Parsing arguments and building values¶

These functions are useful when creating your own extensions functions andmethods. Additional information and examples are available inExtending and Embedding the Python Interpreter.

The first three of these functions described,PyArg_ParseTuple(),PyArg_ParseTupleAndKeywords(), andPyArg_Parse(), all useformat strings which are used to tell the function about the expectedarguments. The format strings use the same syntax for each of thesefunctions.

A format string consists of zero or more “format units.” A format unitdescribes one Python object; it is usually a single character or aparenthesized sequence of format units. With a few exceptions, a format unitthat is not a parenthesized sequence normally corresponds to a single addressargument to these functions. In the following description, the quoted form isthe format unit; the entry in (round) parentheses is the Python object typethat matches the format unit; and the entry in [square] brackets is the typeof the C variable(s) whose address should be passed.

These formats allow accessing an object as a contiguous chunk of memory.You don’t have to provide raw storage for the returned unicode or bytesarea. Also, you won’t have to release any memory yourself, except with thees,es#,et andet# formats.

s (string or Unicode) [const char *]

Convert a Python string or Unicode object to a C pointer to a characterstring. You must not provide storage for the string itself; a pointer toan existing string is stored into the character pointer variable whoseaddress you pass. The C string is NUL-terminated. The Python string mustnot contain embedded NUL bytes; if it does, aTypeError exception israised. Unicode objects are converted to C strings using the defaultencoding. If this conversion fails, aUnicodeError is raised.

s# (string, Unicode or any read buffer compatible object) [const char *, int (orPy_ssize_t, see below)]

This variant ons stores into two C variables, the first one a pointerto a character string, the second one its length. In this case the Pythonstring may contain embedded null bytes. Unicode objects pass back apointer to the default encoded string version of the object if such aconversion is possible. All other read-buffer compatible objects pass backa reference to the raw internal data representation.

Starting with Python 2.5 the type of the length argument can be controlledby defining the macroPY_SSIZE_T_CLEAN before includingPython.h. If the macro is defined, length is aPy_ssize_trather than an int.

s* (string, Unicode, or any buffer compatible object) [Py_buffer]

Similar tos#, this code fills a Py_buffer structure provided by thecaller. The buffer gets locked, so that the caller can subsequently usethe buffer even inside aPy_BEGIN_ALLOW_THREADS block; the caller isresponsible for callingPyBuffer_Release with the structure after ithas processed the data.

New in version 2.6.

z (string, Unicode orNone) [const char *]

Likes, but the Python object may also beNone, in which case the Cpointer is set toNULL.

z# (string, Unicode,None or any read buffer compatible object) [const char *, int]

This is tos# asz is tos.

z* (string, Unicode,None or any buffer compatible object) [Py_buffer]

This is tos* asz is tos.

New in version 2.6.

u (Unicode) [Py_UNICODE *]

Convert a Python Unicode object to a C pointer to a NUL-terminated bufferof 16-bit Unicode (UTF-16) data. As withs, there is no need toprovide storage for the Unicode data buffer; a pointer to the existingUnicode data is stored into thePy_UNICODE pointer variable whoseaddress you pass.

u# (Unicode) [Py_UNICODE *, int]

This variant onu stores into two C variables, the first one a pointerto a Unicode data buffer, the second one its length. Non-Unicode objectsare handled by interpreting their read-buffer pointer as pointer to aPy_UNICODE array.

es (string, Unicode or character buffer compatible object) [const char *encoding, char **buffer]

This variant ons is used for encoding Unicode and objects convertibleto Unicode into a character buffer. It only works for encoded data withoutembedded NUL bytes.

This format requires two arguments. The first is only used as input, andmust be aconstchar* which points to the name of an encoding asa NUL-terminated string, orNULL, in which case the default encoding isused. An exception is raised if the named encoding is not known to Python.The second argument must be achar**; the value of the pointerit references will be set to a buffer with the contents of the argumenttext. The text will be encoded in the encoding specified by the firstargument.

PyArg_ParseTuple() will allocate a buffer of the needed size, copythe encoded data into this buffer and adjust*buffer to reference thenewly allocated storage. The caller is responsible for callingPyMem_Free() to free the allocated buffer after use.

et (string, Unicode or character buffer compatible object) [const char *encoding, char **buffer]

Same ases except that 8-bit string objects are passed through withoutrecoding them. Instead, the implementation assumes that the string objectuses the encoding passed in as parameter.

es# (string, Unicode or character buffer compatible object) [const char *encoding, char **buffer, int *buffer_length]

This variant ons# is used for encoding Unicode and objects convertibleto Unicode into a character buffer. Unlike thees format, this variantallows input data which contains NUL characters.

It requires three arguments. The first is only used as input, and must beaconstchar* which points to the name of an encoding as aNUL-terminated string, orNULL, in which case the default encoding isused. An exception is raised if the named encoding is not known to Python.The second argument must be achar**; the value of the pointerit references will be set to a buffer with the contents of the argumenttext. The text will be encoded in the encoding specified by the firstargument. The third argument must be a pointer to an integer; thereferenced integer will be set to the number of bytes in the output buffer.

There are two modes of operation:

If*buffer points aNULL pointer, the function will allocate a bufferof the needed size, copy the encoded data into this buffer and set*buffer to reference the newly allocated storage. The caller isresponsible for callingPyMem_Free() to free the allocated bufferafter usage.

If*buffer points to a non-NULL pointer (an already allocated buffer),PyArg_ParseTuple() will use this location as the buffer andinterpret the initial value of*buffer_length as the buffer size. Itwill then copy the encoded data into the buffer and NUL-terminate it. Ifthe buffer is not large enough, aTypeError will be set.Note: starting from Python 3.6 aValueError will be set.

In both cases,*buffer_length is set to the length of the encoded datawithout the trailing NUL byte.

et# (string, Unicode or character buffer compatible object) [const char *encoding, char **buffer, int *buffer_length]

Same ases# except that string objects are passed through withoutrecoding them. Instead, the implementation assumes that the string objectuses the encoding passed in as parameter.

b (integer) [unsigned char]

Convert a nonnegative Python integer to an unsigned tiny int, stored in a Cunsignedchar.

B (integer) [unsigned char]

Convert a Python integer to a tiny int without overflow checking, stored ina Cunsignedchar.

New in version 2.3.

h (integer) [short int]

Convert a Python integer to a Cshortint.

H (integer) [unsigned short int]

Convert a Python integer to a Cunsignedshortint, withoutoverflow checking.

New in version 2.3.

i (integer) [int]

Convert a Python integer to a plain Cint.

I (integer) [unsigned int]

Convert a Python integer to a Cunsignedint, without overflowchecking.

New in version 2.3.

l (integer) [long int]

Convert a Python integer to a Clongint.

k (integer) [unsigned long]

Convert a Python integer or long integer to a Cunsignedlongwithout overflow checking.

New in version 2.3.

L (integer) [PY_LONG_LONG]

Convert a Python integer to a Clonglong. This format is onlyavailable on platforms that supportlonglong (or_int64on Windows).

K (integer) [unsigned PY_LONG_LONG]

Convert a Python integer or long integer to a Cunsignedlonglongwithout overflow checking. This format is only available on platforms thatsupportunsignedlonglong (orunsigned_int64 onWindows).

New in version 2.3.

n (integer) [Py_ssize_t]

Convert a Python integer or long integer to a CPy_ssize_t.

New in version 2.5.

c (string of length 1) [char]

Convert a Python character, represented as a string of length 1, to a Cchar.

f (float) [float]

Convert a Python floating point number to a Cfloat.

d (float) [double]

Convert a Python floating point number to a Cdouble.

D (complex) [Py_complex]

Convert a Python complex number to a CPy_complex structure.

O (object) [PyObject *]

Store a Python object (without any conversion) in a C object pointer. TheC program thus receives the actual object that was passed. The object’sreference count is not increased. The pointer stored is notNULL.

O! (object) [typeobject, PyObject *]

Store a Python object in a C object pointer. This is similar toO, buttakes two C arguments: the first is the address of a Python type object,the second is the address of the C variable (of typePyObject*)into which the object pointer is stored. If the Python object does nothave the required type,TypeError is raised.

O& (object) [converter,anything]

Convert a Python object to a C variable through aconverter function.This takes two arguments: the first is a function, the second is theaddress of a C variable (of arbitrary type), converted tovoid*.Theconverter function in turn is called as follows:

status=converter(object,address);

whereobject is the Python object to be converted andaddress is thevoid* argument that was passed to thePyArg_Parse*()function. The returnedstatus should be1 for a successfulconversion and0 if the conversion has failed. When the conversionfails, theconverter function should raise an exception and leave thecontent ofaddress unmodified.

S (string) [PyStringObject *]

LikeO but requires that the Python object is a string object. RaisesTypeError if the object is not a string object. The C variable mayalso be declared asPyObject*.

U (Unicode string) [PyUnicodeObject *]

LikeO but requires that the Python object is a Unicode object. RaisesTypeError if the object is not a Unicode object. The C variable mayalso be declared asPyObject*.

t# (read-only character buffer) [char *, int]

Likes#, but accepts any object which implements the read-only bufferinterface. Thechar* variable is set to point to the first byteof the buffer, and theint is set to the length of the buffer.Only single-segment buffer objects are accepted;TypeError is raisedfor all others.

w (read-write character buffer) [char *]

Similar tos, but accepts any object which implements the read-writebuffer interface. The caller must determine the length of the buffer byother means, or usew# instead. Only single-segment buffer objects areaccepted;TypeError is raised for all others.

w# (read-write character buffer) [char *, Py_ssize_t]

Likes#, but accepts any object which implements the read-write bufferinterface. Thechar* variable is set to point to the first byteof the buffer, and thePy_ssize_t is set to the length of thebuffer. Only single-segment buffer objects are accepted;TypeErroris raised for all others.

w* (read-write byte-oriented buffer) [Py_buffer]

This is tow whats* is tos.

New in version 2.6.

(items) (tuple) [matching-items]

The object must be a Python sequence whose length is the number of formatunits initems. The C arguments must correspond to the individual formatunits initems. Format units for sequences may be nested.

Note

Prior to Python version 1.5.2, this format specifier only accepted atuple containing the individual parameters, not an arbitrary sequence.Code which previously causedTypeError to be raised here may nowproceed without an exception. This is not expected to be a problem forexisting code.

It is possible to pass Python long integers where integers are requested;however no proper range checking is done — the most significant bits aresilently truncated when the receiving field is too small to receive the value(actually, the semantics are inherited from downcasts in C — your mileagemay vary).

A few other characters have a meaning in a format string. These may not occurinside nested parentheses. They are:

|

Indicates that the remaining arguments in the Python argument list areoptional. The C variables corresponding to optional arguments should beinitialized to their default value — when an optional argument is notspecified,PyArg_ParseTuple() does not touch the contents of thecorresponding C variable(s).

:

The list of format units ends here; the string after the colon is used asthe function name in error messages (the “associated value” of theexception thatPyArg_ParseTuple() raises).

;

The list of format units ends here; the string after the semicolon is usedas the error messageinstead of the default error message.: and; mutually exclude each other.

Note that any Python object references which are provided to the caller areborrowed references; do not decrement their reference count!

Additional arguments passed to these functions must be addresses of variableswhose type is determined by the format string; these are used to store valuesfrom the input tuple. There are a few cases, as described in the list offormat units above, where these parameters are used as input values; theyshould match what is specified for the corresponding format unit in that case.

For the conversion to succeed, thearg object must match the format and theformat must be exhausted. On success, thePyArg_Parse*() functionsreturn true, otherwise they return false and raise an appropriate exception.When thePyArg_Parse*() functions fail due to conversion failure inone of the format units, the variables at the addresses corresponding to thatand the following format units are left untouched.

intPyArg_ParseTuple(PyObject *args, const char *format, ...)¶

Parse the parameters of a function that takes only positional parametersinto local variables. Returns true on success; on failure, it returnsfalse and raises the appropriate exception.

intPyArg_VaParse(PyObject *args, const char *format, va_list vargs)¶

Identical toPyArg_ParseTuple(), except that it accepts a va_listrather than a variable number of arguments.

intPyArg_ParseTupleAndKeywords(PyObject *args,PyObject *kw, const char *format, char *keywords[], ...)¶

Parse the parameters of a function that takes both positional and keywordparameters into local variables. Returns true on success; on failure, itreturns false and raises the appropriate exception.

intPyArg_VaParseTupleAndKeywords(PyObject *args,PyObject *kw, const char *format, char *keywords[], va_list vargs)¶

Identical toPyArg_ParseTupleAndKeywords(), except that it accepts ava_list rather than a variable number of arguments.

intPyArg_Parse(PyObject *args, const char *format, ...)¶

Function used to deconstruct the argument lists of “old-style” functions— these are functions which use theMETH_OLDARGS parameterparsing method. This is not recommended for use in parameter parsing innew code, and most code in the standard interpreter has been modified to nolonger use this for that purpose. It does remain a convenient way todecompose other tuples, however, and may continue to be used for thatpurpose.

intPyArg_UnpackTuple(PyObject *args, const char *name, Py_ssize_t min, Py_ssize_t max, ...)¶

A simpler form of parameter retrieval which does not use a format string tospecify the types of the arguments. Functions which use this method toretrieve their parameters should be declared asMETH_VARARGS infunction or method tables. The tuple containing the actual parametersshould be passed asargs; it must actually be a tuple. The length of thetuple must be at leastmin and no more thanmax;min andmax may beequal. Additional arguments must be passed to the function, each of whichshould be a pointer to aPyObject* variable; these will be filledin with the values fromargs; they will contain borrowed references. Thevariables which correspond to optional parameters not given byargs willnot be filled in; these should be initialized by the caller. This functionreturns true on success and false ifargs is not a tuple or contains thewrong number of elements; an exception will be set if there was a failure.

This is an example of the use of this function, taken from the sources forthe_weakref helper module for weak references:

staticPyObject*weakref_ref(PyObject*self,PyObject*args){PyObject*object;PyObject*callback=NULL;PyObject*result=NULL;if(PyArg_UnpackTuple(args,"ref",1,2,&object,&callback)){result=PyWeakref_NewRef(object,callback);}returnresult;}

The call toPyArg_UnpackTuple() in this example is entirelyequivalent to this call toPyArg_ParseTuple():

PyArg_ParseTuple(args,"O|O:ref",&object,&callback)

New in version 2.2.

Changed in version 2.5:This function used anint type formin andmax. This mightrequire changes in your code for properly supporting 64-bit systems.

PyObject*Py_BuildValue(const char *format, ...)¶

Return value: New reference.

Create a new value based on a format string similar to those accepted bythePyArg_Parse*() family of functions and a sequence of values.Returns the value orNULL in the case of an error; an exception will beraised ifNULL is returned.

Py_BuildValue() does not always build a tuple. It builds a tupleonly if its format string contains two or more format units. If the formatstring is empty, it returnsNone; if it contains exactly one formatunit, it returns whatever object is described by that format unit. Toforce it to return a tuple of size0 or one, parenthesize the formatstring.

When memory buffers are passed as parameters to supply data to buildobjects, as for thes ands# formats, the required data is copied.Buffers provided by the caller are never referenced by the objects createdbyPy_BuildValue(). In other words, if your code invokesmalloc() and passes the allocated memory toPy_BuildValue(),your code is responsible for callingfree() for that memory oncePy_BuildValue() returns.

In the following description, the quoted form is the format unit; the entryin (round) parentheses is the Python object type that the format unit willreturn; and the entry in [square] brackets is the type of the C value(s) tobe passed.

The characters space, tab, colon and comma are ignored in format strings(but not within format units such ass#). This can be used to makelong format strings a tad more readable.

s (string) [char *]

Convert a null-terminated C string to a Python object. If the C stringpointer isNULL,None is used.

s# (string) [char *, int]

Convert a C string and its length to a Python object. If the C stringpointer isNULL, the length is ignored andNone is returned.

z (string orNone) [char *]

Same ass.

z# (string orNone) [char *, int]

Same ass#.

u (Unicode string) [Py_UNICODE *]

Convert a null-terminated buffer of Unicode (UCS-2 or UCS-4) data to aPython Unicode object. If the Unicode buffer pointer isNULL,None is returned.

u# (Unicode string) [Py_UNICODE *, int]

Convert a Unicode (UCS-2 or UCS-4) data buffer and its length to aPython Unicode object. If the Unicode buffer pointer isNULL, thelength is ignored andNone is returned.

i (integer) [int]

Convert a plain Cint to a Python integer object.

b (integer) [char]

Convert a plain Cchar to a Python integer object.

h (integer) [short int]

Convert a plain Cshortint to a Python integer object.

l (integer) [long int]

Convert a Clongint to a Python integer object.

B (integer) [unsigned char]

Convert a Cunsignedchar to a Python integer object.

H (integer) [unsigned short int]

Convert a Cunsignedshortint to a Python integer object.

I (integer/long) [unsigned int]

Convert a Cunsignedint to a Python integer object or a Pythonlong integer object, if it is larger thansys.maxint.

k (integer/long) [unsigned long]

Convert a Cunsignedlong to a Python integer object or aPython long integer object, if it is larger thansys.maxint.

L (long) [PY_LONG_LONG]

Convert a Clonglong to a Python long integer object. Onlyavailable on platforms that supportlonglong.

K (long) [unsigned PY_LONG_LONG]

Convert a Cunsignedlonglong to a Python long integer object.Only available on platforms that supportunsignedlonglong.

n (int) [Py_ssize_t]

Convert a CPy_ssize_t to a Python integer or long integer.

New in version 2.5.

c (string of length 1) [char]

Convert a Cint representing a character to a Python string oflength 1.

d (float) [double]

Convert a Cdouble to a Python floating point number.

f (float) [float]

Same asd.

D (complex) [Py_complex *]

Convert a CPy_complex structure to a Python complex number.

O (object) [PyObject *]

Pass a Python object untouched (except for its reference count, which isincremented by one). If the object passed in is aNULL pointer, it isassumed that this was caused because the call producing the argumentfound an error and set an exception. Therefore,Py_BuildValue()will returnNULL but won’t raise an exception. If no exception hasbeen raised yet,SystemError is set.

S (object) [PyObject *]

Same asO.

N (object) [PyObject *]

Same asO, except it doesn’t increment the reference count on theobject. Useful when the object is created by a call to an objectconstructor in the argument list.

O& (object) [converter,anything]

Convertanything to a Python object through aconverter function.The function is called withanything (which should be compatible withvoid*) as its argument and should return a “new” Pythonobject, orNULL if an error occurred.

(items) (tuple) [matching-items]

Convert a sequence of C values to a Python tuple with the same number ofitems.

[items] (list) [matching-items]

Convert a sequence of C values to a Python list with the same number ofitems.

{items} (dictionary) [matching-items]

Convert a sequence of C values to a Python dictionary. Each pair ofconsecutive C values adds one item to the dictionary, serving as key andvalue, respectively.

If there is an error in the format string, theSystemError exceptionis set andNULL returned.

PyObject*Py_VaBuildValue(const char *format, va_list vargs)¶

Identical toPy_BuildValue(), except that it accepts a va_listrather than a variable number of arguments.