Mailman 3 August 2014 - Python-checkins

cpython (merge 3.4 -> default): merge 3.4 (closes #22307)
by benjamin.peterson Aug. 31, 2014

Aug. 31, 2014

http://hg.python.org/cpython/rev/164a17eca081 changeset: 92276:164a17eca081 parent: 92272:d0d4ab0ba70e parent: 92275:c30163548f64 user: Benjamin Peterson <benjamin(a)python.org> date: Sat Aug 30 21:06:00 2014 -0400 summary: merge 3.4 (closes #22307) files: Doc/library/os.rst | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-) diff --git a/Doc/library/os.rst b/Doc/library/os.rst --- a/Doc/library/os.rst +++ b/Doc/library/os.rst @@ -282,10 +282,10 @@ .. function:: getlogin() Return the name of the user logged in on the controlling terminal of the - process. For most purposes, it is more useful to use the environment variables - :envvar:`LOGNAME` or :envvar:`USERNAME` to find out who the user is, or - ``pwd.getpwuid(os.getuid())[0]`` to get the login name of the currently - effective user id. + process. For most purposes, it is more useful to use the environment + variables :envvar:`LOGNAME` or :envvar:`USERNAME` to find out who the user + is, or ``pwd.getpwuid(os.getuid())[0]`` to get the login name of the current + real user id. Availability: Unix, Windows. -- Repository URL: http://hg.python.org/cpython

1 0

cpython (3.4): getuid() returns real process id not effective (closes #22307)
by benjamin.peterson Aug. 31, 2014

Aug. 31, 2014

http://hg.python.org/cpython/rev/c30163548f64 changeset: 92275:c30163548f64 branch: 3.4 parent: 92271:a058760cb069 user: Benjamin Peterson <benjamin(a)python.org> date: Sat Aug 30 21:04:15 2014 -0400 summary: getuid() returns real process id not effective (closes #22307) files: Doc/library/os.rst | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-) diff --git a/Doc/library/os.rst b/Doc/library/os.rst --- a/Doc/library/os.rst +++ b/Doc/library/os.rst @@ -281,10 +281,10 @@ .. function:: getlogin() Return the name of the user logged in on the controlling terminal of the - process. For most purposes, it is more useful to use the environment variables - :envvar:`LOGNAME` or :envvar:`USERNAME` to find out who the user is, or - ``pwd.getpwuid(os.getuid())[0]`` to get the login name of the currently - effective user id. + process. For most purposes, it is more useful to use the environment + variables :envvar:`LOGNAME` or :envvar:`USERNAME` to find out who the user + is, or ``pwd.getpwuid(os.getuid())[0]`` to get the login name of the current + real user id. Availability: Unix, Windows. -- Repository URL: http://hg.python.org/cpython

1 0

cpython (2.7): getuid() returns real process id not effective (closes #22307)
by benjamin.peterson Aug. 31, 2014

Aug. 31, 2014

http://hg.python.org/cpython/rev/59f2edeb8443 changeset: 92274:59f2edeb8443 branch: 2.7 user: Benjamin Peterson <benjamin(a)python.org> date: Sat Aug 30 21:04:15 2014 -0400 summary: getuid() returns real process id not effective (closes #22307) files: Doc/library/os.rst | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-) diff --git a/Doc/library/os.rst b/Doc/library/os.rst --- a/Doc/library/os.rst +++ b/Doc/library/os.rst @@ -188,10 +188,10 @@ .. function:: getlogin() Return the name of the user logged in on the controlling terminal of the - process. For most purposes, it is more useful to use the environment variable - :envvar:`LOGNAME` to find out who the user is, or - ``pwd.getpwuid(os.getuid())[0]`` to get the login name of the currently - effective user id. + process. For most purposes, it is more useful to use the environment + variable :envvar:`LOGNAME` to find out who the user is, or + ``pwd.getpwuid(os.getuid())[0]`` to get the login name of the process's real + user id. Availability: Unix. -- Repository URL: http://hg.python.org/cpython

1 0

peps: updates for downloads move
by benjamin.peterson Aug. 30, 2014

Aug. 30, 2014

http://hg.python.org/peps/rev/8b98fcf2029d changeset: 5532:8b98fcf2029d user: Benjamin Peterson <benjamin(a)python.org> date: Sat Aug 30 17:25:09 2014 -0400 summary: updates for downloads move files: pep-0101.txt | 30 +++++++++++++++--------------- 1 files changed, 15 insertions(+), 15 deletions(-) diff --git a/pep-0101.txt b/pep-0101.txt --- a/pep-0101.txt +++ b/pep-0101.txt @@ -35,8 +35,8 @@ which hopefully will be on the "web of trust" with at least one of the other release managers. - * Access to ``dinsdale.python.org``, the Python web host. You'll - be uploading files directly here. + * Access to ``dl-files.psf.io``, the server that hosts download files. + You'll be uploading files directly here. * Shell access to ``hg.python.org``, the Python Mercurial host. You'll have to adapt repository configuration there. @@ -335,11 +335,11 @@ - Now he runs msi.py with ActivePython or Python with pywin32. The WE checksums the files (*.msi, *.chm, *-pdb.zip), uploads them to - dinsdale together with gpg signature files, and emails you the location - and md5sums. + dl-files.psf.io together with gpg signature files, and emails you the + location and md5sums. - ___ The ME builds Mac installer packages and uploads them to dinsdale together - with gpg signature files. + ___ The ME builds Mac installer packages and uploads them to dl-files.psf.io + together with gpg signature files. ___ Time to build the source tarball. Be sure to update your clone to the correct branch. E.g. @@ -364,7 +364,7 @@ tarballs and signatures in a subdirectory called 'X.Y.ZaN/src', and the built docs in 'X.Y.ZaN/docs' (for final releases). - ___ scp or rsync all the files to your home directory on dinsdale.python.org. + ___ scp or rsync all the files to your home directory on dl-files.psf.io. While you're waiting for the files to finish uploading, you can continue on with the remaining tasks. You can also ask folks on #python-dev @@ -401,32 +401,32 @@ freshly unpacked directory looks weird, you better stop now and figure out what the problem is. - ___ Now you need to go to dinsdale.python.org and move all the files in + ___ Now you need to go to dl-files.psf.io and move all the files in place over there. Our policy is that every Python version gets its own directory, but each directory contains all releases of that version. - ___ On dinsdale, cd /data/ftp.python.org/pub/python/X.Y.Z - creating it if necessary. Make sure it is owned by group 'webmaster' - and group-writable. + ___ On dl-files.psf.io, cd /srv/www.python.org/ftp/python/X.Y.Z creating + it if necessary. Make sure it is owned by group 'downloads' and + group-writable. ___ Move the release .tgz, and .tar.xz files into place, as well as the .asc GPG signature files. The Win/Mac binaries are usually put there by the experts themselves. Make sure they are world readable. They should also be group - writable, and group-owned by webmaster. + writable, and group-owned by downloads. ___ Use `gpg --verify` to make sure they got uploaded intact. ___ If this is a final release: Move the doc zips and tarballs to - /data/ftp.python.org/pub/python/doc/X.Y.Z creating the directory + /srv/www.python.org/ftp/python/doc/X.Y.Z, creating the directory if necessary, and adapt the "current" symlink in .../doc to point to that directory. Note though that if you're releasing a maintenance release for an older version, don't change the current link. ___ If this is a final release (even a maintenance release), also unpack the HTML docs to - /data/ftp.python.org/pub/docs.python.org/release/X.Y.Z. + /data/ftp.python.org/pub/docs.python.org/release/X.Y.Z on dinsdale. If it is a release of a security-fix-only version, tell the DE to build a version with the "version switcher" and put it there. @@ -492,7 +492,7 @@ production release" paragraph as appropriate Note, you don't have to copy any release files into this directory; - they only live on dinsdale in the ftp directory. + they only live on dl-files.psf.io in the ftp directory. ___ Edit `download/releases/content.ht` to update the version numbers for this release. There are a bunch of places you need to touch: -- Repository URL: http://hg.python.org/peps

1 0

cpython (2.7): #22215: have the smtplib 'quit' command reset the state.
by r.david.murray Aug. 30, 2014

Aug. 30, 2014

http://hg.python.org/cpython/rev/7288519594de changeset: 92273:7288519594de branch: 2.7 parent: 92261:3e7f88550788 user: R David Murray <rdmurray(a)bitdance.com> date: Sat Aug 30 16:56:49 2014 -0400 summary: #22215: have the smtplib 'quit' command reset the state. Without this reset, starttls would fail if a connect/starttls was done after a quit, because smtplib assumed the existing value of emspt_features was accurate, but it gets reset when starttls completes (and the new value does not contain the starttls capability, since tls is already started at that point). (There may be additional places where this lack of reset was an issue as well.) Patch by Milan Oberkirch. files: Lib/smtplib.py | 4 ++++ Lib/test/test_smtplib.py | 15 +++++++++++++++ Misc/NEWS | 4 ++++ 3 files changed, 23 insertions(+), 0 deletions(-) diff --git a/Lib/smtplib.py b/Lib/smtplib.py --- a/Lib/smtplib.py +++ b/Lib/smtplib.py @@ -754,6 +754,10 @@ def quit(self): """Terminate the SMTP session.""" res = self.docmd("quit") + # A new EHLO is required after reconnecting with connect() + self.ehlo_resp = self.helo_resp = None + self.esmtp_features = {} + self.does_esmtp = False self.close() return res diff --git a/Lib/test/test_smtplib.py b/Lib/test/test_smtplib.py --- a/Lib/test/test_smtplib.py +++ b/Lib/test/test_smtplib.py @@ -507,6 +507,21 @@ #TODO: add tests for correct AUTH method fallback now that the #test infrastructure can support it. + def test_quit_resets_greeting(self): + smtp = smtplib.SMTP(HOST, self.port, + local_hostname='localhost', + timeout=15) + code, message = smtp.ehlo() + self.assertEqual(code, 250) + self.assertIn('size', smtp.esmtp_features) + smtp.quit() + self.assertNotIn('size', smtp.esmtp_features) + smtp.connect(HOST, self.port) + self.assertNotIn('size', smtp.esmtp_features) + smtp.ehlo_or_helo_if_needed() + self.assertIn('size', smtp.esmtp_features) + smtp.quit() + def test_main(verbose=None): test_support.run_unittest(GeneralTests, DebuggingServerTests, diff --git a/Misc/NEWS b/Misc/NEWS --- a/Misc/NEWS +++ b/Misc/NEWS @@ -19,6 +19,10 @@ Library ------- +- Issue #22216: smtplib now resets its state more completely after a quit. The + most obvious consequence of the previous behavior was a STARTTLS failure + during a connect/starttls/quit/connect/starttls sequence. + - Issue #21305: os.urandom now caches a fd to /dev/urandom. This is a PEP 466 backport from Python 3. -- Repository URL: http://hg.python.org/cpython

1 0

cpython (merge 3.4 -> default): Merge #22215: have the smtplib 'quit' command reset the state.
by r.david.murray Aug. 30, 2014

Aug. 30, 2014

http://hg.python.org/cpython/rev/d0d4ab0ba70e changeset: 92272:d0d4ab0ba70e parent: 92270:c499cc2c4a06 parent: 92271:a058760cb069 user: R David Murray <rdmurray(a)bitdance.com> date: Sat Aug 30 16:55:45 2014 -0400 summary: Merge #22215: have the smtplib 'quit' command reset the state. files: Lib/smtplib.py | 4 ++++ Lib/test/test_smtplib.py | 15 +++++++++++++++ Misc/NEWS | 4 ++++ 3 files changed, 23 insertions(+), 0 deletions(-) diff --git a/Lib/smtplib.py b/Lib/smtplib.py --- a/Lib/smtplib.py +++ b/Lib/smtplib.py @@ -891,6 +891,10 @@ def quit(self): """Terminate the SMTP session.""" res = self.docmd("quit") + # A new EHLO is required after reconnecting with connect() + self.ehlo_resp = self.helo_resp = None + self.esmtp_features = {} + self.does_esmtp = False self.close() return res diff --git a/Lib/test/test_smtplib.py b/Lib/test/test_smtplib.py --- a/Lib/test/test_smtplib.py +++ b/Lib/test/test_smtplib.py @@ -876,6 +876,21 @@ str(err)) smtp.close() + def test_quit_resets_greeting(self): + smtp = smtplib.SMTP(HOST, self.port, + local_hostname='localhost', + timeout=15) + code, message = smtp.ehlo() + self.assertEqual(code, 250) + self.assertIn('size', smtp.esmtp_features) + smtp.quit() + self.assertNotIn('size', smtp.esmtp_features) + smtp.connect(HOST, self.port) + self.assertNotIn('size', smtp.esmtp_features) + smtp.ehlo_or_helo_if_needed() + self.assertIn('size', smtp.esmtp_features) + smtp.quit() + def test_with_statement(self): with smtplib.SMTP(HOST, self.port) as smtp: code, message = smtp.noop() diff --git a/Misc/NEWS b/Misc/NEWS --- a/Misc/NEWS +++ b/Misc/NEWS @@ -124,6 +124,10 @@ Library ------- +- Issue #22216: smtplib now resets its state more completely after a quit. The + most obvious consequence of the previous behavior was a STARTTLS failure + during a connect/starttls/quit/connect/starttls sequence. + - Issue #22098: ctypes' BigEndianStructure and LittleEndianStructure now define an empty __slots__ so that subclasses don't always get an instance dict. Patch by Claudiu Popa. -- Repository URL: http://hg.python.org/cpython

1 0

cpython (3.4): #22215: have the smtplib 'quit' command reset the state.
by r.david.murray Aug. 30, 2014

Aug. 30, 2014

http://hg.python.org/cpython/rev/a058760cb069 changeset: 92271:a058760cb069 branch: 3.4 parent: 92268:4cce39cfe46c user: R David Murray <rdmurray(a)bitdance.com> date: Sat Aug 30 16:51:59 2014 -0400 summary: #22215: have the smtplib 'quit' command reset the state. Without this reset, starttls would fail if a connect/starttls was done after a quit, because smtplib assumed the existing value of emspt_features was accurate, but it gets reset when starttls completes (and the new value does not contain the starttls capability, since tls is already started at that point). (There may be additional places where this lack of reset was an issue as well.) Patch by Milan Oberkirch. files: Lib/smtplib.py | 4 ++++ Lib/test/test_smtplib.py | 15 +++++++++++++++ Misc/NEWS | 4 ++++ 3 files changed, 23 insertions(+), 0 deletions(-) diff --git a/Lib/smtplib.py b/Lib/smtplib.py --- a/Lib/smtplib.py +++ b/Lib/smtplib.py @@ -866,6 +866,10 @@ def quit(self): """Terminate the SMTP session.""" res = self.docmd("quit") + # A new EHLO is required after reconnecting with connect() + self.ehlo_resp = self.helo_resp = None + self.esmtp_features = {} + self.does_esmtp = False self.close() return res diff --git a/Lib/test/test_smtplib.py b/Lib/test/test_smtplib.py --- a/Lib/test/test_smtplib.py +++ b/Lib/test/test_smtplib.py @@ -858,6 +858,21 @@ self.assertIn(sim_auth_login_password, str(err)) smtp.close() + def test_quit_resets_greeting(self): + smtp = smtplib.SMTP(HOST, self.port, + local_hostname='localhost', + timeout=15) + code, message = smtp.ehlo() + self.assertEqual(code, 250) + self.assertIn('size', smtp.esmtp_features) + smtp.quit() + self.assertNotIn('size', smtp.esmtp_features) + smtp.connect(HOST, self.port) + self.assertNotIn('size', smtp.esmtp_features) + smtp.ehlo_or_helo_if_needed() + self.assertIn('size', smtp.esmtp_features) + smtp.quit() + def test_with_statement(self): with smtplib.SMTP(HOST, self.port) as smtp: code, message = smtp.noop() diff --git a/Misc/NEWS b/Misc/NEWS --- a/Misc/NEWS +++ b/Misc/NEWS @@ -27,6 +27,10 @@ Library ------- +- Issue #22216: smtplib now resets its state more completely after a quit. The + most obvious consequence of the previous behavior was a STARTTLS failure + during a connect/starttls/quit/connect/starttls sequence. + - Issue #22185: Fix an occasional RuntimeError in threading.Condition.wait() caused by mutation of the waiters queue without holding the lock. Patch by Doug Zongker. -- Repository URL: http://hg.python.org/cpython

1 0

peps: PEP 476: Updates based on feedback from python-dev
by alex.gaynor Aug. 30, 2014

Aug. 30, 2014

http://hg.python.org/peps/rev/92150c23e648 changeset: 5531:92150c23e648 user: Alex Gaynor <alex.gaynor(a)gmail.com> date: Sat Aug 30 10:55:06 2014 -0700 summary: PEP 476: Updates based on feedback from python-dev files: pep-0476.txt | 8 ++++++++ 1 files changed, 8 insertions(+), 0 deletions(-) diff --git a/pep-0476.txt b/pep-0476.txt --- a/pep-0476.txt +++ b/pep-0476.txt @@ -87,6 +87,10 @@ * Using the platform stores makes it easier to handle situations such as corporate internal CAs. +OpenSSL also has a pair of environment variables, ``SSL_CERT_DIR`` and +``SSL_CERT_FILE`` which can be used to point Python at a different certificate +database. + Backwards compatibility ----------------------- @@ -107,6 +111,10 @@ Twisted's 14.0 release made this same change, and it has been met with almost no opposition. +In order to make this transition as smooth as possible, the next 3.4.x release +following this PEP will be modified to emit a warning in cases that would raise +an Exception in Python 3.5. + Other protocols =============== -- Repository URL: http://hg.python.org/peps

1 0

Daily reference leaks (c499cc2c4a06): sum=149135
by solipsis＠pitrou.net Aug. 30, 2014

Aug. 30, 2014

results for c499cc2c4a06 on branch "default" -------------------------------------------- test_codecs leaked [5825, 5825, 5825] references, sum=17475 test_codecs leaked [1172, 1174, 1174] memory blocks, sum=3520 test_distutils leaked [37000, 37000, 37000] references, sum=111000 test_distutils leaked [5711, 5713, 5713] memory blocks, sum=17137 test_functools leaked [0, 0, 3] memory blocks, sum=3 test_site leaked [0, 2, -2] references, sum=0 test_site leaked [0, 2, -2] memory blocks, sum=0 Command line was: ['./python', '-m', 'test.regrtest', '-uall', '-R', '3:3:/home/antoine/cpython/refleaks/reflog3wyCw8', '-x']

1 0

cpython: Issue #22098: ctypes' BigEndianStructure and LittleEndianStructure now define
by antoine.pitrou Aug. 29, 2014

Aug. 29, 2014

http://hg.python.org/cpython/rev/c499cc2c4a06 changeset: 92270:c499cc2c4a06 user: Antoine Pitrou <solipsis(a)pitrou.net> date: Sat Aug 30 00:37:18 2014 +0200 summary: Issue #22098: ctypes' BigEndianStructure and LittleEndianStructure now define an empty __slots__ so that subclasses don't always get an instance dict. Patch by Claudiu Popa. files: Lib/ctypes/_endian.py | 2 ++ Lib/ctypes/test/test_byteswap.py | 20 ++++++++++++++++++++ Misc/NEWS | 4 ++++ 3 files changed, 26 insertions(+), 0 deletions(-) diff --git a/Lib/ctypes/_endian.py b/Lib/ctypes/_endian.py --- a/Lib/ctypes/_endian.py +++ b/Lib/ctypes/_endian.py @@ -45,6 +45,7 @@ class BigEndianStructure(Structure, metaclass=_swapped_meta): """Structure with big endian byte order""" + __slots__ = () _swappedbytes_ = None elif sys.byteorder == "big": @@ -53,6 +54,7 @@ BigEndianStructure = Structure class LittleEndianStructure(Structure, metaclass=_swapped_meta): """Structure with little endian byte order""" + __slots__ = () _swappedbytes_ = None else: diff --git a/Lib/ctypes/test/test_byteswap.py b/Lib/ctypes/test/test_byteswap.py --- a/Lib/ctypes/test/test_byteswap.py +++ b/Lib/ctypes/test/test_byteswap.py @@ -22,6 +22,26 @@ setattr(bits, "i%s" % i, 1) dump(bits) + def test_slots(self): + class BigPoint(BigEndianStructure): + __slots__ = () + _fields_ = [("x", c_int), ("y", c_int)] + + class LowPoint(LittleEndianStructure): + __slots__ = () + _fields_ = [("x", c_int), ("y", c_int)] + + big = BigPoint() + little = LowPoint() + big.x = 4 + big.y = 2 + little.x = 2 + little.y = 4 + with self.assertRaises(AttributeError): + big.z = 42 + with self.assertRaises(AttributeError): + little.z = 24 + def test_endian_short(self): if sys.byteorder == "little": self.assertIs(c_short.__ctype_le__, c_short) diff --git a/Misc/NEWS b/Misc/NEWS --- a/Misc/NEWS +++ b/Misc/NEWS @@ -124,6 +124,10 @@ Library ------- +- Issue #22098: ctypes' BigEndianStructure and LittleEndianStructure now + define an empty __slots__ so that subclasses don't always get an instance + dict. Patch by Claudiu Popa. + - Issue #22185: Fix an occasional RuntimeError in threading.Condition.wait() caused by mutation of the waiters queue without holding the lock. Patch by Doug Zongker. -- Repository URL: http://hg.python.org/cpython

1 0