cpython/Doc/c-api/buffer.rst

321 lines
17 KiB
ReStructuredText
Raw Normal View History

2008-01-20 05:30:57 -04:00
.. highlightlang:: c
.. _bufferobjects:
Buffer Objects
--------------
.. sectionauthor:: Greg Stein <gstein@lyra.org>
.. sectionauthor:: Benjamin Peterson
2008-01-20 05:30:57 -04:00
.. index::
single: buffer interface
Python objects implemented in C can export a "buffer interface." These
functions can be used by an object to expose its data in a raw, byte-oriented
Merged revisions 71898-71900,71910,71914-71919 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r71898 | jeroen.ruigrok | 2009-04-25 16:24:30 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71899 | jeroen.ruigrok | 2009-04-25 16:27:00 +0200 (za, 25 apr 2009) | 3 lines The type for ppos has been Py_ssize_t since 2.5, reflect this in the documentation. ........ r71900 | jeroen.ruigrok | 2009-04-25 16:28:02 +0200 (za, 25 apr 2009) | 2 lines Reformat paragraph. ........ r71910 | jeroen.ruigrok | 2009-04-25 19:59:03 +0200 (za, 25 apr 2009) | 4 lines Issue #4129: Belatedly document which C API functions had their argument(s) or return type changed from int or int * to Py_ssize_t or Py_ssize_t * as this might cause problems on 64-bit platforms. ........ r71914 | jeroen.ruigrok | 2009-04-25 20:31:20 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71915 | jeroen.ruigrok | 2009-04-25 20:46:03 +0200 (za, 25 apr 2009) | 2 lines Issue #4129: Document more int -> Py_ssize_t changes. ........ r71916 | jeroen.ruigrok | 2009-04-25 20:53:48 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71917 | jeroen.ruigrok | 2009-04-25 20:57:32 +0200 (za, 25 apr 2009) | 2 lines Reference to an int type, whereas it's a Py_ssize_t as the synopsis states. ........ r71918 | jeroen.ruigrok | 2009-04-25 21:04:15 +0200 (za, 25 apr 2009) | 2 lines Since I edited this file, reformat for future edits. ........ r71919 | jeroen.ruigrok | 2009-04-25 21:10:52 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........
2009-04-26 18:06:15 -03:00
format. Clients of the object can use the buffer interface to access the
object data directly, without needing to copy it first.
2008-01-20 05:30:57 -04:00
Two examples of objects that support the buffer interface are bytes and
arrays. The bytes object exposes the character contents in the buffer
2008-01-20 05:30:57 -04:00
interface's byte-oriented form. An array can also expose its contents, but it
should be noted that array elements may be multi-byte values.
An example user of the buffer interface is the file object's :meth:`write`
method. Any object that can export a series of bytes through the buffer
interface can be written to a file. There are a number of format codes to
:cfunc:`PyArg_ParseTuple` that operate against an object's buffer interface,
returning data from the target object.
.. index:: single: PyBufferProcs
2009-01-03 17:18:54 -04:00
More information on the buffer interface is provided in the section
2008-01-20 05:30:57 -04:00
:ref:`buffer-structs`, under the description for :ctype:`PyBufferProcs`.
Buffer objects are useful as a way to expose the data from another object's
buffer interface to the Python programmer. They can also be used as a zero-copy
slicing mechanism. Using their ability to reference a block of memory, it is
possible to expose any data to the Python programmer quite easily. The memory
2008-01-20 05:30:57 -04:00
could be a large, constant array in a C extension, it could be a raw block of
memory for manipulation before passing to an operating system library, or it
could be used to pass around structured data in its native, in-memory format.
.. ctype:: Py_buffer
2008-01-20 05:30:57 -04:00
.. cmember:: void *buf
2008-01-20 05:30:57 -04:00
A pointer to the start of the memory for the object.
2008-01-20 05:30:57 -04:00
.. cmember:: Py_ssize_t len
2008-09-17 19:59:21 -03:00
:noindex:
2008-01-20 05:30:57 -04:00
The total length of the memory in bytes.
2008-01-20 05:30:57 -04:00
.. cmember:: int readonly
2008-01-20 05:30:57 -04:00
An indicator of whether the buffer is read only.
2008-01-20 05:30:57 -04:00
.. cmember:: const char *format
2008-09-17 19:59:21 -03:00
:noindex:
2008-01-20 05:30:57 -04:00
Merged revisions 71898-71900,71910,71914-71919 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r71898 | jeroen.ruigrok | 2009-04-25 16:24:30 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71899 | jeroen.ruigrok | 2009-04-25 16:27:00 +0200 (za, 25 apr 2009) | 3 lines The type for ppos has been Py_ssize_t since 2.5, reflect this in the documentation. ........ r71900 | jeroen.ruigrok | 2009-04-25 16:28:02 +0200 (za, 25 apr 2009) | 2 lines Reformat paragraph. ........ r71910 | jeroen.ruigrok | 2009-04-25 19:59:03 +0200 (za, 25 apr 2009) | 4 lines Issue #4129: Belatedly document which C API functions had their argument(s) or return type changed from int or int * to Py_ssize_t or Py_ssize_t * as this might cause problems on 64-bit platforms. ........ r71914 | jeroen.ruigrok | 2009-04-25 20:31:20 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71915 | jeroen.ruigrok | 2009-04-25 20:46:03 +0200 (za, 25 apr 2009) | 2 lines Issue #4129: Document more int -> Py_ssize_t changes. ........ r71916 | jeroen.ruigrok | 2009-04-25 20:53:48 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71917 | jeroen.ruigrok | 2009-04-25 20:57:32 +0200 (za, 25 apr 2009) | 2 lines Reference to an int type, whereas it's a Py_ssize_t as the synopsis states. ........ r71918 | jeroen.ruigrok | 2009-04-25 21:04:15 +0200 (za, 25 apr 2009) | 2 lines Since I edited this file, reformat for future edits. ........ r71919 | jeroen.ruigrok | 2009-04-25 21:10:52 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........
2009-04-26 18:06:15 -03:00
A *NULL* terminated string in :mod:`struct` module style syntax giving
the contents of the elements available through the buffer. If this is
*NULL*, ``"B"`` (unsigned bytes) is assumed.
2008-01-20 05:30:57 -04:00
.. cmember:: int ndim
2008-01-20 05:30:57 -04:00
The number of dimensions the memory represents as a multi-dimensional
array. If it is 0, :cdata:`strides` and :cdata:`suboffsets` must be
*NULL*.
2008-01-20 05:30:57 -04:00
.. cmember:: Py_ssize_t *shape
2008-01-20 05:30:57 -04:00
An array of :ctype:`Py_ssize_t`\s the length of :cdata:`ndim` giving the
shape of the memory as a multi-dimensional array. Note that
``((*shape)[0] * ... * (*shape)[ndims-1])*itemsize`` should be equal to
:cdata:`len`.
2008-01-20 05:30:57 -04:00
.. cmember:: Py_ssize_t *strides
2008-01-20 05:30:57 -04:00
An array of :ctype:`Py_ssize_t`\s the length of :cdata:`ndim` giving the
number of bytes to skip to get to a new element in each dimension.
2008-01-20 05:30:57 -04:00
.. cmember:: Py_ssize_t *suboffsets
2008-01-20 05:30:57 -04:00
An array of :ctype:`Py_ssize_t`\s the length of :cdata:`ndim`. If these
suboffset numbers are greater than or equal to 0, then the value stored
along the indicated dimension is a pointer and the suboffset value
dictates how many bytes to add to the pointer after de-referencing. A
suboffset value that it negative indicates that no de-referencing should
occur (striding in a contiguous memory block).
2008-01-20 05:30:57 -04:00
Here is a function that returns a pointer to the element in an N-D array
2009-03-13 16:04:40 -03:00
pointed to by an N-dimensional index when there are both non-NULL strides
and suboffsets::
void *get_item_pointer(int ndim, void *buf, Py_ssize_t *strides,
Py_ssize_t *suboffsets, Py_ssize_t *indices) {
char *pointer = (char*)buf;
int i;
for (i = 0; i < ndim; i++) {
pointer += strides[i] * indices[i];
if (suboffsets[i] >=0 ) {
pointer = *((char**)pointer) + suboffsets[i];
2009-01-03 17:18:54 -04:00
}
}
return (void*)pointer;
}
.. cmember:: Py_ssize_t itemsize
This is a storage for the itemsize (in bytes) of each element of the
Merged revisions 71898-71900,71910,71914-71919 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r71898 | jeroen.ruigrok | 2009-04-25 16:24:30 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71899 | jeroen.ruigrok | 2009-04-25 16:27:00 +0200 (za, 25 apr 2009) | 3 lines The type for ppos has been Py_ssize_t since 2.5, reflect this in the documentation. ........ r71900 | jeroen.ruigrok | 2009-04-25 16:28:02 +0200 (za, 25 apr 2009) | 2 lines Reformat paragraph. ........ r71910 | jeroen.ruigrok | 2009-04-25 19:59:03 +0200 (za, 25 apr 2009) | 4 lines Issue #4129: Belatedly document which C API functions had their argument(s) or return type changed from int or int * to Py_ssize_t or Py_ssize_t * as this might cause problems on 64-bit platforms. ........ r71914 | jeroen.ruigrok | 2009-04-25 20:31:20 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71915 | jeroen.ruigrok | 2009-04-25 20:46:03 +0200 (za, 25 apr 2009) | 2 lines Issue #4129: Document more int -> Py_ssize_t changes. ........ r71916 | jeroen.ruigrok | 2009-04-25 20:53:48 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71917 | jeroen.ruigrok | 2009-04-25 20:57:32 +0200 (za, 25 apr 2009) | 2 lines Reference to an int type, whereas it's a Py_ssize_t as the synopsis states. ........ r71918 | jeroen.ruigrok | 2009-04-25 21:04:15 +0200 (za, 25 apr 2009) | 2 lines Since I edited this file, reformat for future edits. ........ r71919 | jeroen.ruigrok | 2009-04-25 21:10:52 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........
2009-04-26 18:06:15 -03:00
shared memory. It is technically un-necessary as it can be obtained
using :cfunc:`PyBuffer_SizeFromFormat`, however an exporter may know
this information without parsing the format string and it is necessary
to know the itemsize for proper interpretation of striding. Therefore,
storing it is more convenient and faster.
.. cmember:: void *internal
This is for use internally by the exporting object. For example, this
might be re-cast as an integer by the exporter and used to store flags
about whether or not the shape, strides, and suboffsets arrays must be
freed when the buffer is released. The consumer should never alter this
value.
Buffer related functions
========================
.. cfunction:: int PyObject_CheckBuffer(PyObject *obj)
Return 1 if *obj* supports the buffer interface otherwise 0.
Merged revisions 72506,72525-72526,72551,72558,72616,72654-72655,72689,72745,72750,72802,72812,72822,72824,72826-72827,72833,72876,72890,72923,72946,73026,73042,73045,73047,73065,73068-73069 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r72506 | vinay.sajip | 2009-05-09 07:07:17 -0500 (Sat, 09 May 2009) | 1 line Issue #5971: StreamHandler.handleError now swallows IOErrors which occur when trying to print a traceback. ........ r72525 | benjamin.peterson | 2009-05-09 20:38:02 -0500 (Sat, 09 May 2009) | 1 line close file explicitly ........ r72526 | benjamin.peterson | 2009-05-09 21:29:00 -0500 (Sat, 09 May 2009) | 1 line make sure files are closed using the with statement ........ r72551 | benjamin.peterson | 2009-05-10 09:16:47 -0500 (Sun, 10 May 2009) | 1 line use isinstance ........ r72558 | benjamin.peterson | 2009-05-10 18:52:09 -0500 (Sun, 10 May 2009) | 1 line sys.setdefaultencoding() strikes me as a bad example ........ r72616 | benjamin.peterson | 2009-05-13 19:33:10 -0500 (Wed, 13 May 2009) | 1 line importlib.import_module is better these days ........ r72654 | benjamin.peterson | 2009-05-14 17:37:49 -0500 (Thu, 14 May 2009) | 1 line prevent refleaks from threads ........ r72655 | benjamin.peterson | 2009-05-14 17:40:34 -0500 (Thu, 14 May 2009) | 1 line a useful decorator for cleaning up threads ........ r72689 | benjamin.peterson | 2009-05-16 13:44:34 -0500 (Sat, 16 May 2009) | 1 line use skipTest() ........ r72745 | benjamin.peterson | 2009-05-17 09:16:29 -0500 (Sun, 17 May 2009) | 1 line ignore .rst files in sphinx its self ........ r72750 | benjamin.peterson | 2009-05-17 11:59:27 -0500 (Sun, 17 May 2009) | 1 line chop off slash ........ r72802 | georg.brandl | 2009-05-20 13:35:27 -0500 (Wed, 20 May 2009) | 1 line #6051: refer to email examples for better way to construct email messages. ........ r72812 | michael.foord | 2009-05-21 17:57:02 -0500 (Thu, 21 May 2009) | 1 line Rename TestCase._result to _resultForDoCleanups to avoid potential clashes in TestCase subclasses. Issue 6072. ........ r72822 | georg.brandl | 2009-05-22 04:33:25 -0500 (Fri, 22 May 2009) | 1 line #6084: fix example. ........ r72824 | georg.brandl | 2009-05-22 04:43:17 -0500 (Fri, 22 May 2009) | 1 line Fix references to file-related functions and methods (os.* vs file.*). ........ r72826 | georg.brandl | 2009-05-22 04:49:42 -0500 (Fri, 22 May 2009) | 1 line Fix confusing wording. ........ r72827 | georg.brandl | 2009-05-22 04:50:30 -0500 (Fri, 22 May 2009) | 1 line s/use/call/ ........ r72833 | georg.brandl | 2009-05-22 12:00:17 -0500 (Fri, 22 May 2009) | 1 line #6078: _warnings is a builtin module and has no standard init_warnings function. ........ r72876 | benjamin.peterson | 2009-05-23 15:59:09 -0500 (Sat, 23 May 2009) | 1 line remove mention of old ctypes version ........ r72890 | gregory.p.smith | 2009-05-24 13:00:13 -0500 (Sun, 24 May 2009) | 2 lines add a versionadded tag for set_tunnel ........ r72923 | michael.foord | 2009-05-25 15:36:56 -0500 (Mon, 25 May 2009) | 1 line Make assertSequenceEqual error messages less cryptic, particularly for nested sequences. ........ r72946 | ronald.oussoren | 2009-05-26 13:44:48 -0500 (Tue, 26 May 2009) | 2 lines Fixes issue 6110 ........ r73026 | r.david.murray | 2009-05-29 14:30:27 -0500 (Fri, 29 May 2009) | 3 lines Issue 6141: document that the first item of args is still the command name even when executable is specified. ........ r73042 | benjamin.peterson | 2009-05-29 22:10:52 -0500 (Fri, 29 May 2009) | 1 line no fdatasync on macos ........ r73045 | georg.brandl | 2009-05-30 02:26:04 -0500 (Sat, 30 May 2009) | 1 line #6146: fix markup bug. ........ r73047 | georg.brandl | 2009-05-30 05:33:23 -0500 (Sat, 30 May 2009) | 1 line Fix some more small markup problems. ........ r73065 | antoine.pitrou | 2009-05-30 16:39:25 -0500 (Sat, 30 May 2009) | 3 lines The test for #5330 wasn't correct. ........ r73068 | antoine.pitrou | 2009-05-30 16:45:40 -0500 (Sat, 30 May 2009) | 3 lines Update ACKS ........ r73069 | benjamin.peterson | 2009-05-30 19:42:42 -0500 (Sat, 30 May 2009) | 1 line fix signature ........
2009-06-01 19:42:33 -03:00
.. cfunction:: int PyObject_GetBuffer(PyObject *obj, Py_buffer *view, int flags)
Export *obj* into a :ctype:`Py_buffer`, *view*. These arguments must
Merged revisions 71898-71900,71910,71914-71919 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r71898 | jeroen.ruigrok | 2009-04-25 16:24:30 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71899 | jeroen.ruigrok | 2009-04-25 16:27:00 +0200 (za, 25 apr 2009) | 3 lines The type for ppos has been Py_ssize_t since 2.5, reflect this in the documentation. ........ r71900 | jeroen.ruigrok | 2009-04-25 16:28:02 +0200 (za, 25 apr 2009) | 2 lines Reformat paragraph. ........ r71910 | jeroen.ruigrok | 2009-04-25 19:59:03 +0200 (za, 25 apr 2009) | 4 lines Issue #4129: Belatedly document which C API functions had their argument(s) or return type changed from int or int * to Py_ssize_t or Py_ssize_t * as this might cause problems on 64-bit platforms. ........ r71914 | jeroen.ruigrok | 2009-04-25 20:31:20 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71915 | jeroen.ruigrok | 2009-04-25 20:46:03 +0200 (za, 25 apr 2009) | 2 lines Issue #4129: Document more int -> Py_ssize_t changes. ........ r71916 | jeroen.ruigrok | 2009-04-25 20:53:48 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71917 | jeroen.ruigrok | 2009-04-25 20:57:32 +0200 (za, 25 apr 2009) | 2 lines Reference to an int type, whereas it's a Py_ssize_t as the synopsis states. ........ r71918 | jeroen.ruigrok | 2009-04-25 21:04:15 +0200 (za, 25 apr 2009) | 2 lines Since I edited this file, reformat for future edits. ........ r71919 | jeroen.ruigrok | 2009-04-25 21:10:52 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........
2009-04-26 18:06:15 -03:00
never be *NULL*. The *flags* argument is a bit field indicating what
kind of buffer the caller is prepared to deal with and therefore what
kind of buffer the exporter is allowed to return. The buffer interface
allows for complicated memory sharing possibilities, but some caller may
Merged revisions 73941-73943,74076,74094,74186,74211-74214,74247,74254,74262,74311,74334,74368 via svnmerge from svn+ssh://svn.python.org/python/branches/py3k ................ r73941 | georg.brandl | 2009-07-11 12:39:00 +0200 (Sa, 11 Jul 2009) | 9 lines Merged revisions 73940 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r73940 | georg.brandl | 2009-07-11 12:37:38 +0200 (Sa, 11 Jul 2009) | 1 line #6430: add note about size of "u" type. ........ ................ r73942 | georg.brandl | 2009-07-11 12:39:23 +0200 (Sa, 11 Jul 2009) | 1 line #6430: remove mention of "w" array typecode. ................ r73943 | georg.brandl | 2009-07-11 12:43:08 +0200 (Sa, 11 Jul 2009) | 1 line #6421: The self argument of module-level PyCFunctions is now a reference to the module object. ................ r74076 | georg.brandl | 2009-07-18 11:07:48 +0200 (Sa, 18 Jul 2009) | 1 line #6502: add missing comma in docstring. ................ r74094 | georg.brandl | 2009-07-19 09:25:56 +0200 (So, 19 Jul 2009) | 10 lines Recorded merge of revisions 74089 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r74089 | senthil.kumaran | 2009-07-19 04:43:43 +0200 (So, 19 Jul 2009) | 3 lines Fix for issue5102, timeout value propages between redirects, proxy, digest and auth handlers. Fixed tests to reflect the same. ........ ................ r74186 | georg.brandl | 2009-07-23 11:19:09 +0200 (Do, 23 Jul 2009) | 9 lines Recorded merge of revisions 74185 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r74185 | georg.brandl | 2009-07-23 11:17:09 +0200 (Do, 23 Jul 2009) | 1 line Fix the "pylocals" gdb command. ........ ................ r74211 | georg.brandl | 2009-07-26 16:48:09 +0200 (So, 26 Jul 2009) | 9 lines Recorded merge of revisions 74210 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r74210 | georg.brandl | 2009-07-26 16:44:23 +0200 (So, 26 Jul 2009) | 1 line Move member descriptions inside the classes. ........ ................ r74212 | georg.brandl | 2009-07-26 16:54:51 +0200 (So, 26 Jul 2009) | 9 lines Merged revisions 74209 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r74209 | georg.brandl | 2009-07-26 16:37:28 +0200 (So, 26 Jul 2009) | 1 line builtin -> built-in. ........ ................ r74213 | georg.brandl | 2009-07-26 17:02:41 +0200 (So, 26 Jul 2009) | 9 lines Merged revisions 74207 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r74207 | georg.brandl | 2009-07-26 16:19:57 +0200 (So, 26 Jul 2009) | 1 line #6577: fix (hopefully) all links to builtin instead of module/class-specific objects. ........ ................ r74214 | georg.brandl | 2009-07-26 17:03:49 +0200 (So, 26 Jul 2009) | 9 lines Merged revisions 74205 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r74205 | georg.brandl | 2009-07-26 15:36:39 +0200 (So, 26 Jul 2009) | 1 line #6576: fix cross-refs in re docs. ........ ................ r74247 | georg.brandl | 2009-07-29 09:27:08 +0200 (Mi, 29 Jul 2009) | 9 lines Merged revisions 74239 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r74239 | georg.brandl | 2009-07-28 18:55:32 +0000 (Di, 28 Jul 2009) | 1 line Clarify quote_plus() usage. ........ ................ r74254 | georg.brandl | 2009-07-29 18:14:16 +0200 (Mi, 29 Jul 2009) | 1 line #6586: fix return/argument type doc for os.read() and os.write(). ................ r74262 | alexandre.vassalotti | 2009-07-29 21:54:39 +0200 (Mi, 29 Jul 2009) | 57 lines Merged revisions 74074,74077,74111,74188,74192-74193,74200,74252-74253,74258-74261 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r74074 | georg.brandl | 2009-07-18 05:03:10 -0400 (Sat, 18 Jul 2009) | 1 line #6513: fix example code: warning categories are classes, not instances. ........ r74077 | georg.brandl | 2009-07-18 05:43:40 -0400 (Sat, 18 Jul 2009) | 1 line #6489: fix an ambiguity in getiterator() documentation. ........ r74111 | benjamin.peterson | 2009-07-20 09:30:10 -0400 (Mon, 20 Jul 2009) | 1 line remove docs for deprecated -p option ........ r74188 | benjamin.peterson | 2009-07-23 10:25:31 -0400 (Thu, 23 Jul 2009) | 1 line use bools ........ r74192 | georg.brandl | 2009-07-24 12:28:38 -0400 (Fri, 24 Jul 2009) | 1 line Fix arg types of et#. ........ r74193 | georg.brandl | 2009-07-24 12:46:38 -0400 (Fri, 24 Jul 2009) | 1 line Dont put "void" in signature for nullary functions. ........ r74200 | georg.brandl | 2009-07-25 09:02:15 -0400 (Sat, 25 Jul 2009) | 1 line #6571: add index entries for more operators. ........ r74252 | georg.brandl | 2009-07-29 12:06:31 -0400 (Wed, 29 Jul 2009) | 1 line #6593: fix link targets. ........ r74253 | georg.brandl | 2009-07-29 12:09:17 -0400 (Wed, 29 Jul 2009) | 1 line #6591: add reference to ioctl in fcntl module for platforms other than Windows. ........ r74258 | georg.brandl | 2009-07-29 12:57:05 -0400 (Wed, 29 Jul 2009) | 1 line Add a link to readline, and mention IPython and bpython. ........ r74259 | georg.brandl | 2009-07-29 13:07:21 -0400 (Wed, 29 Jul 2009) | 1 line Fix some markup and small factual glitches found by M. Markert. ........ r74260 | georg.brandl | 2009-07-29 13:15:20 -0400 (Wed, 29 Jul 2009) | 1 line Fix a few markup glitches. ........ r74261 | georg.brandl | 2009-07-29 13:50:25 -0400 (Wed, 29 Jul 2009) | 1 line Rewrite the section about classes a bit; mostly tidbits, and a larger update to the section about "private" variables to reflect the Pythonic consensus better. ........ ................ r74311 | georg.brandl | 2009-08-04 22:29:27 +0200 (Di, 04 Aug 2009) | 1 line Slightly improve buffer-related error message. ................ r74334 | georg.brandl | 2009-08-06 19:51:03 +0200 (Do, 06 Aug 2009) | 1 line #6648: mention surrogateescape handler where all standard handlers are listed. ................ r74368 | georg.brandl | 2009-08-13 09:56:35 +0200 (Do, 13 Aug 2009) | 21 lines Merged revisions 74328,74332-74333,74365 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r74328 | georg.brandl | 2009-08-06 17:06:25 +0200 (Do, 06 Aug 2009) | 1 line Fix base keyword arg name for int() and long(). ........ r74332 | georg.brandl | 2009-08-06 19:23:21 +0200 (Do, 06 Aug 2009) | 1 line Fix punctuation and one copy-paste error. ........ r74333 | georg.brandl | 2009-08-06 19:43:55 +0200 (Do, 06 Aug 2009) | 1 line #6658: fix two typos. ........ r74365 | georg.brandl | 2009-08-13 09:48:05 +0200 (Do, 13 Aug 2009) | 1 line #6679: Remove mention that sub supports no flags. ........ ................
2009-08-13 05:26:44 -03:00
not be able to handle all the complexity but may want to see if the
Merged revisions 71898-71900,71910,71914-71919 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r71898 | jeroen.ruigrok | 2009-04-25 16:24:30 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71899 | jeroen.ruigrok | 2009-04-25 16:27:00 +0200 (za, 25 apr 2009) | 3 lines The type for ppos has been Py_ssize_t since 2.5, reflect this in the documentation. ........ r71900 | jeroen.ruigrok | 2009-04-25 16:28:02 +0200 (za, 25 apr 2009) | 2 lines Reformat paragraph. ........ r71910 | jeroen.ruigrok | 2009-04-25 19:59:03 +0200 (za, 25 apr 2009) | 4 lines Issue #4129: Belatedly document which C API functions had their argument(s) or return type changed from int or int * to Py_ssize_t or Py_ssize_t * as this might cause problems on 64-bit platforms. ........ r71914 | jeroen.ruigrok | 2009-04-25 20:31:20 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71915 | jeroen.ruigrok | 2009-04-25 20:46:03 +0200 (za, 25 apr 2009) | 2 lines Issue #4129: Document more int -> Py_ssize_t changes. ........ r71916 | jeroen.ruigrok | 2009-04-25 20:53:48 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71917 | jeroen.ruigrok | 2009-04-25 20:57:32 +0200 (za, 25 apr 2009) | 2 lines Reference to an int type, whereas it's a Py_ssize_t as the synopsis states. ........ r71918 | jeroen.ruigrok | 2009-04-25 21:04:15 +0200 (za, 25 apr 2009) | 2 lines Since I edited this file, reformat for future edits. ........ r71919 | jeroen.ruigrok | 2009-04-25 21:10:52 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........
2009-04-26 18:06:15 -03:00
exporter will let them take a simpler view to its memory.
Some exporters may not be able to share memory in every possible way and
may need to raise errors to signal to some consumers that something is
just not possible. These errors should be a :exc:`BufferError` unless
Merged revisions 71898-71900,71910,71914-71919 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r71898 | jeroen.ruigrok | 2009-04-25 16:24:30 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71899 | jeroen.ruigrok | 2009-04-25 16:27:00 +0200 (za, 25 apr 2009) | 3 lines The type for ppos has been Py_ssize_t since 2.5, reflect this in the documentation. ........ r71900 | jeroen.ruigrok | 2009-04-25 16:28:02 +0200 (za, 25 apr 2009) | 2 lines Reformat paragraph. ........ r71910 | jeroen.ruigrok | 2009-04-25 19:59:03 +0200 (za, 25 apr 2009) | 4 lines Issue #4129: Belatedly document which C API functions had their argument(s) or return type changed from int or int * to Py_ssize_t or Py_ssize_t * as this might cause problems on 64-bit platforms. ........ r71914 | jeroen.ruigrok | 2009-04-25 20:31:20 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71915 | jeroen.ruigrok | 2009-04-25 20:46:03 +0200 (za, 25 apr 2009) | 2 lines Issue #4129: Document more int -> Py_ssize_t changes. ........ r71916 | jeroen.ruigrok | 2009-04-25 20:53:48 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71917 | jeroen.ruigrok | 2009-04-25 20:57:32 +0200 (za, 25 apr 2009) | 2 lines Reference to an int type, whereas it's a Py_ssize_t as the synopsis states. ........ r71918 | jeroen.ruigrok | 2009-04-25 21:04:15 +0200 (za, 25 apr 2009) | 2 lines Since I edited this file, reformat for future edits. ........ r71919 | jeroen.ruigrok | 2009-04-25 21:10:52 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........
2009-04-26 18:06:15 -03:00
there is another error that is actually causing the problem. The
exporter can use flags information to simplify how much of the
:cdata:`Py_buffer` structure is filled in with non-default values and/or
raise an error if the object can't support a simpler view of its memory.
0 is returned on success and -1 on error.
The following table gives possible values to the *flags* arguments.
2008-11-20 18:38:20 -04:00
+------------------------------+---------------------------------------------------+
| Flag | Description |
+==============================+===================================================+
| :cmacro:`PyBUF_SIMPLE` | This is the default flag state. The returned |
| | buffer may or may not have writable memory. The |
| | format of the data will be assumed to be unsigned |
| | bytes. This is a "stand-alone" flag constant. It |
| | never needs to be '|'d to the others. The exporter|
| | will raise an error if it cannot provide such a |
| | contiguous buffer of bytes. |
| | |
+------------------------------+---------------------------------------------------+
| :cmacro:`PyBUF_WRITABLE` | The returned buffer must be writable. If it is |
| | not writable, then raise an error. |
+------------------------------+---------------------------------------------------+
| :cmacro:`PyBUF_STRIDES` | This implies :cmacro:`PyBUF_ND`. The returned |
| | buffer must provide strides information (i.e. the |
| | strides cannot be NULL). This would be used when |
| | the consumer can handle strided, discontiguous |
| | arrays. Handling strides automatically assumes |
| | you can handle shape. The exporter can raise an |
| | error if a strided representation of the data is |
| | not possible (i.e. without the suboffsets). |
| | |
+------------------------------+---------------------------------------------------+
| :cmacro:`PyBUF_ND` | The returned buffer must provide shape |
| | information. The memory will be assumed C-style |
| | contiguous (last dimension varies the |
| | fastest). The exporter may raise an error if it |
| | cannot provide this kind of contiguous buffer. If |
| | this is not given then shape will be *NULL*. |
| | |
| | |
| | |
+------------------------------+---------------------------------------------------+
|:cmacro:`PyBUF_C_CONTIGUOUS` | These flags indicate that the contiguity returned |
|:cmacro:`PyBUF_F_CONTIGUOUS` | buffer must be respectively, C-contiguous (last |
|:cmacro:`PyBUF_ANY_CONTIGUOUS`| dimension varies the fastest), Fortran contiguous |
| | (first dimension varies the fastest) or either |
| | one. All of these flags imply |
| | :cmacro:`PyBUF_STRIDES` and guarantee that the |
| | strides buffer info structure will be filled in |
| | correctly. |
| | |
+------------------------------+---------------------------------------------------+
| :cmacro:`PyBUF_INDIRECT` | This flag indicates the returned buffer must have |
| | suboffsets information (which can be NULL if no |
| | suboffsets are needed). This can be used when |
| | the consumer can handle indirect array |
| | referencing implied by these suboffsets. This |
| | implies :cmacro:`PyBUF_STRIDES`. |
| | |
| | |
| | |
+------------------------------+---------------------------------------------------+
| :cmacro:`PyBUF_FORMAT` | The returned buffer must have true format |
| | information if this flag is provided. This would |
| | be used when the consumer is going to be checking |
| | for what 'kind' of data is actually stored. An |
| | exporter should always be able to provide this |
| | information if requested. If format is not |
| | explicitly requested then the format must be |
| | returned as *NULL* (which means ``'B'``, or |
| | unsigned bytes) |
+------------------------------+---------------------------------------------------+
| :cmacro:`PyBUF_STRIDED` | This is equivalent to ``(PyBUF_STRIDES | |
| | PyBUF_WRITABLE)``. |
+------------------------------+---------------------------------------------------+
| :cmacro:`PyBUF_STRIDED_RO` | This is equivalent to ``(PyBUF_STRIDES)``. |
| | |
+------------------------------+---------------------------------------------------+
| :cmacro:`PyBUF_RECORDS` | This is equivalent to ``(PyBUF_STRIDES | |
| | PyBUF_FORMAT | PyBUF_WRITABLE)``. |
+------------------------------+---------------------------------------------------+
| :cmacro:`PyBUF_RECORDS_RO` | This is equivalent to ``(PyBUF_STRIDES | |
| | PyBUF_FORMAT)``. |
+------------------------------+---------------------------------------------------+
| :cmacro:`PyBUF_FULL` | This is equivalent to ``(PyBUF_INDIRECT | |
| | PyBUF_FORMAT | PyBUF_WRITABLE)``. |
+------------------------------+---------------------------------------------------+
Merged revisions 70980,71059,71225,71234,71241,71243,71249,71251,71255,71266,71299,71329,71397-71398,71486 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r70980 | jack.diederich | 2009-04-01 15:26:13 -0500 (Wed, 01 Apr 2009) | 3 lines bounds check arguments to mmap.move(). All of them. Really. fixes crasher on OS X 10.5 ........ r71059 | mark.dickinson | 2009-04-02 13:39:37 -0500 (Thu, 02 Apr 2009) | 2 lines sys.long_info attributes should be ints, not longs ........ r71225 | georg.brandl | 2009-04-05 06:54:07 -0500 (Sun, 05 Apr 2009) | 1 line #5580: no need to use parentheses when converterr() argument is actually a type description. ........ r71234 | georg.brandl | 2009-04-05 08:16:35 -0500 (Sun, 05 Apr 2009) | 1 line Whitespace normalization. ........ r71241 | georg.brandl | 2009-04-05 09:48:49 -0500 (Sun, 05 Apr 2009) | 1 line #5471: fix expanduser() for $HOME set to "/". ........ r71243 | georg.brandl | 2009-04-05 10:14:29 -0500 (Sun, 05 Apr 2009) | 1 line #5432: make plistlib docstring a raw string, since it contains examples with backslash escapes. ........ r71249 | georg.brandl | 2009-04-05 11:30:43 -0500 (Sun, 05 Apr 2009) | 1 line #5444: adapt make.bat to new htmlhelp output file name. ........ r71251 | georg.brandl | 2009-04-05 12:17:42 -0500 (Sun, 05 Apr 2009) | 1 line #5298: clarify docs about GIL by using more consistent wording. ........ r71255 | georg.brandl | 2009-04-05 13:34:58 -0500 (Sun, 05 Apr 2009) | 1 line #602893: add indicator for current line in cgitb that doesnt rely on styling alone. ........ r71266 | georg.brandl | 2009-04-05 15:23:13 -0500 (Sun, 05 Apr 2009) | 1 line Normalize issue referencing style. ........ r71299 | gregory.p.smith | 2009-04-05 18:43:58 -0500 (Sun, 05 Apr 2009) | 3 lines Fixes issue5705: os.setuid() and friends did not accept the same range of values that pwd.getpwnam() returns. ........ r71329 | benjamin.peterson | 2009-04-06 16:53:33 -0500 (Mon, 06 Apr 2009) | 1 line add create_connection to __all__ #5711 ........ r71397 | georg.brandl | 2009-04-08 11:36:39 -0500 (Wed, 08 Apr 2009) | 1 line Remove redundant backtick. ........ r71398 | georg.brandl | 2009-04-08 11:39:04 -0500 (Wed, 08 Apr 2009) | 1 line Update ignore file for suspicious builder. ........ r71486 | andrew.kuchling | 2009-04-11 11:18:14 -0500 (Sat, 11 Apr 2009) | 1 line Re-word ........
2009-04-11 16:48:14 -03:00
| :cmacro:`PyBUF_FULL_RO` | This is equivalent to ``(PyBUF_INDIRECT | |
2008-11-20 18:38:20 -04:00
| | PyBUF_FORMAT)``. |
+------------------------------+---------------------------------------------------+
| :cmacro:`PyBUF_CONTIG` | This is equivalent to ``(PyBUF_ND | |
| | PyBUF_WRITABLE)``. |
+------------------------------+---------------------------------------------------+
| :cmacro:`PyBUF_CONTIG_RO` | This is equivalent to ``(PyBUF_ND)``. |
| | |
+------------------------------+---------------------------------------------------+
.. cfunction:: void PyBuffer_Release(PyObject *obj, Py_buffer *view)
2009-03-02 01:49:44 -04:00
Release the buffer *view* over *obj*. This should be called when the buffer
is no longer being used as it may free memory from it.
.. cfunction:: Py_ssize_t PyBuffer_SizeFromFormat(const char *)
Return the implied :cdata:`~Py_buffer.itemsize` from the struct-stype
:cdata:`~Py_buffer.format`.
.. cfunction:: int PyObject_CopyToObject(PyObject *obj, void *buf, Py_ssize_t len, char fortran)
Merged revisions 71898-71900,71910,71914-71919 via svnmerge from svn+ssh://pythondev@svn.python.org/python/trunk ........ r71898 | jeroen.ruigrok | 2009-04-25 16:24:30 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71899 | jeroen.ruigrok | 2009-04-25 16:27:00 +0200 (za, 25 apr 2009) | 3 lines The type for ppos has been Py_ssize_t since 2.5, reflect this in the documentation. ........ r71900 | jeroen.ruigrok | 2009-04-25 16:28:02 +0200 (za, 25 apr 2009) | 2 lines Reformat paragraph. ........ r71910 | jeroen.ruigrok | 2009-04-25 19:59:03 +0200 (za, 25 apr 2009) | 4 lines Issue #4129: Belatedly document which C API functions had their argument(s) or return type changed from int or int * to Py_ssize_t or Py_ssize_t * as this might cause problems on 64-bit platforms. ........ r71914 | jeroen.ruigrok | 2009-04-25 20:31:20 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71915 | jeroen.ruigrok | 2009-04-25 20:46:03 +0200 (za, 25 apr 2009) | 2 lines Issue #4129: Document more int -> Py_ssize_t changes. ........ r71916 | jeroen.ruigrok | 2009-04-25 20:53:48 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........ r71917 | jeroen.ruigrok | 2009-04-25 20:57:32 +0200 (za, 25 apr 2009) | 2 lines Reference to an int type, whereas it's a Py_ssize_t as the synopsis states. ........ r71918 | jeroen.ruigrok | 2009-04-25 21:04:15 +0200 (za, 25 apr 2009) | 2 lines Since I edited this file, reformat for future edits. ........ r71919 | jeroen.ruigrok | 2009-04-25 21:10:52 +0200 (za, 25 apr 2009) | 2 lines Reformat prior to editing. ........
2009-04-26 18:06:15 -03:00
Copy *len* bytes of data pointed to by the contiguous chunk of memory
pointed to by *buf* into the buffer exported by obj. The buffer must of
course be writable. Return 0 on success and return -1 and raise an error
on failure. If the object does not have a writable buffer, then an error
is raised. If *fortran* is ``'F'``, then if the object is
multi-dimensional, then the data will be copied into the array in
Fortran-style (first dimension varies the fastest). If *fortran* is
``'C'``, then the data will be copied into the array in C-style (last
dimension varies the fastest). If *fortran* is ``'A'``, then it does not
matter and the copy will be made in whatever way is more efficient.
.. cfunction:: int PyBuffer_IsContiguous(Py_buffer *view, char fortran)
Return 1 if the memory defined by the *view* is C-style (*fortran* is
``'C'``) or Fortran-style (*fortran* is ``'F'``) contiguous or either one
(*fortran* is ``'A'``). Return 0 otherwise.
.. cfunction:: void PyBuffer_FillContiguousStrides(int ndim, Py_ssize_t *shape, Py_ssize_t *strides, Py_ssize_t itemsize, char fortran)
Fill the *strides* array with byte-strides of a contiguous (C-style if
*fortran* is ``'C'`` or Fortran-style if *fortran* is ``'F'`` array of the
given shape with the given number of bytes per element.
.. cfunction:: int PyBuffer_FillInfo(Py_buffer *view, void *buf, Py_ssize_t len, int readonly, int infoflags)
Fill in a buffer-info structure, *view*, correctly for an exporter that can
only share a contiguous chunk of memory of "unsigned bytes" of the given
length. Return 0 on success and -1 (with raising an error) on error.
.. index::
object: memoryview
MemoryView objects
==================
A memoryview object exposes the C level buffer interface to Python.
.. cfunction:: PyObject* PyMemoryView_FromObject(PyObject *obj)
Return a memoryview object from an object that defines the buffer interface.
.. cfunction:: PyObject * PyMemoryView_GetContiguous(PyObject *obj, int buffertype, char order)
Return a memoryview object to a contiguous chunk of memory (in either
'C' or 'F'ortran order) from an object that defines the buffer
interface. If memory is contiguous, the memoryview object points to the
original memory. Otherwise copy is made and the memoryview points to a
new bytes object.