Move the following private functions and structures to
pycore_modsupport.h internal C API:
* _PyArg_BadArgument()
* _PyArg_CheckPositional()
* _PyArg_NoKeywords()
* _PyArg_NoPositional()
* _PyArg_ParseStack()
* _PyArg_ParseStackAndKeywords()
* _PyArg_Parser structure
* _PyArg_UnpackKeywords()
* _PyArg_UnpackKeywordsWithVararg()
* _PyArg_UnpackStack()
* _Py_ANY_VARARGS()
Changes:
* Python/getargs.h now includes pycore_modsupport.h to export
functions.
* clinic.py now adds pycore_modsupport.h when one of these functions
is used.
* Add pycore_modsupport.h includes when a C extension uses one of
these functions.
* Define Py_BUILD_CORE_MODULE in C extensions which now include
directly or indirectly (via code generated by Argument Clinic)
pycore_modsupport.h:
* _csv
* _curses_panel
* _dbm
* _gdbm
* _multiprocessing.posixshmem
* _sqlite.row
* _statistics
* grp
* resource
* syslog
* _testcapi: bad_get() no longer uses METH_FASTCALL calling
convention but METH_VARARGS. Replace _PyArg_UnpackStack() with
PyArg_ParseTuple().
* _testcapi: add PYTESTCAPI_NEED_INTERNAL_API macro which is defined
by _testcapi sub-modules which need the internal C API
(pycore_modsupport.h): exceptions.c, float.c, vectorcall.c,
watchers.c.
* Remove Include/cpython/modsupport.h header file.
Include/modsupport.h no longer includes the removed header file.
* Fix mypy clinic.py
* Move <ctype.h>, <limits.h> and <stdarg.h> standard includes to
Python.h.
* Move "pystats.h" include from object.h to Python.h.
* Remove redundant "pymem.h" include in objimpl.h and "pyport.h"
include in pymem.h; Python.h already includes them earlier.
* Remove redundant <wchar.h> include in unicodeobject.h; Python.h
already includes it.
* Move _SGI_MP_SOURCE define from Python.h to pyport.h.
* pycore_condvar.h includes explicitly <unistd.h> for the
_POSIX_THREADS macro.
Python built with "configure --with-trace-refs" (tracing references)
is now ABI compatible with Python release build and debug build.
Moreover, it now also supports the Limited API.
Change Py_TRACE_REFS build:
* Remove _PyObject_EXTRA_INIT macro.
* The PyObject structure no longer has two extra members (_ob_prev
and _ob_next).
* Use a hash table (_Py_hashtable_t) to trace references (all
objects): PyInterpreterState.object_state.refchain.
* Py_TRACE_REFS build is now ABI compatible with release build and
debug build.
* Limited C API extensions can now be built with Py_TRACE_REFS:
xxlimited, xxlimited_35, _testclinic_limited.
* No longer rename PyModule_Create2() and PyModule_FromDefAndSpec2()
functions to PyModule_Create2TraceRefs() and
PyModule_FromDefAndSpec2TraceRefs().
* _Py_PrintReferenceAddresses() is now called before
finalize_interp_delete() which deletes the refchain hash table.
* test_tracemalloc find_trace() now also filters by size to ignore
the memory allocated by _PyRefchain_Trace().
Test changes for Py_TRACE_REFS:
* Add test.support.Py_TRACE_REFS constant.
* Add test_sys.test_getobjects() to test sys.getobjects() function.
* test_exceptions skips test_recursion_normalizing_with_no_memory()
and test_memory_error_in_PyErr_PrintEx() if Python is built with
Py_TRACE_REFS.
* test_repl skips test_no_memory().
* test_capi skisp test_set_nomemory().
A bunch of code currently uses PyInterpreterState.modules directly instead of PyImport_GetModuleDict(). This complicates efforts to make changes relative to sys.modules. This patch switches to using PyImport_GetModuleDict() uniformly. Also, a number of related uses of sys.modules are updated for uniformity for the same reason.
Note that this code was already reviewed and merged as part of #1638. I reverted that and am now splitting it up into more focused parts.
PR #1638, for bpo-28411, causes problems in some (very) edge cases. Until that gets sorted out, we're reverting the merge. PR #3506, a fix on top of #1638, is also getting reverted.
PyArg_VaParse() and PyArg_VaParseTupleAndKeywords() were not available in
limited API. PyArg_ValidateKeywordArguments(), PyArg_UnpackTuple() and
Py_BuildValue() were not available in limited API of version < 3.3 when
PY_SSIZE_T_CLEAN is defined.
PyArg_VaParse() and PyArg_VaParseTupleAndKeywords() were not available in
limited API. PyArg_ValidateKeywordArguments(), PyArg_UnpackTuple() and
Py_BuildValue() were not available in limited API of version < 3.3 when
PY_SSIZE_T_CLEAN is defined.
PyArg_VaParse() and PyArg_VaParseTupleAndKeywords() were not available in
limited API. PyArg_ValidateKeywordArguments(), PyArg_UnpackTuple() and
Py_BuildValue() were not available in limited API of version < 3.3 when
PY_SSIZE_T_CLEAN is defined.
Issue #27810:
* Modify vgetargskeywordsfast() to work on a C array of PyObject* rather than
working on a tuple directly.
* Add _PyArg_ParseStack()
* Argument Clinic now emits code using the new METH_FASTCALL calling convention
Known limitations of the current implementation:
- documentation changes are incomplete
- there's a reference leak I haven't tracked down yet
The leak is most visible by running:
./python -m test -R3:3 test_importlib
However, you can also see it by running:
./python -X showrefcount
Importing the array or _testmultiphase modules, and
then deleting them from both sys.modules and the local
namespace shows significant increases in the total
number of active references each cycle. By contrast,
with _testcapi (which continues to use single-phase
initialisation) the global refcounts stabilise after
a couple of cycles.
svn+ssh://pythondev@svn.python.org/python/trunk
........
r60178 | georg.brandl | 2008-01-21 22:05:49 +0100 (Mon, 21 Jan 2008) | 2 lines
#1715: include sub-extension modules in pydoc text output.
........
r60179 | georg.brandl | 2008-01-21 22:14:21 +0100 (Mon, 21 Jan 2008) | 2 lines
Add a "const" to make gcc happy.
........
r60180 | georg.brandl | 2008-01-21 22:19:07 +0100 (Mon, 21 Jan 2008) | 2 lines
Add the correct build dir when building with pydebug.
........
r60181 | georg.brandl | 2008-01-21 22:23:15 +0100 (Mon, 21 Jan 2008) | 3 lines
Patch #1720595: add T_BOOL to the range of structmember types.
Patch by Angelo Mottola, reviewed by MvL, tests by me.
........
r60182 | georg.brandl | 2008-01-21 22:28:32 +0100 (Mon, 21 Jan 2008) | 2 lines
Reformat some ugly code.
........
r60187 | brett.cannon | 2008-01-22 00:50:16 +0100 (Tue, 22 Jan 2008) | 4 lines
Make's MAKEFLAGS variable is set to a string containing the single-letter
arguments to Make. This means there are no hyphens. Fix the '-s' check to
silence distutils to now work.
........
r60188 | gregory.p.smith | 2008-01-22 01:19:41 +0100 (Tue, 22 Jan 2008) | 3 lines
accepts and closes issue #1221598: adds an optional callback to ftplib.FTP
storbinary() and storlines() methods.
........
r60189 | gregory.p.smith | 2008-01-22 02:12:02 +0100 (Tue, 22 Jan 2008) | 2 lines
Replace spam.acquire() try: ... finally: spam.release() with "with spam:"
........
r60190 | gregory.p.smith | 2008-01-22 02:20:42 +0100 (Tue, 22 Jan 2008) | 4 lines
- Fix Issue #1703448: A joined thread could show up in the
threading.enumerate() list after the join() for a brief period until
it actually exited.
........
r60193 | georg.brandl | 2008-01-22 08:53:31 +0100 (Tue, 22 Jan 2008) | 2 lines
Fix \xhh specs, #1889.
........
r60198 | christian.heimes | 2008-01-22 16:01:25 +0100 (Tue, 22 Jan 2008) | 1 line
Fixed a missing (X) in define
........
r60199 | christian.heimes | 2008-01-22 16:25:18 +0100 (Tue, 22 Jan 2008) | 2 lines
Don't repeat yourself
Added the macros PyModule_AddIntMacro and PyModule_AddStringMacro. They shorten PyModule_AddIntConstant(m, "AF_INET", AF_INET) to PyModule_AddIntMacro(m, AF_INET)
........
r60201 | raymond.hettinger | 2008-01-22 20:51:41 +0100 (Tue, 22 Jan 2008) | 1 line
Document when to use izip_longest().
........
r60202 | georg.brandl | 2008-01-22 20:56:03 +0100 (Tue, 22 Jan 2008) | 2 lines
Fix for #1087741 patch.
........
r60203 | raymond.hettinger | 2008-01-22 21:18:53 +0100 (Tue, 22 Jan 2008) | 1 line
Give zip() the same guarantee as izip() for left-to-right evaluation.
........
r60204 | raymond.hettinger | 2008-01-22 23:09:26 +0100 (Tue, 22 Jan 2008) | 1 line
Improve variable name in sample code
........
r60205 | gregory.p.smith | 2008-01-23 00:15:34 +0100 (Wed, 23 Jan 2008) | 2 lines
docstring and comment updates suggested by Giampaolo Rodola'
........
r60207 | raymond.hettinger | 2008-01-23 01:04:40 +0100 (Wed, 23 Jan 2008) | 1 line
Let pprint() support sets and frozensets (suggested by David Mertz).
........
r60208 | guido.van.rossum | 2008-01-23 02:18:27 +0100 (Wed, 23 Jan 2008) | 4 lines
I'm tired of these tests breaking at Google due to our large number of
users and groups in LDAP/NIS. So I'm limiting the extra-heavy part of
the tests to passwd/group files with at most 1000 entries.
........
Victor Stinner
Serhiy Storchaka
Inada Naoki
Batuhan Taskaya
Dong-hee Na
Rémi Lapeyre
Eric Snow
Victor Stinner
Yury Selivanov
Nick Coghlan
Larry Hastings
Benjamin Peterson
Martin v. Löwis
Amaury Forgeot d'Arc
Christian Heimes