[issue27164] zlib can't decompress DEFLATE using shared dictionary

Xiang Zhang added the comment: Thanks for your review, Martin. :) I have replied with after thinking and try. Yes, this is more like a feature request. But for the end user, this behaviour is quite like a bug. > If we were to add full support of inflateSetDictionary() for raw deflate > mode,

[issue27179] subprocess uses wrong encoding on Windows

Martin Panter added the comment: I think Issue 6135 has a bit of discussion on adding encoding and error parameters to subprocess.Popen etc. -- ___ Python tracker

[issue27025] More human readable generated widget names

Raymond Hettinger added the comment: +1 I think would be a nice improvement. -- nosy: +rhettinger ___ Python tracker ___

[issue27179] subprocess uses wrong encoding on Windows

Eryk Sun added the comment: There is no right encoding as far as I can see. If it's attached to a console (i.e. conhost.exe), then cmd.exe uses the console's output codepage when writing to a pipe or file, which is the scenario that your patch attempts to address. But if you pass

[issue27179] subprocess uses wrong encoding on Windows

Steve Dower added the comment: Even sys.__stdout__ can be missing. In this context, falling back on the default encoding is probably fine, but for 3.6 I'd like to make everything default to UTF-8 on Windows, and force the console mode on startup (restore on finalize) - apart from the input()

[issue27164] zlib can't decompress DEFLATE using shared dictionary

Martin Panter added the comment: Thanks for the patches. Maybe Gregory knows more about this than I do (I never used zdict). But I think I have figured it out today, so I left some review comments :) I’m not sure whether this really is a bug fix, although it would have little if any impact

[issue27130] zlib: OverflowError while trying to compress 2^32 bytes or more

Xiang Zhang added the comment: Yes. It's compression object not compress. I find more. The overflow checking is introduced to solve problem in issue8650. It seems the explicit overflow checking is introduced to keep compatibility with py2 (py2 raises the overflowerror in pyargparse). I

[issue20825] containment test for "ip_network in ip_network"

James Schneider added the comment: I'd like to ask for a status on getting this merged? As a network administrator, these changes would have a magical effect on my code dealing with routing tables and ACL's. -- nosy: +James Schneider ___ Python

[issue18085] Verifying refcounts.dat

Fred L. Drake, Jr. added the comment: I don't think this is a duplicate of issue 9755; this relates to verifying the data, and that revolves around possible process improvements. Whether this issue should be closed is tied to whether the file has been verified, as the issue title suggests. I

[issue9755] Fix refcounting details in Py3k C API documentation

Fred L. Drake, Jr. added the comment: As mentioned in issue 18085, the original file was not generated, but crafted by hand (though I don't think that really matters). -- ___ Python tracker

[issue18384] Add devhelp build instructions to the documentation makefile

Berker Peksag added the comment: I think we should avoid adding distro specific tools into CPython repository (and the another reason to reject this is that the other distros will probably want to add their tools). -- nosy: +berker.peksag resolution: -> rejected stage: patch review

[issue18085] Verifying refcounts.dat

Berker Peksag added the comment: Can we now close this as a duplicate of issue 9755? -- nosy: +berker.peksag ___ Python tracker ___

[issue27179] subprocess uses wrong encoding on Windows

Martin Panter added the comment: I don’t know much about the conventions for stdout etc encoding on Windows. But in general, the patch does not seem robust. Does it work if sys.stdout is a pipe or file (not a console)? I doubt it will work when sys.stdout has been replaced by e.g. StringIO,

[issue20693] Sidebar scrolls down 2x as fast as page content

Changes by Berker Peksag : -- status: pending -> closed ___ Python tracker ___ ___

[issue8519] doc: termios and ioctl reference links

Berker Peksag added the comment: +1 for using the manpage directives. -- keywords: +easy nosy: +berker.peksag stage: -> needs patch type: -> enhancement versions: +Python 3.5, Python 3.6 ___ Python tracker

[issue22374] Replace contextmanager example and improve explanation

Berker Peksag added the comment: We also have swap_attr() and swap_item() helpers in Lib/test/support/__init__.py. I've used a simplified version of swap_attr() in my patch. Since this is basically a document improvement I removed Python 2.7 from the versions field. -- keywords:

[issue6443] _winreg.QueryValueEx confused on double null terminated strings

Tony Grue added the comment: I see this was closed,though it sounds like it's because a work around is found, not because the code now handles multiple null values in a row. Looking at code from the github mirror (pasted below); it looks like the issue is that the countStrings function breaks

[issue27179] subprocess uses wrong encoding on Windows

Dāvis added the comment: I looked at #27048 and indeed it's affected by this bug, it happens to me too (I've non-ASCII symbols in %PATH%) and this my patch fixes that. on my system without patch > python -c "from distutils import _msvccompiler; _msvccompiler._get_vc_env('')" Traceback (most

[issue27025] More human readable generated widget names

Terry J. Reedy added the comment: Try this instead. import tkinter as tk from idlelib.configdialog import ConfigDialog from idlelib import macosx _realsetup = tk.BaseWidget._setup def _wrapsetup(self, master, cnf): _realsetup(self, master, cnf) print(self.widgetName, self._w)

[issue12243] getpass.getuser works on OSX

Roundup Robot added the comment: New changeset 98870f27d2ed by Berker Peksag in branch '2.7': Issue #12243: Remove redundant availability notes from getpass docs https://hg.python.org/cpython/rev/98870f27d2ed -- ___ Python tracker

[issue22558] Missing doc links to source code for Python-coded modules.

Yoni Lavi added the comment: Ah, thanks for the explanation. I think in this case we're ok too. I linked io.rst to Lib/io.py, which relies on the C version. And in the case of decimal.rst, it was already linked to Lib/decimal.py which very explicitly attempts to load the C version and falls

[issue12243] getpass.getuser works on OSX

Berker Peksag added the comment: See issue 11233 for the availability directive patch. Thanks for the patch, Bryce. -- nosy: +berker.peksag resolution: -> fixed stage: -> resolved status: open -> closed type: -> behavior versions: +Python 3.6 -Python 3.4

[issue12243] getpass.getuser works on OSX

Roundup Robot added the comment: New changeset c13198a2007e by Berker Peksag in branch '3.5': Issue #12243: Remove redundant availability notes from getpass docs https://hg.python.org/cpython/rev/c13198a2007e New changeset 1bd3daae351d by Berker Peksag in branch 'default': Issue #12243: Merge

[issue27179] subprocess uses wrong encoding on Windows

Dāvis added the comment: there's no such "ā" command, it's just used to get non-ASCII output cmd will return: 'ā' is not recognized as an internal or external command, operable program or batch file. and this will be encoded in consoles encoding (UTF8 in my example or whatever chcp is set

[issue27179] subprocess uses wrong encoding on Windows

STINNER Victor added the comment: What is the ā command? Where does it come from? What is its output? Do you know the encoding of its output? It remembers me the "set" issue in distutils: issue #27048. -- ___ Python tracker

[issue27025] More human readable generated widget names

Terry J. Reedy added the comment: To make the experiment easy to repeat, say after the patch, I wrote this. (Perhaps inspired by Ned Batchelder's 'machete debugging' PyCon talk.) import tkinter as tk from idlelib.configdialog import ConfigDialog _realsetup = tk.BaseWidget._setup def

[issue27179] subprocess uses wrong encoding on Windows

New submission from Dāvis: subprocess uses wrong encoding on Windows. On Windows 10 with Python 3.5.1 from Command Prompt (cmd.exe) > chcp 65001 > python -c "import subprocess; subprocess.getstatusoutput('ā')" Traceback (most recent call last): File "", line 1, in File

[issue22558] Missing doc links to source code for Python-coded modules.

R. David Murray added the comment: Yes, those should be OK. I was thinking of _pyio and _pydecimal, but I don't know if you are linking to those. -- ___ Python tracker

[issue27130] zlib: OverflowError while trying to compress 2^32 bytes or more

Martin Panter added the comment: You should be able to use a compression (or decompression) object as a workaround. But calling zlib.compress() multiple times would generate multiple separate deflated streams, which is different. I think it is reasonable for Python to handle larger data sizes

[issue16964] Add 'm' format specifier for mon_grouping etc.

Changes by R. David Murray : -- stage: needs patch -> patch review versions: +Python 3.6 -Python 3.4 ___ Python tracker ___

[issue19012] urllib2: bad proxy configuration throws "getaddrinfo" error

R. David Murray added the comment: Closing because no reproducer provided. -- resolution: -> works for me stage: -> resolved status: open -> closed ___ Python tracker

[issue22558] Missing doc links to source code for Python-coded modules.

Yoni Lavi added the comment: Thank you for looking at this. David, concerning modules that have multiple sources, are you referring to modules such as `stat` and `heapq`? In these, it's the Python source that tries to import the C module if it's available, so I'd consider it ok to leave just

[issue22558] Missing doc links to source code for Python-coded modules.

Terry J. Reedy added the comment: If someone does a full review (looks at each diff in Rietveld to see if it matches the intended format -- see Yoni's 4-19 message, with :source:`Lib/mod.py` (lib2to3 is an exception) following **Source code:** -- and then downloads patch, applies, builds new

[issue25931] os.fork() command distributed in windows Python27 (in SocketServer module)

R. David Murray added the comment: They aren't currently defined conditionally, though, and haven't ever been, so I'm not sure it is worth chaning that. However, we should probably add an 'availability' line, since they aren't (effectively) available on Windows. -- assignee: ->

[issue26582] asyncio documentation links to wrong CancelledError

R. David Murray added the comment: ping: Victor or Yuri, do agree that asyncio.CancelledError should be documented? I know that this duality has confused me in the past as well. I don't know if having it documented would have helped me ahead of time, but at least it could have cleared up my

[issue27164] zlib can't decompress DEFLATE using shared dictionary

Changes by Gregory P. Smith : -- assignee: -> gregory.p.smith nosy: +gregory.p.smith ___ Python tracker ___

[issue27178] Unconverted RST marking in interpreter tutorial

Berker Peksag added the comment: Good catch! Thanks for patch, Nathan. -- nosy: +berker.peksag resolution: -> fixed stage: -> resolved status: open -> closed type: enhancement -> behavior ___ Python tracker

[issue27178] Unconverted RST marking in interpreter tutorial

Roundup Robot added the comment: New changeset a2ffc966363e by Berker Peksag in branch 'default': Issue #27178: Fix markup in tutorial https://hg.python.org/cpython/rev/a2ffc966363e -- nosy: +python-dev ___ Python tracker

[issue27178] Unconverted RST marking in interpreter tutorial

New submission from Nathan Harold: There's a bit of uninterpreted RST markup: 3.6.. _tut-using: at the top of the second section of the tutorial (https://docs.python.org/3.6/tutorial/interpreter.html) in the documentation for 3.6 only. (I can see this in the offline copy I received with

[issue27025] More human readable generated widget names

Serhiy Storchaka added the comment: You can apply show_widget_names.diff to see all widget names. Run IDLE and open the configuration dialog for example. -- Added file: http://bugs.python.org/file43092/show_widget_names.diff ___ Python tracker

[issue22558] Missing doc links to source code for Python-coded modules.

R. David Murray added the comment: I meant full review, not file review. -- ___ Python tracker ___ ___

[issue22558] Missing doc links to source code for Python-coded modules.

R. David Murray added the comment: You mean someone doing a file review at the sprints? I can add it to the list. No guarantees about it getting committed at the sprints, though. If we are linking to the python version of modules that have both python and C versions, IMO the existence of

[issue27173] Modern Unix key bindings for IDLE

Serhiy Storchaka added the comment: New patch makes IDLE silently fall back to default key set and adds a note about new key set on Keys configuration tab. -- Added file: http://bugs.python.org/file43091/idle_modern_unix_key_set3.patch ___ Python

[issue27172] Add skip_bound_arg argument to inspect.Signature.from_callable()

Changes by Ryan Petrello : Added file: http://bugs.python.org/file43090/signature-from-callable-skip-bound-arg.patch ___ Python tracker

[issue22558] Missing doc links to source code for Python-coded modules.

Terry J. Reedy added the comment: [David, could this be a sprint issue?] Here is the difference between python and C coded modules. The x.y docs are for all implementations of x.y. The python-coded modules are intended for use by all implementations, perhaps with minor modifications. In

[issue27172] Add skip_bound_arg argument to inspect.Signature.from_callable()

Changes by Ryan Petrello : Added file: http://bugs.python.org/file43089/signature-from-callable-skip-bound-arg.patch ___ Python tracker

[issue23116] Python Tutorial 4.7.1: Improve ask_ok() to cover more input values

Berker Peksag added the comment: Here is a patch. -- keywords: +patch stage: needs patch -> patch review Added file: http://bugs.python.org/file43088/issue23116.diff ___ Python tracker

[issue27172] Add skip_bound_arg argument to inspect.Signature.from_callable()

Changes by SilentGhost : -- assignee: -> yselivanov stage: patch review -> commit review ___ Python tracker ___

[issue23116] Python Tutorial 4.7.1: Improve ask_ok() to cover more input values

Fred L. Drake, Jr. added the comment: +1 for ValueError instead of OSError. -- nosy: +fdrake ___ Python tracker ___

[issue27172] Add skip_bound_arg argument to inspect.Signature.from_callable()

Yury Selivanov added the comment: Please don't commit this without my review. -- ___ Python tracker ___ ___

[issue23116] Python Tutorial 4.7.1: Improve ask_ok() to cover more input values

Berker Peksag added the comment: I agree with all of Carol's suggestions. Also, OSError could be replaced with a more appropriate exception such as ValueError. -- nosy: +berker.peksag stage: -> needs patch versions: +Python 3.5, Python 3.6 -Python 3.4

[issue27173] Modern Unix key bindings for IDLE

Terry J. Reedy added the comment: Original IDLE Dark patch: #24820. Follow-up for cross-version issues (initially for themes, but with a eye to keys, etc): #25313. I intended that 'name2' be future proof (msg252372). I believe my idea was that if it were to point to a future 4th default

[issue27172] Add skip_bound_arg argument to inspect.Signature.from_callable()

Changes by Berker Peksag : -- stage: test needed -> patch review ___ Python tracker ___

[issue27172] Add skip_bound_arg argument to inspect.Signature.from_callable()

Changes by Ryan Petrello : Added file: http://bugs.python.org/file43087/signature-from-callable-skip-bound-arg.patch ___ Python tracker

[issue26589] Add HTTP Response code 451

Nathan Harold added the comment: Added it to the table at https://docs.python.org/3.6/library/http.html#http-status-codes and put a versionadded notice at the bottom. -- nosy: +nharold Added file: http://bugs.python.org/file43086/doctable.diff ___

[issue27051] Create PIP gui

Upendra Kumar added the comment: Actually now we don't need "pip list" command to get list of installed distributions. We can directly get it : from pip_tkinter._vendor import pkg_resources installed_packages = [p.project_name for p in pkg_resources.working_set] --

[issue27051] Create PIP gui

Upendra Kumar added the comment: @Terry, code updates can be followed on https://github.com/upendra-k14/pip_gui/tree/dump_code Please review the code in dump_code branch. For now, only pip_gui_v3.py is relevant. 1. I had doubt about including a private module from pip ( "pip._vendor" ). As

[issue27174] Update URL to IPython in interactive.rst

Berker Peksag added the comment: Thanks for the report, kulla. -- nosy: +berker.peksag resolution: -> fixed stage: needs patch -> resolved status: open -> closed ___ Python tracker

[issue27174] Update URL to IPython in interactive.rst

Roundup Robot added the comment: New changeset 92fd6935e882 by Berker Peksag in branch '3.5': Issue #27174: Update IPython URL in tutorial https://hg.python.org/cpython/rev/92fd6935e882 New changeset 87d76611470c by Berker Peksag in branch 'default': Issue #27174: Merge from 3.5

[issue27176] Addition of assertNotRaises

Raymond Hettinger added the comment: I concur with David snd Steven, so closing this as rejected. -- nosy: +rhettinger resolution: -> rejected status: open -> closed ___ Python tracker

[issue1521950] shlex.split() does not tokenize like the shell

Andrey Kislyuk added the comment: Is there any chance of getting this into 3.6? We are still in a situation where the shlex module misleads developers into believing that it has functionality to parse things the way the shell does. I've had to vendor the copy of shlex with patches from this

[issue27177] re match.group should support __index__

Serhiy Storchaka added the comment: This looks as new feature and can go only in 3.6. -- nosy: +serhiy.storchaka stage: -> needs patch type: -> enhancement versions: +Python 3.6 -Python 2.7 ___ Python tracker

[issue27176] Addition of assertNotRaises

R. David Murray added the comment: Agreed. I've never used the distinction between errors and failures myself. -- nosy: +r.david.murray ___ Python tracker

[issue27176] Addition of assertNotRaises

Steven D'Aprano added the comment: Yes. What does such an assertion actually mean? Why would I write `self.assertNotRaises(ValueError, spam, arg)` rather than just call `spam(arg)`? The only difference is that assertNotRaises will treat one specific exception as a test failure rather than a

[issue27164] zlib can't decompress DEFLATE using shared dictionary

Xiang Zhang added the comment: Forget to add test. Upload a new one with test included. -- Added file: http://bugs.python.org/file43085/issue27164_with_test.patch ___ Python tracker

[issue7105] weak dict iterators are fragile because of unpredictable GC runs

Charles-François Natali added the comment: Shouldn't the documentation be updated? https://docs.python.org/3.6/library/weakref.html#weakref.WeakKeyDictionary Note Caution: Because a WeakKeyDictionary is built on top of a Python dictionary, it must not change size when iterating over it. This

[issue27176] Addition of assertNotRaises

Changes by SilentGhost : -- nosy: +ezio.melotti, michael.foord, rbcollins versions: +Python 3.6 ___ Python tracker ___

[issue27177] re match.group should support __index__

New submission from Jeroen Demeyer: ``` >>> class zero(object): ... def __index__(self): ... return 0 ... >>> z = zero() >>> import re >>> p = re.compile('(a)b') >>> m = p.match('ab') >>> m.group(0) 'ab' >>> m.group(z) Traceback (most recent call last): File "", line 1, in

Re: [issue23459] Linux: expose the new execveat() syscall

On 05/30/2016 06:12 PM, Марк Коренберг wrote: Марк Коренберг added the comment: It is actual to me due to it's AT_EMPTY_PATH feature (I download script, verify signature, and want to run it) See also `man 3 fexecve` Why not to add `dir_fd=None` ? (as in `os.rmdir()` for example). It

[issue27176] Addition of assertNotRaises

New submission from Bar Harel: I thought of implementing an assertNotRaises to solve the issue of using try...except...fail in order to prevent showing the tests as an error and instead show them as a failure. Is there anything I should consider while implementing it? -- components:

[issue27172] Add skip_bound_arg argument to inspect.Signature.from_callable()

Changes by ryan.petrello : -- nosy: +eric.snow ___ Python tracker ___ ___

[issue27173] Modern Unix key bindings for IDLE

Serhiy Storchaka added the comment: > 3 or more years ago, there was an issue about needing to define pairs like > Control-Key-c and Control-Key-C in order to have crtl + c/C-key work the same > regardless of the caps lock setting. I also remember that the effect of Caps > lock was system

[issue27164] zlib can't decompress DEFLATE using shared dictionary

Xiang Zhang added the comment: I think my conclusion in last message is right. A similar bug in CPAN can be found in https://rt.cpan.org/Public/Bug/Display.html?id=36046. And I can find in zlib's changelog inflateSetDictionary starts to support raw inflate from 1.2.2.1. Patch attached now

[issue27172] Add skip_bound_arg argument to inspect.Signature.from_callable()

Changes by ryan.petrello : Added file: http://bugs.python.org/file43081/signature-from-callable-skip-bound-arg.patch ___ Python tracker

[issue27172] Add skip_bound_arg argument to inspect.Signature.from_callable()

ryan.petrello added the comment: Thanks, I'll upload a new patch shortly that addresses these. -- ___ Python tracker ___

[issue12855] linebreak sequences should be better documented

Martin Panter added the comment: Here is an updated patch for Python 2, based on Benjamin’s commit, Matthew’s earlier py27 patch, and Alexander’s backport of related changes from Python 3. Let me know what you think. -- nosy: +Alexander Schrijver stage: resolved -> patch review Added

[issue25570] urllib.request > Request.add_header("abcd", "efgh") fails with character ":" in first parameter string

Changes by Martin Panter : -- resolution: -> fixed stage: commit review -> resolved status: open -> closed ___ Python tracker ___

[issue27175] Unpickling Path objects

Changes by SilentGhost : -- nosy: +pitrou type: -> behavior versions: -Python 3.5 ___ Python tracker ___

[issue26739] idle: Errno 10035 a non-blocking socket operation could not be completed immediately

Kristján Valur Jónsson added the comment: Hi there, everyone. I'm sorry for my rash remarks about the state of IDLE, I'm sure it is alive and well and its good to see that fine people like Terry are working on keeping it up to date. Michael, please understand that python developers are

[issue27175] Unpickling Path objects

New submission from Antony Lee: Currently, pickling Path objects lead to issues when working across platforms: Paths (and, thus, objects that contain Paths) created on a POSIX platform (PosixPaths) cannot be unpickled on Windows and vice versa. There are a few possibilities around this

[issue27173] Modern Unix key bindings for IDLE

Terry J. Reedy added the comment: Seems reasonable. 3 or more years ago, there was an issue about needing to define pairs like Control-Key-c and Control-Key-C in order to have crtl + c/C-key work the same regardless of the caps lock setting. I also remember that the effect of Caps lock was

[issue25570] urllib.request > Request.add_header("abcd", "efgh") fails with character ":" in first parameter string

Roundup Robot added the comment: New changeset 320b9a65ac07 by Martin Panter in branch '3.5': Issue #25570: Add example of customizing User-Agent via add_header() https://hg.python.org/cpython/rev/320b9a65ac07 New changeset 75dc64c8c22b by Martin Panter in branch 'default': Issue #25570: Merge

[issue27172] Add skip_bound_arg argument to inspect.Signature.from_callable()

New submission from SilentGhost: This could only go in 3.6, but it needs proper documentation and test(s). -- nosy: +SilentGhost, yselivanov stage: -> test needed versions: -Python 3.5 ___ Python tracker

[issue12855] linebreak sequences should be better documented

Martin Panter added the comment: Reopening to change the Python 2 documentation. A starting point may be Matthew’s patch and/or Alexander’s patch in Issue 22232. -- status: closed -> open ___ Python tracker

[issue27171] Fix various typos

Martin Panter added the comment: Thanks. The py2 version is partly a back port, but I have searched for the same kind of spelling mistakes and fixed them in different places, e.g. “overriden” in Lib/bsddb/test/test_all.py, which does not exist in Py 3. --

[issue27164] zlib can't decompress DEFLATE using shared dictionary

Xiang Zhang added the comment: This seems related with the difference between zlib format and raw format. When you do raw inflate, you have to inflateSetDictionary before any inflate. You cannot rely on the first inflate to return Z_NEED_DICT and then do inflateSetDictionary. Although I have

[issue27174] Update URL to IPython in interactive.rst

Changes by Serhiy Storchaka : -- assignee: -> docs@python components: +Documentation -Devguide keywords: +easy nosy: +docs@python stage: -> needs patch type: enhancement -> behavior versions: +Python 2.7, Python 3.5 ___

[issue27174] Update URL to IPython in interactive.rst

Changes by kulla : -- type: -> enhancement ___ Python tracker ___ ___

[issue24759] Idle: require tk 8.5 and ttk widgets, and drop unneeded code.

Serhiy Storchaka added the comment: My patch in issue24750 makes sense only if we want support Tcl/Tk 8.4. I think that if require Tcl/Tk 8.5+, we can use Ttk directly and unconditionally. I don't think that it can be not working. -- ___ Python

[issue27174] Update URL to IPython in interactive.rst

New submission from kulla: The file https://docs.python.org/3.6/tutorial/interactive.html has a dead link for IPython to http://ipython.scipy.org/ I guess the right link shall be https://ipython.org/ -- components: Devguide messages: 266802 nosy: ezio.melotti, kulla, willingc

[issue27173] Modern Unix key bindings for IDLE

Changes by Serhiy Storchaka : -- nosy: +terry.reedy ___ Python tracker ___ ___

[issue27173] Modern Unix key bindings for IDLE

New submission from Serhiy Storchaka: The "IDLE Classic Unix" key set in IDLE provides Emacs-like key bindings. But modern Linux environments (KDE, Gnome, etc) use different key bindings in standard editors. Proposed patch provides new key set named "IDLE Modern Unix" that uses key bindings

[issue22232] str.splitlines splitting on non-\r\n characters

Alexander Schrijver added the comment: I appeared to have missed the reference to that issue when I read this issue the first time. Re-opening that issue makes sense to me. -- ___ Python tracker

[issue27171] Fix various typos

Serhiy Storchaka added the comment: typos.py3.patch LGTM. I expect typos.py2.patch is just backported to 2.7. -- nosy: +serhiy.storchaka ___ Python tracker