncdesc}{PyObject*}{PyErr_Format}{PyObject *exception,

const char *format, \moreargs}

This function sets the error indicator.

\var{exception} should be a Python object.

\var{fmt} should be a string, containing format codes, similar to

\cfunction{printf}. The \code{width.precision} before a format code

is parsed, but the width part is ignored.

\begin{tableii}{c|l}{character}{Character}{Meaning}

\lineii{c}{Character, as an \ctype{int} parameter}

\lineii{d}{Number in decimal, as an \ctype{int} parameter}

\lineii{x}{Number in hexadecimal, as an \ctype{int} parameter}

\lineii{x}{A string, as a \ctype{char *} parameter}

\end{tableii}

An unrecognized format character causes all the rest of

the format string to be copied as-is to the result string,

and any extra arguments discarded.

A new reference is returned, which is owned by the caller.

\end{cfuncdesc}

\begin{cfuncdesc}{void}{PyErr_SetNone}{PyObject *type}

This is a shorthand for \samp{PyErr_SetObject(\var{type}, Py_None)}.

\end{cfuncdesc}

\begin{cfuncdesc}{int}{PyErr_BadArgument}{}

This is a shorthand for \samp{PyErr_SetString(PyExc_TypeError,

\var{message})}, where \var{message} indicates that a built-in operation

was invoked with an illegal argument. It is mostly for internal use.

\end{cfuncdesc}

\begin{cfuncdesc}{PyObject*}{PyErr_NoMemory}{}

This is a shorthand for \samp{PyErr_SetNone(PyExc_MemoryError)}; it

returns \NULL{} so an object allocation function can write

\samp{return PyErr_NoMemory();} when it runs out of memory.

\end{cfuncdesc}

\begin{cfuncdesc}{PyObject*}{PyErr_SetFromErrno}{PyObject *type}

This is a convenience function to raise an exception when a C library

function has returned an error and set the C variable \cdata{errno}.

It constructs a tuple object whose first item is the integer

\cdata{errno} value and whose second item is the corresponding error

message (gotten from \cfunction{strerror()}\ttindex{strerror()}), and

then calls

\samp{PyErr_SetObject(\var{type}, \var{object})}. On \UNIX{}, when

the \cdata{errno} value is \constant{EINTR}, indicating an interrupted

system call, this calls \cfunction{PyErr_CheckSignals()}, and if that set

the error indicator, leaves it set to that. The function always

returns \NULL{}, so a wrapper function around a system call can write

\samp{return PyErr_SetFromErrno();} when the system call returns an

error.

\end{cfuncdesc}

\begin{cfuncdesc}{void}{PyErr_BadInternalCall}{}

This is a shorthand for \samp{PyErr_SetString(PyExc_TypeError,

\var{message})}, where \var{message} indicates that an internal

operation (e.g. a Python/C API function) was invoked with an illegal

argument. It is mostly for internal use.

\end{cfuncdesc}

\begin{cfuncdesc}{int}{PyErr_CheckSignals}{}

This function interacts with Python's signal handling. It checks

whether a signal has been sent to the processes and if so, invokes the

corresponding signal handler. If the

\module{signal}\refbimodindex{signal} module is supported, this can

invoke a signal handler written in Python. In all cases, the default

effect for \constant{SIGINT}\ttindex{SIGINT} is to raise the

\withsubitem{(built-in exception)}{\ttindex{KeyboardInterrupt}}

\exception{KeyboardInterrupt} exception. If an exception is raised the

error indicator is set and the function returns \code{1}; otherwise

the function returns \code{0}. The error indicator may or may not be

cleared if it was previously set.

\end{cfuncdesc}

\begin{cfuncdesc}{void}{PyErr_SetInterrupt}{}

This function is obsolete. It simulates the effect of a

\constant{SIGINT}\ttindex{SIGINT} signal arriving --- the next time

\cfunction{PyErr_CheckSignals()} is called,

\withsubitem{(built-in exception)}{\ttindex{KeyboardInterrupt}}

\exception{KeyboardInterrupt} will be raised.

It may be called without holding the interpreter lock.

\end{cfuncdesc}

\begin{cfuncdesc}{PyObject*}{PyErr_NewException}{char *name,

PyObject *base,

PyObject *dict}

This utility function creates and returns a new exception object. The

\var{name} argument must be the name of the new exception, a C string

of the form \code{module.class}. The \var{base} and

\var{dict} arguments are normally \NULL{}. This creates a

class object derived from the root for all exceptions, the built-in

name \exception{Exception} (accessible in C as

\cdata{PyExc_Exception}). The \member{__module__} attribute of the

new class is set to the first part (up to the last dot) of the

\var{name} argument, and the class name is set to the last part (after

the last dot). The \var{base} argument can be used to specify an

alternate base class. The \var{dict} argument can be used to specify

a dictionary of class variables and methods.

\end{cfuncdesc}

\begin{cfuncdesc}{void}{PyErr_WriteUnraisable}{PyObject *obj}

This utility function prints a warning message to \var{sys.stderr}

when an exception has been set but it is impossible for the

interpreter to actually raise the exception. It is used, for example,

when an exception occurs in an \member{__del__} method.

The function is called with a single argument \var{obj} that

identifies where the context in which the unraisable exception

occurred. The repr of \var{obj} will be printed in the warning

message.

\end{cfuncdesc}

\section{Standard Exceptions \label{standardExceptions}}

All standard Python exceptions are available as global variables whose

names are \samp{PyExc_} followed by the Python exception name. These

have the type \ctype{PyObject*}; they are all class objects. For