import python3-3.6.8-4.el8_0

a8 imports/c8/python3-3.6.8-4.el8_0
CentOS Sources 4 years ago committed by Andrew Lukoshko
commit 10e0f3ba3c

1
.gitignore vendored

@ -0,0 +1 @@
SOURCES/Python-3.6.8-noexe.tar.xz

@ -0,0 +1 @@
a39802ac8f0c61645c6a50fbdd32e3ca92862ff5 SOURCES/Python-3.6.8-noexe.tar.xz

@ -0,0 +1,19 @@
diff -up Python-3.1.1/Lib/distutils/unixccompiler.py.rpath Python-3.1.1/Lib/distutils/unixccompiler.py
--- Python-3.1.1/Lib/distutils/unixccompiler.py.rpath 2009-09-04 17:29:34.000000000 -0400
+++ Python-3.1.1/Lib/distutils/unixccompiler.py 2009-09-04 17:49:54.000000000 -0400
@@ -141,6 +141,15 @@ class UnixCCompiler(CCompiler):
if sys.platform == "cygwin":
exe_extension = ".exe"
+ def _fix_lib_args(self, libraries, library_dirs, runtime_library_dirs):
+ """Remove standard library path from rpath"""
+ libraries, library_dirs, runtime_library_dirs = super()._fix_lib_args(
+ libraries, library_dirs, runtime_library_dirs)
+ libdir = sysconfig.get_config_var('LIBDIR')
+ if runtime_library_dirs and (libdir in runtime_library_dirs):
+ runtime_library_dirs.remove(libdir)
+ return libraries, library_dirs, runtime_library_dirs
+
def preprocess(self, source, output_file=None, macros=None,
include_dirs=None, extra_preargs=None, extra_postargs=None):
fixed_args = self._fix_compile_args(None, macros, include_dirs)

@ -0,0 +1,202 @@
diff --git a/Lib/distutils/command/install.py b/Lib/distutils/command/install.py
index 9474e9c..c0ce4c6 100644
--- a/Lib/distutils/command/install.py
+++ b/Lib/distutils/command/install.py
@@ -30,14 +30,14 @@ WINDOWS_SCHEME = {
INSTALL_SCHEMES = {
'unix_prefix': {
'purelib': '$base/lib/python$py_version_short/site-packages',
- 'platlib': '$platbase/lib/python$py_version_short/site-packages',
+ 'platlib': '$platbase/lib64/python$py_version_short/site-packages',
'headers': '$base/include/python$py_version_short$abiflags/$dist_name',
'scripts': '$base/bin',
'data' : '$base',
},
'unix_home': {
'purelib': '$base/lib/python',
- 'platlib': '$base/lib/python',
+ 'platlib': '$base/lib64/python',
'headers': '$base/include/python/$dist_name',
'scripts': '$base/bin',
'data' : '$base',
diff --git a/Lib/distutils/sysconfig.py b/Lib/distutils/sysconfig.py
index 026cca7..6d3e077 100644
--- a/Lib/distutils/sysconfig.py
+++ b/Lib/distutils/sysconfig.py
@@ -132,8 +132,12 @@ def get_python_lib(plat_specific=0, standard_lib=0, prefix=None):
prefix = plat_specific and EXEC_PREFIX or PREFIX
if os.name == "posix":
+ if plat_specific or standard_lib:
+ lib = "lib64"
+ else:
+ lib = "lib"
libpython = os.path.join(prefix,
- "lib", "python" + get_python_version())
+ lib, "python" + get_python_version())
if standard_lib:
return libpython
else:
diff a/Lib/distutils/tests/test_install.py b/Lib/distutils/tests/test_install.py
--- a/Lib/distutils/tests/test_install.py
+++ b/Lib/distutils/tests/test_install.py
@@ -57,8 +57,9 @@
self.assertEqual(got, expected)
libdir = os.path.join(destination, "lib", "python")
+ platlibdir = os.path.join(destination, "lib64", "python")
check_path(cmd.install_lib, libdir)
- check_path(cmd.install_platlib, libdir)
+ check_path(cmd.install_platlib, platlibdir)
check_path(cmd.install_purelib, libdir)
check_path(cmd.install_headers,
os.path.join(destination, "include", "python", "foopkg"))
diff --git a/Lib/site.py b/Lib/site.py
index a84e3bb..ba0d3ea 100644
--- a/Lib/site.py
+++ b/Lib/site.py
@@ -303,11 +303,15 @@ def getsitepackages(prefixes=None):
seen.add(prefix)
if os.sep == '/':
+ sitepackages.append(os.path.join(prefix, "lib64",
+ "python" + sys.version[:3],
+ "site-packages"))
sitepackages.append(os.path.join(prefix, "lib",
"python%d.%d" % sys.version_info[:2],
"site-packages"))
else:
sitepackages.append(prefix)
+ sitepackages.append(os.path.join(prefix, "lib64", "site-packages"))
sitepackages.append(os.path.join(prefix, "lib", "site-packages"))
if sys.platform == "darwin":
# for framework builds *only* we add the standard Apple
diff --git a/Lib/sysconfig.py b/Lib/sysconfig.py
index b9bbfe5..2a5f29c 100644
--- a/Lib/sysconfig.py
+++ b/Lib/sysconfig.py
@@ -20,10 +20,10 @@ __all__ = [
_INSTALL_SCHEMES = {
'posix_prefix': {
- 'stdlib': '{installed_base}/lib/python{py_version_short}',
- 'platstdlib': '{platbase}/lib/python{py_version_short}',
+ 'stdlib': '{installed_base}/lib64/python{py_version_short}',
+ 'platstdlib': '{platbase}/lib64/python{py_version_short}',
'purelib': '{base}/lib/python{py_version_short}/site-packages',
- 'platlib': '{platbase}/lib/python{py_version_short}/site-packages',
+ 'platlib': '{platbase}/lib64/python{py_version_short}/site-packages',
'include':
'{installed_base}/include/python{py_version_short}{abiflags}',
'platinclude':
@@ -61,10 +61,10 @@ _INSTALL_SCHEMES = {
'data': '{userbase}',
},
'posix_user': {
- 'stdlib': '{userbase}/lib/python{py_version_short}',
- 'platstdlib': '{userbase}/lib/python{py_version_short}',
+ 'stdlib': '{userbase}/lib64/python{py_version_short}',
+ 'platstdlib': '{userbase}/lib64/python{py_version_short}',
'purelib': '{userbase}/lib/python{py_version_short}/site-packages',
- 'platlib': '{userbase}/lib/python{py_version_short}/site-packages',
+ 'platlib': '{userbase}/lib64/python{py_version_short}/site-packages',
'include': '{userbase}/include/python{py_version_short}',
'scripts': '{userbase}/bin',
'data': '{userbase}',
diff --git a/Lib/test/test_site.py b/Lib/test/test_site.py
index f698927..bc977b5 100644
--- a/Lib/test/test_site.py
+++ b/Lib/test/test_site.py
@@ -248,8 +248,8 @@ class HelperFunctionsTests(unittest.TestCase):
self.assertEqual(dirs[1], wanted)
elif os.sep == '/':
# OS X non-framework builds, Linux, FreeBSD, etc
- self.assertEqual(len(dirs), 1)
- wanted = os.path.join('xoxo', 'lib',
+ self.assertEqual(len(dirs), 2)
+ wanted = os.path.join('xoxo', 'lib64',
'python%d.%d' % sys.version_info[:2],
'site-packages')
self.assertEqual(dirs[0], wanted)
diff --git a/Makefile.pre.in b/Makefile.pre.in
index 8fa7934..a693917 100644
--- a/Makefile.pre.in
+++ b/Makefile.pre.in
@@ -126,7 +126,7 @@ LIBDIR= @libdir@
MANDIR= @mandir@
INCLUDEDIR= @includedir@
CONFINCLUDEDIR= $(exec_prefix)/include
-SCRIPTDIR= $(prefix)/lib
+SCRIPTDIR= $(prefix)/lib64
ABIFLAGS= @ABIFLAGS@
# Detailed destination directories
diff --git a/Modules/getpath.c b/Modules/getpath.c
index 65b47a3..eaa756c 100644
--- a/Modules/getpath.c
+++ b/Modules/getpath.c
@@ -494,7 +494,7 @@ calculate_path(void)
_pythonpath = Py_DecodeLocale(PYTHONPATH, NULL);
_prefix = Py_DecodeLocale(PREFIX, NULL);
_exec_prefix = Py_DecodeLocale(EXEC_PREFIX, NULL);
- lib_python = Py_DecodeLocale("lib/python" VERSION, NULL);
+ lib_python = Py_DecodeLocale("lib64/python" VERSION, NULL);
if (!_pythonpath || !_prefix || !_exec_prefix || !lib_python) {
Py_FatalError(
@@ -683,7 +683,7 @@ calculate_path(void)
}
else
wcsncpy(zip_path, _prefix, MAXPATHLEN);
- joinpath(zip_path, L"lib/python00.zip");
+ joinpath(zip_path, L"lib64/python00.zip");
bufsz = wcslen(zip_path); /* Replace "00" with version */
zip_path[bufsz - 6] = VERSION[0];
zip_path[bufsz - 5] = VERSION[2];
@@ -695,7 +695,7 @@ calculate_path(void)
fprintf(stderr,
"Could not find platform dependent libraries <exec_prefix>\n");
wcsncpy(exec_prefix, _exec_prefix, MAXPATHLEN);
- joinpath(exec_prefix, L"lib/lib-dynload");
+ joinpath(exec_prefix, L"lib64/lib-dynload");
}
/* If we found EXEC_PREFIX do *not* reduce it! (Yet.) */
diff --git a/setup.py b/setup.py
index 0f2dfc4..da37896 100644
--- a/setup.py
+++ b/setup.py
@@ -492,7 +492,7 @@ class PyBuildExt(build_ext):
# directories (i.e. '.' and 'Include') must be first. See issue
# 10520.
if not cross_compiling:
- add_dir_to_list(self.compiler.library_dirs, '/usr/local/lib')
+ add_dir_to_list(self.compiler.library_dirs, '/usr/local/lib64')
add_dir_to_list(self.compiler.include_dirs, '/usr/local/include')
# only change this for cross builds for 3.3, issues on Mageia
if cross_compiling:
@@ -780,11 +780,11 @@ class PyBuildExt(build_ext):
elif curses_library:
readline_libs.append(curses_library)
elif self.compiler.find_library_file(lib_dirs +
- ['/usr/lib/termcap'],
+ ['/usr/lib64/termcap'],
'termcap'):
readline_libs.append('termcap')
exts.append( Extension('readline', ['readline.c'],
- library_dirs=['/usr/lib/termcap'],
+ library_dirs=['/usr/lib64/termcap'],
extra_link_args=readline_extra_link_args,
libraries=readline_libs) )
else:
@@ -821,8 +821,8 @@ class PyBuildExt(build_ext):
if krb5_h:
ssl_incs += krb5_h
ssl_libs = find_library_file(self.compiler, 'ssl',lib_dirs,
- ['/usr/local/ssl/lib',
- '/usr/contrib/ssl/lib/'
+ ['/usr/local/ssl/lib64',
+ '/usr/contrib/ssl/lib64/'
] )
if (ssl_incs is not None and

@ -0,0 +1,53 @@
diff --git a/Makefile.pre.in b/Makefile.pre.in
index 9cd482f..b074b26 100644
--- a/Makefile.pre.in
+++ b/Makefile.pre.in
@@ -549,7 +549,7 @@ clinic: check-clean-src $(srcdir)/Modules/_blake2/blake2s_impl.c
$(PYTHON_FOR_REGEN) ./Tools/clinic/clinic.py --make
# Build the interpreter
-$(BUILDPYTHON): Programs/python.o $(LIBRARY) $(LDLIBRARY) $(PY3LIBRARY)
+$(BUILDPYTHON): Programs/python.o $(LDLIBRARY) $(PY3LIBRARY)
$(LINKCC) $(PY_CORE_LDFLAGS) $(LINKFORSHARED) -o $@ Programs/python.o $(BLDLIBRARY) $(LIBS) $(MODLIBS) $(SYSLIBS) $(LDLAST)
platform: $(BUILDPYTHON) pybuilddir.txt
@@ -597,12 +597,6 @@ sharedmods: $(BUILDPYTHON) pybuilddir.txt Modules/_math.o
_TCLTK_INCLUDES='$(TCLTK_INCLUDES)' _TCLTK_LIBS='$(TCLTK_LIBS)' \
$(PYTHON_FOR_BUILD) $(srcdir)/setup.py $$quiet build
-
-# Build static library
-$(LIBRARY): $(LIBRARY_OBJS)
- -rm -f $@
- $(AR) $(ARFLAGS) $@ $(LIBRARY_OBJS)
-
libpython$(LDVERSION).so: $(LIBRARY_OBJS)
if test $(INSTSONAME) != $(LDLIBRARY); then \
$(BLDSHARED) -Wl,-h$(INSTSONAME) -o $(INSTSONAME) $(LIBRARY_OBJS) $(MODLIBS) $(SHLIBS) $(LIBC) $(LIBM) $(LDLAST); \
@@ -692,7 +686,7 @@ Modules/Setup: $(srcdir)/Modules/Setup.dist
echo "-----------------------------------------------"; \
fi
-Programs/_testembed: Programs/_testembed.o $(LIBRARY) $(LDLIBRARY) $(PY3LIBRARY)
+Programs/_testembed: Programs/_testembed.o $(LDLIBRARY) $(PY3LIBRARY)
$(LINKCC) $(PY_CORE_LDFLAGS) $(LINKFORSHARED) -o $@ Programs/_testembed.o $(BLDLIBRARY) $(LIBS) $(MODLIBS) $(SYSLIBS) $(LDLAST)
############################################################################
@@ -1428,17 +1422,6 @@ libainstall: @DEF_MAKE_RULE@ python-config
else true; \
fi; \
done
- @if test -d $(LIBRARY); then :; else \
- if test "$(PYTHONFRAMEWORKDIR)" = no-framework; then \
- if test "$(SHLIB_SUFFIX)" = .dll; then \
- $(INSTALL_DATA) $(LDLIBRARY) $(DESTDIR)$(LIBPL) ; \
- else \
- $(INSTALL_DATA) $(LIBRARY) $(DESTDIR)$(LIBPL)/$(LIBRARY) ; \
- fi; \
- else \
- echo Skip install of $(LIBRARY) - use make frameworkinstall; \
- fi; \
- fi
$(INSTALL_DATA) Modules/config.c $(DESTDIR)$(LIBPL)/config.c
$(INSTALL_DATA) Programs/python.o $(DESTDIR)$(LIBPL)/python.o
$(INSTALL_DATA) $(srcdir)/Modules/config.c.in $(DESTDIR)$(LIBPL)/config.c.in

@ -0,0 +1,46 @@
diff -up Python-3.2.2/Lib/unittest/case.py.add-rpmbuild-hooks-to-unittest Python-3.2.2/Lib/unittest/case.py
--- Python-3.2.2/Lib/unittest/case.py.add-rpmbuild-hooks-to-unittest 2011-09-03 12:16:44.000000000 -0400
+++ Python-3.2.2/Lib/unittest/case.py 2011-09-09 06:35:16.365568382 -0400
@@ -3,6 +3,7 @@
import sys
import functools
import difflib
+import os
import logging
import pprint
import re
@@ -101,5 +102,21 @@ def expectedFailure(func):
raise self.test_case.failureException(msg)
+# Non-standard/downstream-only hooks for handling issues with specific test
+# cases:
+
+def _skipInRpmBuild(reason):
+ """
+ Non-standard/downstream-only decorator for marking a specific unit test
+ to be skipped when run within the %check of an rpmbuild.
+
+ Specifically, this takes effect when WITHIN_PYTHON_RPM_BUILD is set within
+ the environment, and has no effect otherwise.
+ """
+ if 'WITHIN_PYTHON_RPM_BUILD' in os.environ:
+ return skip(reason)
+ else:
+ return _id
+
class _AssertRaisesBaseContext(_BaseTestCaseContext):
def __init__(self, expected, test_case, expected_regex=None):
diff -up Python-3.2.2/Lib/unittest/__init__.py.add-rpmbuild-hooks-to-unittest Python-3.2.2/Lib/unittest/__init__.py
--- Python-3.2.2/Lib/unittest/__init__.py.add-rpmbuild-hooks-to-unittest 2011-09-03 12:16:44.000000000 -0400
+++ Python-3.2.2/Lib/unittest/__init__.py 2011-09-09 06:35:16.366568382 -0400
@@ -57,7 +57,8 @@ __unittest = True
from .result import TestResult
from .case import (TestCase, FunctionTestCase, SkipTest, skip, skipIf,
- skipUnless, expectedFailure)
+ skipUnless, expectedFailure,
+ _skipInRpmBuild)
from .suite import BaseTestSuite, TestSuite
from .loader import (TestLoader, defaultTestLoader, makeSuite, getTestCaseNames,
findTestCases)

@ -0,0 +1,15 @@
diff -up Python-3.2.3/Lib/ctypes/__init__.py.rhbz814391 Python-3.2.3/Lib/ctypes/__init__.py
--- Python-3.2.3/Lib/ctypes/__init__.py.rhbz814391 2012-04-20 15:12:49.017867692 -0400
+++ Python-3.2.3/Lib/ctypes/__init__.py 2012-04-20 15:15:09.501111408 -0400
@@ -275,11 +275,6 @@ def _reset_cache():
# _SimpleCData.c_char_p_from_param
POINTER(c_char).from_param = c_char_p.from_param
_pointer_type_cache[None] = c_void_p
- # XXX for whatever reasons, creating the first instance of a callback
- # function is needed for the unittests on Win64 to succeed. This MAY
- # be a compiler bug, since the problem occurs only when _ctypes is
- # compiled with the MS SDK compiler. Or an uninitialized variable?
- CFUNCTYPE(c_int)(lambda: None)
def create_unicode_buffer(init, size=None):
"""create_unicode_buffer(aString) -> character array

@ -0,0 +1,11 @@
diff -up cpython-59223da36dec/Lib/test/test_posix.py.disable-test_fs_holes-in-rpm-build cpython-59223da36dec/Lib/test/test_posix.py
--- cpython-59223da36dec/Lib/test/test_posix.py.disable-test_fs_holes-in-rpm-build 2012-08-07 17:15:59.000000000 -0400
+++ cpython-59223da36dec/Lib/test/test_posix.py 2012-08-07 17:16:53.528330330 -0400
@@ -973,6 +973,7 @@ class PosixTester(unittest.TestCase):
posix.RTLD_GLOBAL
posix.RTLD_LOCAL
+ @unittest._skipInRpmBuild('running kernel may not match kernel in chroot')
@unittest.skipUnless(hasattr(os, 'SEEK_HOLE'),
"test needs an OS that reports file holes")
def test_fs_holes(self):

@ -0,0 +1,11 @@
diff -up Python-3.3.0b1/Lib/test/test_socket.py.disable-test_socket-in-rpm-builds Python-3.3.0b1/Lib/test/test_socket.py
--- Python-3.3.0b1/Lib/test/test_socket.py.disable-test_socket-in-rpm-builds 2012-07-24 15:02:30.823355067 -0400
+++ Python-3.3.0b1/Lib/test/test_socket.py 2012-07-24 15:08:13.021354999 -0400
@@ -2188,6 +2188,7 @@ class RecvmsgGenericStreamTests(RecvmsgG
# Tests which require a stream socket and can use either recvmsg()
# or recvmsg_into().
+ @unittest._skipInRpmBuild('fails intermittently when run within Koji')
def testRecvmsgEOF(self):
# Receive end-of-stream indicator (b"", peer socket closed).
msg, ancdata, flags, addr = self.doRecvmsg(self.serv_sock, 1024)

@ -0,0 +1,309 @@
diff --git a/Include/object.h b/Include/object.h
index 63e37b8..613b26c 100644
--- a/Include/object.h
+++ b/Include/object.h
@@ -1071,6 +1071,49 @@ PyAPI_FUNC(void)
_PyObject_DebugTypeStats(FILE *out);
#endif /* ifndef Py_LIMITED_API */
+/*
+ Define a pair of assertion macros.
+
+ These work like the regular C assert(), in that they will abort the
+ process with a message on stderr if the given condition fails to hold,
+ but compile away to nothing if NDEBUG is defined.
+
+ However, before aborting, Python will also try to call _PyObject_Dump() on
+ the given object. This may be of use when investigating bugs in which a
+ particular object is corrupt (e.g. buggy a tp_visit method in an extension
+ module breaking the garbage collector), to help locate the broken objects.
+
+ The WITH_MSG variant allows you to supply an additional message that Python
+ will attempt to print to stderr, after the object dump.
+*/
+#ifdef NDEBUG
+/* No debugging: compile away the assertions: */
+#define PyObject_ASSERT_WITH_MSG(obj, expr, msg) ((void)0)
+#else
+/* With debugging: generate checks: */
+#define PyObject_ASSERT_WITH_MSG(obj, expr, msg) \
+ ((expr) \
+ ? (void)(0) \
+ : _PyObject_AssertFailed((obj), \
+ (msg), \
+ (__STRING(expr)), \
+ (__FILE__), \
+ (__LINE__), \
+ (__PRETTY_FUNCTION__)))
+#endif
+
+#define PyObject_ASSERT(obj, expr) \
+ PyObject_ASSERT_WITH_MSG(obj, expr, NULL)
+
+/*
+ Declare and define the entrypoint even when NDEBUG is defined, to avoid
+ causing compiler/linker errors when building extensions without NDEBUG
+ against a Python built with NDEBUG defined
+*/
+PyAPI_FUNC(void) _PyObject_AssertFailed(PyObject *, const char *,
+ const char *, const char *, int,
+ const char *);
+
#ifdef __cplusplus
}
#endif
diff --git a/Lib/test/test_gc.py b/Lib/test/test_gc.py
index 7e82b24..8ecc3d9 100644
--- a/Lib/test/test_gc.py
+++ b/Lib/test/test_gc.py
@@ -2,9 +2,11 @@ import unittest
from test.support import (verbose, refcount_test, run_unittest,
strip_python_stderr, cpython_only, start_threads,
temp_dir, requires_type_collecting, TESTFN, unlink)
+from test.support import import_module
from test.support.script_helper import assert_python_ok, make_script
import sys
+import sysconfig
import time
import gc
import weakref
@@ -50,6 +52,8 @@ class GC_Detector(object):
# gc collects it.
self.wr = weakref.ref(C1055820(666), it_happened)
+BUILD_WITH_NDEBUG = ('-DNDEBUG' in sysconfig.get_config_vars()['PY_CFLAGS'])
+
@with_tp_del
class Uncollectable(object):
"""Create a reference cycle with multiple __del__ methods.
@@ -877,6 +881,50 @@ class GCCallbackTests(unittest.TestCase):
self.assertEqual(len(gc.garbage), 0)
+ @unittest.skipIf(BUILD_WITH_NDEBUG,
+ 'built with -NDEBUG')
+ def test_refcount_errors(self):
+ self.preclean()
+ # Verify the "handling" of objects with broken refcounts
+ import_module("ctypes") #skip if not supported
+
+ import subprocess
+ code = '''if 1:
+ a = []
+ b = [a]
+
+ # Simulate the refcount of "a" being too low (compared to the
+ # references held on it by live data), but keeping it above zero
+ # (to avoid deallocating it):
+ import ctypes
+ ctypes.pythonapi.Py_DecRef(ctypes.py_object(a))
+
+ # The garbage collector should now have a fatal error when it reaches
+ # the broken object:
+ import gc
+ gc.collect()
+ '''
+ p = subprocess.Popen([sys.executable, "-c", code],
+ stdout=subprocess.PIPE,
+ stderr=subprocess.PIPE)
+ stdout, stderr = p.communicate()
+ p.stdout.close()
+ p.stderr.close()
+ # Verify that stderr has a useful error message:
+ self.assertRegex(stderr,
+ b'Modules/gcmodule.c:[0-9]+: visit_decref: Assertion "\(\(gc\)->gc.gc_refs >> \(1\)\) != 0" failed.')
+ self.assertRegex(stderr,
+ b'refcount was too small')
+ self.assertRegex(stderr,
+ b'object : \[\]')
+ self.assertRegex(stderr,
+ b'type : list')
+ self.assertRegex(stderr,
+ b'refcount: 1')
+ self.assertRegex(stderr,
+ b'address : 0x[0-9a-f]+')
+
+
class GCTogglingTests(unittest.TestCase):
def setUp(self):
gc.enable()
diff --git a/Modules/gcmodule.c b/Modules/gcmodule.c
index 3bddc40..0cc24f7 100644
--- a/Modules/gcmodule.c
+++ b/Modules/gcmodule.c
@@ -342,7 +342,8 @@ update_refs(PyGC_Head *containers)
{
PyGC_Head *gc = containers->gc.gc_next;
for (; gc != containers; gc = gc->gc.gc_next) {
- assert(_PyGCHead_REFS(gc) == GC_REACHABLE);
+ PyObject_ASSERT(FROM_GC(gc),
+ _PyGCHead_REFS(gc) == GC_REACHABLE);
_PyGCHead_SET_REFS(gc, Py_REFCNT(FROM_GC(gc)));
/* Python's cyclic gc should never see an incoming refcount
* of 0: if something decref'ed to 0, it should have been
@@ -362,7 +363,8 @@ update_refs(PyGC_Head *containers)
* so serious that maybe this should be a release-build
* check instead of an assert?
*/
- assert(_PyGCHead_REFS(gc) != 0);
+ PyObject_ASSERT(FROM_GC(gc),
+ _PyGCHead_REFS(gc) != 0);
}
}
@@ -377,7 +379,9 @@ visit_decref(PyObject *op, void *data)
* generation being collected, which can be recognized
* because only they have positive gc_refs.
*/
- assert(_PyGCHead_REFS(gc) != 0); /* else refcount was too small */
+ PyObject_ASSERT_WITH_MSG(FROM_GC(gc),
+ _PyGCHead_REFS(gc) != 0,
+ "refcount was too small"); /* else refcount was too small */
if (_PyGCHead_REFS(gc) > 0)
_PyGCHead_DECREF(gc);
}
@@ -437,9 +441,10 @@ visit_reachable(PyObject *op, PyGC_Head *reachable)
* If gc_refs == GC_UNTRACKED, it must be ignored.
*/
else {
- assert(gc_refs > 0
- || gc_refs == GC_REACHABLE
- || gc_refs == GC_UNTRACKED);
+ PyObject_ASSERT(FROM_GC(gc),
+ gc_refs > 0
+ || gc_refs == GC_REACHABLE
+ || gc_refs == GC_UNTRACKED);
}
}
return 0;
@@ -481,7 +486,7 @@ move_unreachable(PyGC_Head *young, PyGC_Head *unreachable)
*/
PyObject *op = FROM_GC(gc);
traverseproc traverse = Py_TYPE(op)->tp_traverse;
- assert(_PyGCHead_REFS(gc) > 0);
+ PyObject_ASSERT(op, _PyGCHead_REFS(gc) > 0);
_PyGCHead_SET_REFS(gc, GC_REACHABLE);
(void) traverse(op,
(visitproc)visit_reachable,
@@ -544,7 +549,7 @@ move_legacy_finalizers(PyGC_Head *unreachable, PyGC_Head *finalizers)
for (gc = unreachable->gc.gc_next; gc != unreachable; gc = next) {
PyObject *op = FROM_GC(gc);
- assert(IS_TENTATIVELY_UNREACHABLE(op));
+ PyObject_ASSERT(op, IS_TENTATIVELY_UNREACHABLE(op));
next = gc->gc.gc_next;
if (has_legacy_finalizer(op)) {
@@ -620,7 +625,7 @@ handle_weakrefs(PyGC_Head *unreachable, PyGC_Head *old)
PyWeakReference **wrlist;
op = FROM_GC(gc);
- assert(IS_TENTATIVELY_UNREACHABLE(op));
+ PyObject_ASSERT(op, IS_TENTATIVELY_UNREACHABLE(op));
next = gc->gc.gc_next;
if (! PyType_SUPPORTS_WEAKREFS(Py_TYPE(op)))
@@ -641,9 +646,9 @@ handle_weakrefs(PyGC_Head *unreachable, PyGC_Head *old)
* the callback pointer intact. Obscure: it also
* changes *wrlist.
*/
- assert(wr->wr_object == op);
+ PyObject_ASSERT(wr->wr_object, wr->wr_object == op);
_PyWeakref_ClearRef(wr);
- assert(wr->wr_object == Py_None);
+ PyObject_ASSERT(wr->wr_object, wr->wr_object == Py_None);
if (wr->wr_callback == NULL)
continue; /* no callback */
@@ -677,7 +682,7 @@ handle_weakrefs(PyGC_Head *unreachable, PyGC_Head *old)
*/
if (IS_TENTATIVELY_UNREACHABLE(wr))
continue;
- assert(IS_REACHABLE(wr));
+ PyObject_ASSERT(op, IS_REACHABLE(wr));
/* Create a new reference so that wr can't go away
* before we can process it again.
@@ -686,7 +691,8 @@ handle_weakrefs(PyGC_Head *unreachable, PyGC_Head *old)
/* Move wr to wrcb_to_call, for the next pass. */
wrasgc = AS_GC(wr);
- assert(wrasgc != next); /* wrasgc is reachable, but
+ PyObject_ASSERT(op, wrasgc != next);
+ /* wrasgc is reachable, but
next isn't, so they can't
be the same */
gc_list_move(wrasgc, &wrcb_to_call);
@@ -702,11 +708,11 @@ handle_weakrefs(PyGC_Head *unreachable, PyGC_Head *old)
gc = wrcb_to_call.gc.gc_next;
op = FROM_GC(gc);
- assert(IS_REACHABLE(op));
- assert(PyWeakref_Check(op));
+ PyObject_ASSERT(op, IS_REACHABLE(op));
+ PyObject_ASSERT(op, PyWeakref_Check(op));
wr = (PyWeakReference *)op;
callback = wr->wr_callback;
- assert(callback != NULL);
+ PyObject_ASSERT(op, callback != NULL);
/* copy-paste of weakrefobject.c's handle_callback() */
temp = PyObject_CallFunctionObjArgs(callback, wr, NULL);
@@ -820,12 +826,14 @@ check_garbage(PyGC_Head *collectable)
for (gc = collectable->gc.gc_next; gc != collectable;
gc = gc->gc.gc_next) {
_PyGCHead_SET_REFS(gc, Py_REFCNT(FROM_GC(gc)));
- assert(_PyGCHead_REFS(gc) != 0);
+ PyObject_ASSERT(FROM_GC(gc),
+ _PyGCHead_REFS(gc) != 0);
}
subtract_refs(collectable);
for (gc = collectable->gc.gc_next; gc != collectable;
gc = gc->gc.gc_next) {
- assert(_PyGCHead_REFS(gc) >= 0);
+ PyObject_ASSERT(FROM_GC(gc),
+ _PyGCHead_REFS(gc) >= 0);
if (_PyGCHead_REFS(gc) != 0)
return -1;
}
diff --git a/Objects/object.c b/Objects/object.c
index fdd41a6..bfe806c 100644
--- a/Objects/object.c
+++ b/Objects/object.c
@@ -2031,6 +2031,35 @@ _PyTrash_thread_destroy_chain(void)
}
}
+PyAPI_FUNC(void)
+_PyObject_AssertFailed(PyObject *obj, const char *msg, const char *expr,
+ const char *file, int line, const char *function)
+{
+ fprintf(stderr,
+ "%s:%d: %s: Assertion \"%s\" failed.\n",
+ file, line, function, expr);
+ if (msg) {
+ fprintf(stderr, "%s\n", msg);
+ }
+
+ fflush(stderr);
+
+ if (obj) {
+ /* This might succeed or fail, but we're about to abort, so at least
+ try to provide any extra info we can: */
+ _PyObject_Dump(obj);
+ }
+ else {
+ fprintf(stderr, "NULL object\n");
+ }
+
+ fflush(stdout);
+ fflush(stderr);
+
+ /* Terminate the process: */
+ abort();
+}
+
#ifndef Py_TRACE_REFS
/* For Py_LIMITED_API, we need an out-of-line version of _Py_Dealloc.
Define this here, so we can undefine the macro. */

@ -0,0 +1,30 @@
diff -r 39b9b05c3085 Lib/distutils/sysconfig.py
--- a/Lib/distutils/sysconfig.py Wed Apr 10 00:27:23 2013 +0200
+++ b/Lib/distutils/sysconfig.py Wed Apr 10 10:14:18 2013 +0200
@@ -362,7 +362,10 @@
done[n] = item = ""
if found:
after = value[m.end():]
- value = value[:m.start()] + item + after
+ value = value[:m.start()]
+ if item.strip() not in value:
+ value += item
+ value += after
if "$" in after:
notdone[name] = value
else:
diff -r 39b9b05c3085 Lib/sysconfig.py
--- a/Lib/sysconfig.py Wed Apr 10 00:27:23 2013 +0200
+++ b/Lib/sysconfig.py Wed Apr 10 10:14:18 2013 +0200
@@ -296,7 +296,10 @@
if found:
after = value[m.end():]
- value = value[:m.start()] + item + after
+ value = value[:m.start()]
+ if item.strip() not in value:
+ value += item
+ value += after
if "$" in after:
notdone[name] = value
else:

@ -0,0 +1,245 @@
diff --git a/Lib/ensurepip/__init__.py b/Lib/ensurepip/__init__.py
index 4748ba4..986d5e9 100644
--- a/Lib/ensurepip/__init__.py
+++ b/Lib/ensurepip/__init__.py
@@ -1,8 +1,10 @@
import os
import os.path
import pkgutil
+import shutil
import sys
import tempfile
+from ensurepip import rewheel
__all__ = ["version", "bootstrap"]
@@ -24,8 +26,15 @@ def _run_pip(args, additional_paths=None):
sys.path = additional_paths + sys.path
# Install the bundled software
- import pip._internal
- return pip._internal.main(args)
+ try:
+ # pip 10
+ from pip._internal import main
+ except ImportError:
+ # pip 9
+ from pip import main
+ if args[0] in ["install", "list", "wheel"]:
+ args.append('--pre')
+ return main(args)
def version():
@@ -88,20 +97,39 @@ def _bootstrap(*, root=None, upgrade=False, user=False,
# omit pip and easy_install
os.environ["ENSUREPIP_OPTIONS"] = "install"
+ whls = []
+ rewheel_dir = None
+ # try to see if we have system-wide versions of _PROJECTS
+ dep_records = rewheel.find_system_records([p[0] for p in _PROJECTS])
+ # TODO: check if system-wide versions are the newest ones
+ # if --upgrade is used?
+ if all(dep_records):
+ # if we have all _PROJECTS installed system-wide, we'll recreate
+ # wheels from them and install those
+ rewheel_dir = tempfile.TemporaryDirectory()
+ for dr in dep_records:
+ new_whl = rewheel.rewheel_from_record(dr, rewheel_dir.name)
+ whls.append(os.path.join(rewheel_dir.name, new_whl))
+ else:
+ # if we don't have all the _PROJECTS installed system-wide,
+ # let's just fall back to bundled wheels
+ for project, version in _PROJECTS:
+ whl = os.path.join(
+ os.path.dirname(__file__),
+ "_bundled",
+ "{}-{}-py2.py3-none-any.whl".format(project, version)
+ )
+ whls.append(whl)
+
with tempfile.TemporaryDirectory() as tmpdir:
# Put our bundled wheels into a temporary directory and construct the
# additional paths that need added to sys.path
additional_paths = []
- for project, version in _PROJECTS:
- wheel_name = "{}-{}-py2.py3-none-any.whl".format(project, version)
- whl = pkgutil.get_data(
- "ensurepip",
- "_bundled/{}".format(wheel_name),
- )
- with open(os.path.join(tmpdir, wheel_name), "wb") as fp:
- fp.write(whl)
-
- additional_paths.append(os.path.join(tmpdir, wheel_name))
+ for whl in whls:
+ shutil.copy(whl, tmpdir)
+ additional_paths.append(os.path.join(tmpdir, os.path.basename(whl)))
+ if rewheel_dir:
+ rewheel_dir.cleanup()
# Construct the arguments to be passed to the pip command
args = ["install", "--no-index", "--find-links", tmpdir]
diff --git a/Lib/ensurepip/rewheel/__init__.py b/Lib/ensurepip/rewheel/__init__.py
new file mode 100644
index 0000000..753c764
--- /dev/null
+++ b/Lib/ensurepip/rewheel/__init__.py
@@ -0,0 +1,143 @@
+import argparse
+import codecs
+import csv
+import email.parser
+import os
+import io
+import re
+import site
+import subprocess
+import sys
+import zipfile
+
+def run():
+ parser = argparse.ArgumentParser(description='Recreate wheel of package with given RECORD.')
+ parser.add_argument('record_path',
+ help='Path to RECORD file')
+ parser.add_argument('-o', '--output-dir',
+ help='Dir where to place the wheel, defaults to current working dir.',
+ dest='outdir',
+ default=os.path.curdir)
+
+ ns = parser.parse_args()
+ retcode = 0
+ try:
+ print(rewheel_from_record(**vars(ns)))
+ except BaseException as e:
+ print('Failed: {}'.format(e))
+ retcode = 1
+ sys.exit(1)
+
+def find_system_records(projects):
+ """Return list of paths to RECORD files for system-installed projects.
+
+ If a project is not installed, the resulting list contains None instead
+ of a path to its RECORD
+ """
+ records = []
+ # get system site-packages dirs
+ sys_sitepack = site.getsitepackages([sys.base_prefix, sys.base_exec_prefix])
+ sys_sitepack = [sp for sp in sys_sitepack if os.path.exists(sp)]
+ # try to find all projects in all system site-packages
+ for project in projects:
+ path = None
+ for sp in sys_sitepack:
+ dist_info_re = os.path.join(sp, project) + r'-[^\{0}]+\.dist-info'.format(os.sep)
+ candidates = [os.path.join(sp, p) for p in os.listdir(sp)]
+ # filter out candidate dirs based on the above regexp
+ filtered = [c for c in candidates if re.match(dist_info_re, c)]
+ # if we have 0 or 2 or more dirs, something is wrong...
+ if len(filtered) == 1:
+ path = filtered[0]
+ if path is not None:
+ records.append(os.path.join(path, 'RECORD'))
+ else:
+ records.append(None)
+ return records
+
+def rewheel_from_record(record_path, outdir):
+ """Recreates a whee of package with given record_path and returns path
+ to the newly created wheel."""
+ site_dir = os.path.dirname(os.path.dirname(record_path))
+ record_relpath = record_path[len(site_dir):].strip(os.path.sep)
+ to_write, to_omit = get_records_to_pack(site_dir, record_relpath)
+ new_wheel_name = get_wheel_name(record_path)
+ new_wheel_path = os.path.join(outdir, new_wheel_name + '.whl')
+
+ new_wheel = zipfile.ZipFile(new_wheel_path, mode='w', compression=zipfile.ZIP_DEFLATED)
+ # we need to write a new record with just the files that we will write,
+ # e.g. not binaries and *.pyc/*.pyo files
+ new_record = io.StringIO()
+ writer = csv.writer(new_record)
+
+ # handle files that we can write straight away
+ for f, sha_hash, size in to_write:
+ new_wheel.write(os.path.join(site_dir, f), arcname=f)
+ writer.writerow([f, sha_hash,size])
+
+ # rewrite the old wheel file with a new computed one
+ writer.writerow([record_relpath, '', ''])
+ new_wheel.writestr(record_relpath, new_record.getvalue())
+
+ new_wheel.close()
+
+ return new_wheel.filename
+
+def get_wheel_name(record_path):
+ """Return proper name of the wheel, without .whl."""
+
+ wheel_info_path = os.path.join(os.path.dirname(record_path), 'WHEEL')
+ with codecs.open(wheel_info_path, encoding='utf-8') as wheel_info_file:
+ wheel_info = email.parser.Parser().parsestr(wheel_info_file.read())
+
+ metadata_path = os.path.join(os.path.dirname(record_path), 'METADATA')
+ with codecs.open(metadata_path, encoding='utf-8') as metadata_file:
+ metadata = email.parser.Parser().parsestr(metadata_file.read())
+
+ # construct name parts according to wheel spec
+ distribution = metadata.get('Name')
+ version = metadata.get('Version')
+ build_tag = '' # nothing for now
+ lang_tag = []
+ for t in wheel_info.get_all('Tag'):
+ lang_tag.append(t.split('-')[0])
+ lang_tag = '.'.join(lang_tag)
+ abi_tag, plat_tag = wheel_info.get('Tag').split('-')[1:3]
+ # leave out build tag, if it is empty
+ to_join = filter(None, [distribution, version, build_tag, lang_tag, abi_tag, plat_tag])
+ return '-'.join(list(to_join))
+
+def get_records_to_pack(site_dir, record_relpath):
+ """Accepts path of sitedir and path of RECORD file relative to it.
+ Returns two lists:
+ - list of files that can be written to new RECORD straight away
+ - list of files that shouldn't be written or need some processing
+ (pyc and pyo files, scripts)
+ """
+ record_file_path = os.path.join(site_dir, record_relpath)
+ with codecs.open(record_file_path, encoding='utf-8') as record_file:
+ record_contents = record_file.read()
+ # temporary fix for https://github.com/pypa/pip/issues/1376
+ # we need to ignore files under ".data" directory
+ data_dir = os.path.dirname(record_relpath).strip(os.path.sep)
+ data_dir = data_dir[:-len('dist-info')] + 'data'
+
+ to_write = []
+ to_omit = []
+ for l in record_contents.splitlines():
+ spl = l.split(',')
+ if len(spl) == 3:
+ # new record will omit (or write differently):
+ # - abs paths, paths with ".." (entry points),
+ # - pyc+pyo files
+ # - the old RECORD file
+ # TODO: is there any better way to recognize an entry point?
+ if os.path.isabs(spl[0]) or spl[0].startswith('..') or \
+ spl[0].endswith('.pyc') or spl[0].endswith('.pyo') or \
+ spl[0] == record_relpath or spl[0].startswith(data_dir):
+ to_omit.append(spl)
+ else:
+ to_write.append(spl)
+ else:
+ pass # bad RECORD or empty line
+ return to_write, to_omit
diff --git a/Makefile.pre.in b/Makefile.pre.in
index 85e2ee3..4d34130 100644
--- a/Makefile.pre.in
+++ b/Makefile.pre.in
@@ -1256,7 +1256,7 @@ LIBSUBDIRS= tkinter tkinter/test tkinter/test/test_tkinter \
test/test_asyncio \
collections concurrent concurrent/futures encodings \
email email/mime test/test_email test/test_email/data \
- ensurepip ensurepip/_bundled \
+ ensurepip ensurepip/_bundled ensurepip/rewheel \
html json test/test_json http dbm xmlrpc \
sqlite3 sqlite3/test \
logging csv wsgiref urllib \

@ -0,0 +1,12 @@
diff -up Python-3.5.0/Makefile.pre.in.lib Python-3.5.0/Makefile.pre.in
--- Python-3.5.0/Makefile.pre.in.lib 2015-09-21 15:39:47.928286620 +0200
+++ Python-3.5.0/Makefile.pre.in 2015-09-21 15:42:58.004042762 +0200
@@ -1340,7 +1340,7 @@ inclinstall:
# Install the library and miscellaneous stuff needed for extending/embedding
# This goes into $(exec_prefix)
-LIBPL= @LIBPL@
+LIBPL= $(LIBDEST)/config-$(LDVERSION)-$(MULTIARCH)
# pkgconfig directory
LIBPC= $(LIBDIR)/pkgconfig

@ -0,0 +1,46 @@
diff --git a/Lib/distutils/command/install.py b/Lib/distutils/command/install.py
index 0258d3d..4ebf50a 100644
--- a/Lib/distutils/command/install.py
+++ b/Lib/distutils/command/install.py
@@ -418,8 +418,19 @@ class install(Command):
raise DistutilsOptionError(
"must not supply exec-prefix without prefix")
- self.prefix = os.path.normpath(sys.prefix)
- self.exec_prefix = os.path.normpath(sys.exec_prefix)
+ # self.prefix is set to sys.prefix + /local/
+ # if neither RPM build nor virtual environment is
+ # detected to make pip and distutils install packages
+ # into the separate location.
+ if (not (hasattr(sys, 'real_prefix') or
+ sys.prefix != sys.base_prefix) and
+ 'RPM_BUILD_ROOT' not in os.environ):
+ addition = "/local"
+ else:
+ addition = ""
+
+ self.prefix = os.path.normpath(sys.prefix) + addition
+ self.exec_prefix = os.path.normpath(sys.exec_prefix) + addition
else:
if self.exec_prefix is None:
diff --git a/Lib/site.py b/Lib/site.py
index 0fc9200..c95202e 100644
--- a/Lib/site.py
+++ b/Lib/site.py
@@ -322,7 +322,14 @@ def getsitepackages(prefixes=None):
return sitepackages
def addsitepackages(known_paths, prefixes=None):
- """Add site-packages to sys.path"""
+ """Add site-packages to sys.path
+
+ '/usr/local' is included in PREFIXES if RPM build is not detected
+ to make packages installed into this location visible.
+
+ """
+ if ENABLE_USER_SITE and 'RPM_BUILD_ROOT' not in os.environ:
+ PREFIXES.insert(0, "/usr/local")
for sitedir in getsitepackages(prefixes):
if os.path.isdir(sitedir):
addsitedir(sitedir, known_paths)

@ -0,0 +1,901 @@
diff --git a/Doc/using/cmdline.rst b/Doc/using/cmdline.rst
index d14793a..65aa3ad 100644
--- a/Doc/using/cmdline.rst
+++ b/Doc/using/cmdline.rst
@@ -728,6 +728,45 @@ conflict.
.. versionadded:: 3.6
+
+.. envvar:: PYTHONCOERCECLOCALE
+
+ If set to the value ``0``, causes the main Python command line application
+ to skip coercing the legacy ASCII-based C locale to a more capable UTF-8
+ based alternative. Note that this setting is checked even when the
+ :option:`-E` or :option:`-I` options are used, as it is handled prior to
+ the processing of command line options.
+
+ If this variable is *not* set, or is set to a value other than ``0``, and
+ the current locale reported for the ``LC_CTYPE`` category is the default
+ ``C`` locale, then the Python CLI will attempt to configure one of the
+ following locales for the given locale categories before loading the
+ interpreter runtime:
+
+ * ``C.UTF-8`` (``LC_ALL``)
+ * ``C.utf8`` (``LC_ALL``)
+ * ``UTF-8`` (``LC_CTYPE``)
+
+ If setting one of these locale categories succeeds, then the matching
+ environment variables will be set (both ``LC_ALL`` and ``LANG`` for the
+ ``LC_ALL`` category, and ``LC_CTYPE`` for the ``LC_CTYPE`` category) in
+ the current process environment before the Python runtime is initialized.
+
+ Configuring one of these locales (either explicitly or via the above
+ implicit locale coercion) will automatically set the error handler for
+ :data:`sys.stdin` and :data:`sys.stdout` to ``surrogateescape``. This
+ behavior can be overridden using :envvar:`PYTHONIOENCODING` as usual.
+
+ For debugging purposes, setting ``PYTHONCOERCECLOCALE=warn`` will cause
+ Python to emit warning messages on ``stderr`` if either the locale coercion
+ activates, or else if a locale that *would* have triggered coercion is
+ still active when the Python runtime is initialized.
+
+ Availability: \*nix
+
+ .. versionadded:: 3.7
+ See :pep:`538` for more details.
+
Debug-mode variables
~~~~~~~~~~~~~~~~~~~~
diff --git a/Lib/test/support/script_helper.py b/Lib/test/support/script_helper.py
index 507dc48..c3cb720 100644
--- a/Lib/test/support/script_helper.py
+++ b/Lib/test/support/script_helper.py
@@ -56,8 +56,35 @@ def interpreter_requires_environment():
return __cached_interp_requires_environment
-_PythonRunResult = collections.namedtuple("_PythonRunResult",
- ("rc", "out", "err"))
+class _PythonRunResult(collections.namedtuple("_PythonRunResult",
+ ("rc", "out", "err"))):
+ """Helper for reporting Python subprocess run results"""
+ def fail(self, cmd_line):
+ """Provide helpful details about failed subcommand runs"""
+ # Limit to 80 lines to ASCII characters
+ maxlen = 80 * 100
+ out, err = self.out, self.err
+ if len(out) > maxlen:
+ out = b'(... truncated stdout ...)' + out[-maxlen:]
+ if len(err) > maxlen:
+ err = b'(... truncated stderr ...)' + err[-maxlen:]
+ out = out.decode('ascii', 'replace').rstrip()
+ err = err.decode('ascii', 'replace').rstrip()
+ raise AssertionError("Process return code is %d\n"
+ "command line: %r\n"
+ "\n"
+ "stdout:\n"
+ "---\n"
+ "%s\n"
+ "---\n"
+ "\n"
+ "stderr:\n"
+ "---\n"
+ "%s\n"
+ "---"
+ % (self.rc, cmd_line,
+ out,
+ err))
# Executing the interpreter in a subprocess
@@ -115,30 +142,7 @@ def run_python_until_end(*args, **env_vars):
def _assert_python(expected_success, *args, **env_vars):
res, cmd_line = run_python_until_end(*args, **env_vars)
if (res.rc and expected_success) or (not res.rc and not expected_success):
- # Limit to 80 lines to ASCII characters
- maxlen = 80 * 100
- out, err = res.out, res.err
- if len(out) > maxlen:
- out = b'(... truncated stdout ...)' + out[-maxlen:]
- if len(err) > maxlen:
- err = b'(... truncated stderr ...)' + err[-maxlen:]
- out = out.decode('ascii', 'replace').rstrip()
- err = err.decode('ascii', 'replace').rstrip()
- raise AssertionError("Process return code is %d\n"
- "command line: %r\n"
- "\n"
- "stdout:\n"
- "---\n"
- "%s\n"
- "---\n"
- "\n"
- "stderr:\n"
- "---\n"
- "%s\n"
- "---"
- % (res.rc, cmd_line,
- out,
- err))
+ res.fail(cmd_line)
return res
def assert_python_ok(*args, **env_vars):
diff --git a/Lib/test/test_c_locale_coercion.py b/Lib/test/test_c_locale_coercion.py
new file mode 100644
index 0000000..635c98f
--- /dev/null
+++ b/Lib/test/test_c_locale_coercion.py
@@ -0,0 +1,371 @@
+# Tests the attempted automatic coercion of the C locale to a UTF-8 locale
+
+import unittest
+import locale
+import os
+import sys
+import sysconfig
+import shutil
+import subprocess
+from collections import namedtuple
+
+import test.support
+from test.support.script_helper import (
+ run_python_until_end,
+ interpreter_requires_environment,
+)
+
+# Set our expectation for the default encoding used in the C locale
+# for the filesystem encoding and the standard streams
+
+# AIX uses iso8859-1 in the C locale, other *nix platforms use ASCII
+if sys.platform.startswith("aix"):
+ C_LOCALE_STREAM_ENCODING = "iso8859-1"
+else:
+ C_LOCALE_STREAM_ENCODING = "ascii"
+
+# FS encoding is UTF-8 on macOS, other *nix platforms use the locale encoding
+if sys.platform == "darwin":
+ C_LOCALE_FS_ENCODING = "utf-8"
+else:
+ C_LOCALE_FS_ENCODING = C_LOCALE_STREAM_ENCODING
+
+# Note that the above is probably still wrong in some cases, such as:
+# * Windows when PYTHONLEGACYWINDOWSFSENCODING is set
+# * AIX and any other platforms that use latin-1 in the C locale
+#
+# Options for dealing with this:
+# * Don't set PYTHON_COERCE_C_LOCALE on such platforms (e.g. Windows doesn't)
+# * Fix the test expectations to match the actual platform behaviour
+
+# In order to get the warning messages to match up as expected, the candidate
+# order here must much the target locale order in Python/pylifecycle.c
+_C_UTF8_LOCALES = ("C.UTF-8", "C.utf8", "UTF-8")
+
+# There's no reliable cross-platform way of checking locale alias
+# lists, so the only way of knowing which of these locales will work
+# is to try them with locale.setlocale(). We do that in a subprocess
+# to avoid altering the locale of the test runner.
+#
+# If the relevant locale module attributes exist, and we're not on a platform
+# where we expect it to always succeed, we also check that
+# `locale.nl_langinfo(locale.CODESET)` works, as if it fails, the interpreter
+# will skip locale coercion for that particular target locale
+_check_nl_langinfo_CODESET = bool(
+ sys.platform not in ("darwin", "linux") and
+ hasattr(locale, "nl_langinfo") and
+ hasattr(locale, "CODESET")
+)
+
+def _set_locale_in_subprocess(locale_name):
+ cmd_fmt = "import locale; print(locale.setlocale(locale.LC_CTYPE, '{}'))"
+ if _check_nl_langinfo_CODESET:
+ # If there's no valid CODESET, we expect coercion to be skipped
+ cmd_fmt += "; import sys; sys.exit(not locale.nl_langinfo(locale.CODESET))"
+ cmd = cmd_fmt.format(locale_name)
+ result, py_cmd = run_python_until_end("-c", cmd, __isolated=True)
+ return result.rc == 0
+
+
+
+_fields = "fsencoding stdin_info stdout_info stderr_info lang lc_ctype lc_all"
+_EncodingDetails = namedtuple("EncodingDetails", _fields)
+
+class EncodingDetails(_EncodingDetails):
+ # XXX (ncoghlan): Using JSON for child state reporting may be less fragile
+ CHILD_PROCESS_SCRIPT = ";".join([
+ "import sys, os",
+ "print(sys.getfilesystemencoding())",
+ "print(sys.stdin.encoding + ':' + sys.stdin.errors)",
+ "print(sys.stdout.encoding + ':' + sys.stdout.errors)",
+ "print(sys.stderr.encoding + ':' + sys.stderr.errors)",
+ "print(os.environ.get('LANG', 'not set'))",
+ "print(os.environ.get('LC_CTYPE', 'not set'))",
+ "print(os.environ.get('LC_ALL', 'not set'))",
+ ])
+
+ @classmethod
+ def get_expected_details(cls, coercion_expected, fs_encoding, stream_encoding, env_vars):
+ """Returns expected child process details for a given encoding"""
+ _stream = stream_encoding + ":{}"
+ # stdin and stdout should use surrogateescape either because the
+ # coercion triggered, or because the C locale was detected
+ stream_info = 2*[_stream.format("surrogateescape")]
+ # stderr should always use backslashreplace
+ stream_info.append(_stream.format("backslashreplace"))
+ expected_lang = env_vars.get("LANG", "not set").lower()
+ if coercion_expected:
+ expected_lc_ctype = CLI_COERCION_TARGET.lower()
+ else:
+ expected_lc_ctype = env_vars.get("LC_CTYPE", "not set").lower()
+ expected_lc_all = env_vars.get("LC_ALL", "not set").lower()
+ env_info = expected_lang, expected_lc_ctype, expected_lc_all
+ return dict(cls(fs_encoding, *stream_info, *env_info)._asdict())
+
+ @staticmethod
+ def _handle_output_variations(data):
+ """Adjust the output to handle platform specific idiosyncrasies
+
+ * Some platforms report ASCII as ANSI_X3.4-1968
+ * Some platforms report ASCII as US-ASCII
+ * Some platforms report UTF-8 instead of utf-8
+ """
+ data = data.replace(b"ANSI_X3.4-1968", b"ascii")
+ data = data.replace(b"US-ASCII", b"ascii")
+ data = data.lower()
+ return data
+
+ @classmethod
+ def get_child_details(cls, env_vars):
+ """Retrieves fsencoding and standard stream details from a child process
+
+ Returns (encoding_details, stderr_lines):
+
+ - encoding_details: EncodingDetails for eager decoding
+ - stderr_lines: result of calling splitlines() on the stderr output
+
+ The child is run in isolated mode if the current interpreter supports
+ that.
+ """
+ result, py_cmd = run_python_until_end(
+ "-c", cls.CHILD_PROCESS_SCRIPT,
+ __isolated=True,
+ **env_vars
+ )
+ if not result.rc == 0:
+ result.fail(py_cmd)
+ # All subprocess outputs in this test case should be pure ASCII
+ adjusted_output = cls._handle_output_variations(result.out)
+ stdout_lines = adjusted_output.decode("ascii").splitlines()
+ child_encoding_details = dict(cls(*stdout_lines)._asdict())
+ stderr_lines = result.err.decode("ascii").rstrip().splitlines()
+ return child_encoding_details, stderr_lines
+
+
+# Details of the shared library warning emitted at runtime
+LEGACY_LOCALE_WARNING = (
+ "Python runtime initialized with LC_CTYPE=C (a locale with default ASCII "
+ "encoding), which may cause Unicode compatibility problems. Using C.UTF-8, "
+ "C.utf8, or UTF-8 (if available) as alternative Unicode-compatible "
+ "locales is recommended."
+)
+
+# Details of the CLI locale coercion warning emitted at runtime
+CLI_COERCION_WARNING_FMT = (
+ "Python detected LC_CTYPE=C: LC_CTYPE coerced to {} (set another locale "
+ "or PYTHONCOERCECLOCALE=0 to disable this locale coercion behavior)."
+)
+
+
+AVAILABLE_TARGETS = None
+CLI_COERCION_TARGET = None
+CLI_COERCION_WARNING = None
+
+def setUpModule():
+ global AVAILABLE_TARGETS
+ global CLI_COERCION_TARGET
+ global CLI_COERCION_WARNING
+
+ if AVAILABLE_TARGETS is not None:
+ # initialization already done
+ return
+ AVAILABLE_TARGETS = []
+
+ # Find the target locales available in the current system
+ for target_locale in _C_UTF8_LOCALES:
+ if _set_locale_in_subprocess(target_locale):
+ AVAILABLE_TARGETS.append(target_locale)
+
+ if AVAILABLE_TARGETS:
+ # Coercion is expected to use the first available target locale
+ CLI_COERCION_TARGET = AVAILABLE_TARGETS[0]
+ CLI_COERCION_WARNING = CLI_COERCION_WARNING_FMT.format(CLI_COERCION_TARGET)
+
+
+class _LocaleHandlingTestCase(unittest.TestCase):
+ # Base class to check expected locale handling behaviour
+
+ def _check_child_encoding_details(self,
+ env_vars,
+ expected_fs_encoding,
+ expected_stream_encoding,
+ expected_warnings,
+ coercion_expected):
+ """Check the C locale handling for the given process environment
+
+ Parameters:
+ expected_fs_encoding: expected sys.getfilesystemencoding() result
+ expected_stream_encoding: expected encoding for standard streams
+ expected_warning: stderr output to expect (if any)
+ """
+ result = EncodingDetails.get_child_details(env_vars)
+ encoding_details, stderr_lines = result
+ expected_details = EncodingDetails.get_expected_details(
+ coercion_expected,
+ expected_fs_encoding,
+ expected_stream_encoding,
+ env_vars
+ )
+ self.assertEqual(encoding_details, expected_details)
+ if expected_warnings is None:
+ expected_warnings = []
+ self.assertEqual(stderr_lines, expected_warnings)
+
+
+class LocaleConfigurationTests(_LocaleHandlingTestCase):
+ # Test explicit external configuration via the process environment
+
+ def setUpClass():
+ # This relies on setupModule() having been run, so it can't be
+ # handled via the @unittest.skipUnless decorator
+ if not AVAILABLE_TARGETS:
+ raise unittest.SkipTest("No C-with-UTF-8 locale available")
+
+ def test_external_target_locale_configuration(self):
+
+ # Explicitly setting a target locale should give the same behaviour as
+ # is seen when implicitly coercing to that target locale
+ self.maxDiff = None
+
+ expected_fs_encoding = "utf-8"
+ expected_stream_encoding = "utf-8"
+
+ base_var_dict = {
+ "LANG": "",
+ "LC_CTYPE": "",
+ "LC_ALL": "",
+ }
+ for env_var in ("LANG", "LC_CTYPE"):
+ for locale_to_set in AVAILABLE_TARGETS:
+ # XXX (ncoghlan): LANG=UTF-8 doesn't appear to work as
+ # expected, so skip that combination for now
+ # See https://bugs.python.org/issue30672 for discussion
+ if env_var == "LANG" and locale_to_set == "UTF-8":
+ continue
+
+ with self.subTest(env_var=env_var,
+ configured_locale=locale_to_set):
+ var_dict = base_var_dict.copy()
+ var_dict[env_var] = locale_to_set
+ self._check_child_encoding_details(var_dict,
+ expected_fs_encoding,
+ expected_stream_encoding,
+ expected_warnings=None,
+ coercion_expected=False)
+
+
+
+@test.support.cpython_only
+@unittest.skipUnless(sysconfig.get_config_var("PY_COERCE_C_LOCALE"),
+ "C locale coercion disabled at build time")
+class LocaleCoercionTests(_LocaleHandlingTestCase):
+ # Test implicit reconfiguration of the environment during CLI startup
+
+ def _check_c_locale_coercion(self,
+ fs_encoding, stream_encoding,
+ coerce_c_locale,
+ expected_warnings=None,
+ coercion_expected=True,
+ **extra_vars):
+ """Check the C locale handling for various configurations
+
+ Parameters:
+ fs_encoding: expected sys.getfilesystemencoding() result
+ stream_encoding: expected encoding for standard streams
+ coerce_c_locale: setting to use for PYTHONCOERCECLOCALE
+ None: don't set the variable at all
+ str: the value set in the child's environment
+ expected_warnings: expected warning lines on stderr
+ extra_vars: additional environment variables to set in subprocess
+ """
+ self.maxDiff = None
+
+ if not AVAILABLE_TARGETS:
+ # Locale coercion is disabled when there aren't any target locales
+ fs_encoding = C_LOCALE_FS_ENCODING
+ stream_encoding = C_LOCALE_STREAM_ENCODING
+ coercion_expected = False
+ if expected_warnings:
+ expected_warnings = [LEGACY_LOCALE_WARNING]
+
+ base_var_dict = {
+ "LANG": "",
+ "LC_CTYPE": "",
+ "LC_ALL": "",
+ }
+ base_var_dict.update(extra_vars)
+ for env_var in ("LANG", "LC_CTYPE"):
+ for locale_to_set in ("", "C", "POSIX", "invalid.ascii"):
+ # XXX (ncoghlan): *BSD platforms don't behave as expected in the
+ # POSIX locale, so we skip that for now
+ # See https://bugs.python.org/issue30672 for discussion
+ if locale_to_set == "POSIX":
+ continue
+ with self.subTest(env_var=env_var,
+ nominal_locale=locale_to_set,
+ PYTHONCOERCECLOCALE=coerce_c_locale):
+ var_dict = base_var_dict.copy()
+ var_dict[env_var] = locale_to_set
+ if coerce_c_locale is not None:
+ var_dict["PYTHONCOERCECLOCALE"] = coerce_c_locale
+ # Check behaviour on successful coercion
+ self._check_child_encoding_details(var_dict,
+ fs_encoding,
+ stream_encoding,
+ expected_warnings,
+ coercion_expected)
+
+ def test_test_PYTHONCOERCECLOCALE_not_set(self):
+ # This should coerce to the first available target locale by default
+ self._check_c_locale_coercion("utf-8", "utf-8", coerce_c_locale=None)
+
+ def test_PYTHONCOERCECLOCALE_not_zero(self):
+ # *Any* string other than "0" is considered "set" for our purposes
+ # and hence should result in the locale coercion being enabled
+ for setting in ("", "1", "true", "false"):
+ self._check_c_locale_coercion("utf-8", "utf-8", coerce_c_locale=setting)
+
+ def test_PYTHONCOERCECLOCALE_set_to_warn(self):
+ # PYTHONCOERCECLOCALE=warn enables runtime warnings for legacy locales
+ self._check_c_locale_coercion("utf-8", "utf-8",
+ coerce_c_locale="warn",
+ expected_warnings=[CLI_COERCION_WARNING])
+
+
+ def test_PYTHONCOERCECLOCALE_set_to_zero(self):
+ # The setting "0" should result in the locale coercion being disabled
+ self._check_c_locale_coercion(C_LOCALE_FS_ENCODING,
+ C_LOCALE_STREAM_ENCODING,
+ coerce_c_locale="0",
+ coercion_expected=False)
+ # Setting LC_ALL=C shouldn't make any difference to the behaviour
+ self._check_c_locale_coercion(C_LOCALE_FS_ENCODING,
+ C_LOCALE_STREAM_ENCODING,
+ coerce_c_locale="0",
+ LC_ALL="C",
+ coercion_expected=False)
+
+ def test_LC_ALL_set_to_C(self):
+ # Setting LC_ALL should render the locale coercion ineffective
+ self._check_c_locale_coercion(C_LOCALE_FS_ENCODING,
+ C_LOCALE_STREAM_ENCODING,
+ coerce_c_locale=None,
+ LC_ALL="C",
+ coercion_expected=False)
+ # And result in a warning about a lack of locale compatibility
+ self._check_c_locale_coercion(C_LOCALE_FS_ENCODING,
+ C_LOCALE_STREAM_ENCODING,
+ coerce_c_locale="warn",
+ LC_ALL="C",
+ expected_warnings=[LEGACY_LOCALE_WARNING],
+ coercion_expected=False)
+
+def test_main():
+ test.support.run_unittest(
+ LocaleConfigurationTests,
+ LocaleCoercionTests
+ )
+ test.support.reap_children()
+
+if __name__ == "__main__":
+ test_main()
diff --git a/Lib/test/test_cmd_line.py b/Lib/test/test_cmd_line.py
index 38156b4..5922ed9 100644
--- a/Lib/test/test_cmd_line.py
+++ b/Lib/test/test_cmd_line.py
@@ -153,6 +153,7 @@ class CmdLineTest(unittest.TestCase):
env = os.environ.copy()
# Use C locale to get ascii for the locale encoding
env['LC_ALL'] = 'C'
+ env['PYTHONCOERCECLOCALE'] = '0'
code = (
b'import locale; '
b'print(ascii("' + undecodable + b'"), '
diff --git a/Lib/test/test_sys.py b/Lib/test/test_sys.py
index 7866a5c..b41239a 100644
--- a/Lib/test/test_sys.py
+++ b/Lib/test/test_sys.py
@@ -680,6 +680,7 @@ class SysModuleTest(unittest.TestCase):
# Force the POSIX locale
env = os.environ.copy()
env["LC_ALL"] = "C"
+ env["PYTHONCOERCECLOCALE"] = "0"
code = '\n'.join((
'import sys',
'def dump(name):',
diff --git a/Modules/main.c b/Modules/main.c
index 585d696..96d8be4 100644
--- a/Modules/main.c
+++ b/Modules/main.c
@@ -107,7 +107,11 @@ static const char usage_6[] =
" predictable seed.\n"
"PYTHONMALLOC: set the Python memory allocators and/or install debug hooks\n"
" on Python memory allocators. Use PYTHONMALLOC=debug to install debug\n"
-" hooks.\n";
+" hooks.\n"
+
+"PYTHONCOERCECLOCALE: if this variable is set to 0, it disables the locale\n"
+" coercion behavior. Use PYTHONCOERCECLOCALE=warn to request display of\n"
+" locale coercion and locale compatibility warnings on stderr.\n";
static int
usage(int exitcode, const wchar_t* program)
diff --git a/Programs/_testembed.c b/Programs/_testembed.c
index 813cf30..2a64092 100644
--- a/Programs/_testembed.c
+++ b/Programs/_testembed.c
@@ -1,4 +1,5 @@
#include <Python.h>
+#include "pyconfig.h"
#include "pythread.h"
#include <stdio.h>
diff --git a/Programs/python.c b/Programs/python.c
index a7afbc7..03f8295 100644
--- a/Programs/python.c
+++ b/Programs/python.c
@@ -15,6 +15,21 @@ wmain(int argc, wchar_t **argv)
}
#else
+/* Access private pylifecycle helper API to better handle the legacy C locale
+ *
+ * The legacy C locale assumes ASCII as the default text encoding, which
+ * causes problems not only for the CPython runtime, but also other
+ * components like GNU readline.
+ *
+ * Accordingly, when the CLI detects it, it attempts to coerce it to a
+ * more capable UTF-8 based alternative.
+ *
+ * See the documentation of the PYTHONCOERCECLOCALE setting for more details.
+ *
+ */
+extern int _Py_LegacyLocaleDetected(void);
+extern void _Py_CoerceLegacyLocale(void);
+
int
main(int argc, char **argv)
{
@@ -25,7 +40,11 @@ main(int argc, char **argv)
char *oldloc;
/* Force malloc() allocator to bootstrap Python */
+#ifdef Py_DEBUG
+ (void)_PyMem_SetupAllocators("malloc_debug");
+# else
(void)_PyMem_SetupAllocators("malloc");
+# endif
argv_copy = (wchar_t **)PyMem_RawMalloc(sizeof(wchar_t*) * (argc+1));
argv_copy2 = (wchar_t **)PyMem_RawMalloc(sizeof(wchar_t*) * (argc+1));
@@ -49,7 +68,21 @@ main(int argc, char **argv)
return 1;
}
+#ifdef __ANDROID__
+ /* Passing "" to setlocale() on Android requests the C locale rather
+ * than checking environment variables, so request C.UTF-8 explicitly
+ */
+ setlocale(LC_ALL, "C.UTF-8");
+#else
+ /* Reconfigure the locale to the default for this process */
setlocale(LC_ALL, "");
+#endif
+
+ if (_Py_LegacyLocaleDetected()) {
+ _Py_CoerceLegacyLocale();
+ }
+
+ /* Convert from char to wchar_t based on the locale settings */
for (i = 0; i < argc; i++) {
argv_copy[i] = Py_DecodeLocale(argv[i], NULL);
if (!argv_copy[i]) {
@@ -70,7 +103,11 @@ main(int argc, char **argv)
/* Force again malloc() allocator to release memory blocks allocated
before Py_Main() */
+#ifdef Py_DEBUG
+ (void)_PyMem_SetupAllocators("malloc_debug");
+# else
(void)_PyMem_SetupAllocators("malloc");
+# endif
for (i = 0; i < argc; i++) {
PyMem_RawFree(argv_copy2[i]);
diff --git a/Python/pylifecycle.c b/Python/pylifecycle.c
index ecfdfee..4fee178 100644
--- a/Python/pylifecycle.c
+++ b/Python/pylifecycle.c
@@ -167,6 +167,7 @@ Py_SetStandardStreamEncoding(const char *encoding, const char *errors)
return 0;
}
+
/* Global initializations. Can be undone by Py_FinalizeEx(). Don't
call this twice without an intervening Py_FinalizeEx() call. When
initializations fail, a fatal error is issued and the function does
@@ -301,6 +302,183 @@ import_init(PyInterpreterState *interp, PyObject *sysmod)
}
+/* Helper functions to better handle the legacy C locale
+ *
+ * The legacy C locale assumes ASCII as the default text encoding, which
+ * causes problems not only for the CPython runtime, but also other
+ * components like GNU readline.
+ *
+ * Accordingly, when the CLI detects it, it attempts to coerce it to a
+ * more capable UTF-8 based alternative as follows:
+ *
+ * if (_Py_LegacyLocaleDetected()) {
+ * _Py_CoerceLegacyLocale();
+ * }
+ *
+ * See the documentation of the PYTHONCOERCECLOCALE setting for more details.
+ *
+ * Locale coercion also impacts the default error handler for the standard
+ * streams: while the usual default is "strict", the default for the legacy
+ * C locale and for any of the coercion target locales is "surrogateescape".
+ */
+
+int
+_Py_LegacyLocaleDetected(void)
+{
+#ifndef MS_WINDOWS
+ /* On non-Windows systems, the C locale is considered a legacy locale */
+ /* XXX (ncoghlan): some platforms (notably Mac OS X) don't appear to treat
+ * the POSIX locale as a simple alias for the C locale, so
+ * we may also want to check for that explicitly.
+ */
+ const char *ctype_loc = setlocale(LC_CTYPE, NULL);
+ return ctype_loc != NULL && strcmp(ctype_loc, "C") == 0;
+#else
+ /* Windows uses code pages instead of locales, so no locale is legacy */
+ return 0;
+#endif
+}
+
+
+static const char *_C_LOCALE_WARNING =
+ "Python runtime initialized with LC_CTYPE=C (a locale with default ASCII "
+ "encoding), which may cause Unicode compatibility problems. Using C.UTF-8, "
+ "C.utf8, or UTF-8 (if available) as alternative Unicode-compatible "
+ "locales is recommended.\n";
+
+static int
+_legacy_locale_warnings_enabled(void)
+{
+ const char *coerce_c_locale = getenv("PYTHONCOERCECLOCALE");
+ return (coerce_c_locale != NULL &&
+ strncmp(coerce_c_locale, "warn", 5) == 0);
+}
+
+static void
+_emit_stderr_warning_for_legacy_locale(void)
+{
+ if (_legacy_locale_warnings_enabled()) {
+ if (_Py_LegacyLocaleDetected()) {
+ fprintf(stderr, "%s", _C_LOCALE_WARNING);
+ }
+ }
+}
+
+typedef struct _CandidateLocale {
+ const char *locale_name; /* The locale to try as a coercion target */
+} _LocaleCoercionTarget;
+
+static _LocaleCoercionTarget _TARGET_LOCALES[] = {
+ {"C.UTF-8"},
+ {"C.utf8"},
+ {"UTF-8"},
+ {NULL}
+};
+
+static char *
+get_default_standard_stream_error_handler(void)
+{
+ const char *ctype_loc = setlocale(LC_CTYPE, NULL);
+ if (ctype_loc != NULL) {
+ /* "surrogateescape" is the default in the legacy C locale */
+ if (strcmp(ctype_loc, "C") == 0) {
+ return "surrogateescape";
+ }
+
+#ifdef PY_COERCE_C_LOCALE
+ /* "surrogateescape" is the default in locale coercion target locales */
+ const _LocaleCoercionTarget *target = NULL;
+ for (target = _TARGET_LOCALES; target->locale_name; target++) {
+ if (strcmp(ctype_loc, target->locale_name) == 0) {
+ return "surrogateescape";
+ }
+ }