[issue25220] Enhance and refactor test.regrtest (convert regrtest.py to a package)

Serhiy Storchaka added the comment: середа, 23-вер-2015 10:12:35 STINNER Victor написано: > New submission from STINNER Victor: > > The Lib/test/regrtest.py file became a monster. It's very long, it uses a > lot of global variables. It's hard to review and hard to maintain. I > propose to: > >

[issue25220] Enhance and refactor test.regrtest (convert regrtest.py to a package)

Changes by Jeremy Kloth : -- nosy: +jkloth ___ Python tracker ___ ___

[issue25214] asyncio ssl transport regression

Andrew Svetlov added the comment: I've missed your patch, sorry. Everything is fixed by http://bugs.python.org/issue25114 -- resolution: -> duplicate stage: -> resolved ___ Python tracker

[issue25131] The AST for dict and set displays has the lineno of the first value

Changes by Berker Peksag : -- keywords: +3.5regression ___ Python tracker ___ ___

[issue25220] Enhance and refactor test.regrtest (convert regrtest.py to a package)

New submission from STINNER Victor: The Lib/test/regrtest.py file became a monster. It's very long, it uses a lot of global variables. It's hard to review and hard to maintain. I propose to: - split single file regrtest.py into multiple files in a new Lib/test/regrtest/ package - refactor the

[issue25220] Enhance and refactor test.regrtest (convert regrtest.py to a package)

Berker Peksag added the comment: I like the new features, but I think we should take a look at issue 10967 before converting it to a package or refactoring it. The buildbot part is also a bit complicated. For example, support.verbose doesn't work correctly on builtbots: issue 23235.

[issue25220] Enhance and refactor test.regrtest (convert regrtest.py to a package)

STINNER Victor added the comment: Serhiy Storchaka wrote: """ You can just mover parts of the code into utility files (or to the test.support package) without converting regrtest.py to a package. This preserves compatibility with running Lib/test/regrtest.py as a script. """ Does it really

[issue25220] Enhance and refactor test.regrtest (convert regrtest.py to a package)

Martin Panter added the comment: I never used --single, but quickly looking at the code, it looks like the pynexttest file got stored in /tmp or similar, via tempfile.gettempdir(), so it should usually survive. However it looks like your patch now creates “pynexttest” in a temporary directory

[issue25214] asyncio ssl transport regression

STINNER Victor added the comment: > I've missed your patch, sorry. There is no need to be sorry :-) > Everything is fixed by http://bugs.python.org/issue25114 Wow, great :-) Thanks again Mathieu Pasquet who reported the issue #22768. -- status: open -> closed superseder: ->

[issue25220] Enhance and refactor test.regrtest (convert regrtest.py to a package)

STINNER Victor added the comment: Another recent example: http://buildbot.python.org/all/builders/AMD64%20Windows7%20SP1%203.x/builds/6712/steps/test/logs/stdio ... [395/399] test_asyncio [396/399] test_email [397/399] test_threaded_import [398/399] test_tools command timed out: 3900 seconds

[issue25205] setattr accepts invalid identifiers

eryksun added the comment: Even MvL appears to slip up when he states "some may accept non-strings as attribute names". That would be pointless in a __setattr__ method since setattr / PyObject_SetAttr reject non-string values as 'names'. -- ___

[issue23329] _ssl cannot be compiled with LibreSSL anymore (on OpenBSD 5.5) because of ALPN

Remi Pointel added the comment: With this patch it works fine on OpenBSD with LibreSSL. Could you test on a Linux machine with OpenSSL supporting ALPN please? Thanks spil@ for the idea. -- keywords: +patch Added file: http://bugs.python.org/file40556/python_have_alpn.diff

[issue16701] Docs missing the behavior of += (in-place add) for lists.

R. David Murray added the comment: Wording looks ok...except that technically it is not that 'n' is an integer, it's that 'n' can play the role of an integer (ie: it has an __index__ method). -- ___ Python tracker

[issue25210] Special-case NoneType() in do_richcompare()

R. David Murray added the comment: There are other places NoneType shows up (eg: attribute error). Why should this case be special? Do we want to fix them all? -- nosy: +r.david.murray ___ Python tracker

[issue25219] Update doc for Idle command line argument.

New submission from Terry J. Reedy: idle.rst is either wrong or obsolete. Some options are missing and the notes are not now correct. I am editing so it matches code and help message in PyShell.py. I believe there is a mismatch between code and help message with respect to interaction with

[issue22052] Comparison operators called in reverse order for subclasses with no override.

Mark Dickinson added the comment: > the most we should do is document the behaviour somewhere And indeed, perhaps this issue counts as sufficient documentation... -- ___ Python tracker

[issue25117] Windows installer: precompiling stdlib fails with missing DLL errors

Marius Gedminas added the comment: For the record, I rebooted once, after installing both 32-bit and 64-bit versions of Python 3.5. (Also, it seems that the Python 3.5 installer didn't install pip for me, which could be fallout from this bug? I had to run python -m ensurepip to get it.)

[issue12067] Doc: remove errors about mixed-type comparisons.

Martin Panter added the comment: I committed the changes to expressions.rst for 3.4+. That still leaves the changes to test_compare.py, and possibly changes for 2.7. Andy: In msg229721 you mentioned a potential 2.7 patch. Did you get anywhere with that? Even if it is only half finished,

[issue23329] _ssl cannot be compiled with LibreSSL anymore (on OpenBSD 5.5) because of ALPN

Bernard Spil added the comment: Checking if the method is actually defined is the smart way to go >From include/openssl/tls1.h # define TLSEXT_TYPE_application_layer_protocol_negotiation So #ifdef TLSEXT_TYPE_application_layer_protocol_negotiation would show you if the feature is available

[issue25198] Idle: improve idle.html help viewer.

Terry J. Reedy added the comment: List indent glitch: In the command line section, after the box, idle.rst has If there are arguments: #. If ``-``, ``-c``, or ``r`` is used, all arguments are placed in ``sys.argv[1:...]`` and ``sys.argv[0]`` is set to ``''``, ``'-c'``, or ``'-r'``. No

[issue22052] Comparison operators called in reverse order for subclasses with no override.

Mark Dickinson added the comment: For Python 2, I think the most we should do is document the behaviour somewhere; changing it in a bugfix release seems both unnecessary and potentially risky. -- ___ Python tracker

[issue19143] Finding the Windows version getting messier (detect windows 8.1?)

Marc-Andre Lemburg added the comment: On 23.09.2015 02:33, Steve Dower wrote: > > Done. Thanks. > As this is (meant to be) a purely informational/diagnostic module, seems like > a good idea to fix every version we're supporting in any way. Yep, that's the intention behind the platform

[issue25148] Windows registry PythonCore key changed inconsistent with other releases

Barry Scott added the comment: To clarify: 3.5 means 3.5 64 bit 3.5-32 means 3.5 32 bit You do not add the -64 as it is the default. However: this change should have been in the "What's New" as it breaks Windows installation code for extensions. --

[issue1401] urllib2 302 POST

RuixiaZhang added the comment: I am using python 2.7,the same ,come to 302 lost cookies, I try add the code : req.headers.pop('content-length') to urllib2.py:536 but, have another error: KeyError: 'content-length' So, I want to ask is there any better method to solver this error? --

[issue18967] Find a less conflict prone approach to Misc/NEWS

Larry Hastings added the comment: Okay. I got tired of the constant Misc/NEWS merge conflicts, so I wrote a tool to fix the problem. It's checked in to Bitbucket here: https://bitbucket.org/larry/mergenews/ There's a readme, which you'll see rendered on that page. I don't know what you

[issue25219] Update doc for Idle command line options.

Changes by Terry J. Reedy : -- title: Update doc for Idle command line arguments. -> Update doc for Idle command line options. ___ Python tracker

[issue25219] Update doc for Idle command line arguments.

Changes by Terry J. Reedy : -- assignee: -> terry.reedy ___ Python tracker ___ ___

[issue25219] Update doc for Idle command line arguments.

Changes by Terry J. Reedy : -- title: Update doc for Idle command line argument. -> Update doc for Idle command line arguments. ___ Python tracker

[issue25219] Update doc for Idle command line argument.

Roundup Robot added the comment: New changeset 80e92eba23e0 by Terry Jan Reedy in branch '2.7': Issue #25219: Update doc for Idle command line options. https://hg.python.org/cpython/rev/80e92eba23e0 New changeset 97f3d7749d3f by Terry Jan Reedy in branch '3.4': Issue #25219: Update doc for Idle

[issue9928] weird oddity with bz2 context manager

Roman Valls added the comment: It still seems to be failing with Python 2.7.10... or am I doing sth wrong? :/ $ ipython Python 2.7.10 |Anaconda 2.2.0 (x86_64)| (default, May 28 2015, 17:04:42) Type "copyright", "credits" or "license" for more information. IPython 4.0.0 -- An enhanced

[issue25154] Drop the pyvenv script

Brett Cannon added the comment: python-dev did not seem to object to the idea. -- assignee: -> brett.cannon ___ Python tracker ___

[issue25220] Enhance and refactor test.regrtest (convert regrtest.py to a package)

R. David Murray added the comment: Victor and I have been discussing this on IRC. We agreed that importing from regrtest and running regrtest as a script needs to keep working. Summary of my disagreements: I'm leery of wholesale changes to regrtest because we have gotten bug reports when

[issue25224] Replace Idle's README.txt with annotated file list

New submission from Terry J. Reedy: idlelib currently has user information that belongs in either the Idle reference idle.rst or possibly a separate Idle howto #17583. This information should be removed and replaced with an annotated list of files in idlelib. The latter would help people

[issue24894] iso-8859-11 missing from codecs table

STINNER Victor added the comment: Issue24894.patch is wrong: latin is an alias to ISO 8859-1, not to ISO 8859-11. >>> codecs.lookup('latin').name 'iso8859-1' iso8859_11 has 3 aliases: thai, iso_8859_11, iso_8859_11_2001. You can mention iso-8859-11 and thai in the doc. --

[issue25222] 3.5.0 regression - Fatal Python error: Cannot recover from stack overflow

Bernie Hackett added the comment: > Well, Python has no perfect protection again stack overflow. It's only best > effect. That's interesting. I thought there was a stronger contract, and it appeared that way in all previous cpython releases back to 2.4. Again, this failure is new with 3.5.0.

[issue25210] Special-case NoneType() in do_richcompare()

R. David Murray added the comment: I was agreeing with you, I think the comment should just be dropped. My point was the alternative is really to change it everywhere, not just here :) -- ___ Python tracker

[issue25220] Enhance and refactor test.regrtest (convert regrtest.py to a package)

Roundup Robot added the comment: New changeset c92d893fd3c8 by Victor Stinner in branch 'default': Issue #25220: Backed out changeset eaf9a99b6bb8 https://hg.python.org/cpython/rev/c92d893fd3c8 -- ___ Python tracker

[issue25220] Enhance and refactor test.regrtest (convert regrtest.py to a package)

STINNER Victor added the comment: I pushed the changeset eaf9a99b6bb8, but R. David Murray asked me to wait for a review before pushing changes. So I reverted my change. I checked if Lib/test/regrtest.py is called directly. The answer is yes: it's called inside Python in various places, but

[issue13253] 2to3 fix_renames renames sys.maxint only in imports

Nikita Klimov added the comment: I'm close to solution, but I need another 1 week. -- ___ Python tracker ___

[issue25222] 3.5.0 regression - Fatal Python error: Cannot recover from stack overflow

STINNER Victor added the comment: I'm not sure that I understood the bug. You are testing a function which enters an unlimited loop. You expect to get a RecursionError, but Python does crash. Well, Python has no perfect protection again stack overflow. It's only best effect. You should

[issue25223] Statically or dynamically linked to the VC++runtime ? or how Python install fails o Vista despite the VC redist packages - api-ms-win-crt-runtime-l1-1-0.dll

New submission from acx01bc: I think you should build Python.exe with the option to be statically linked to the VC++-runtime. This way, the executable is a little bigger, but you ensure nobody will have troubles with some VC-runtime different Dll versions. DETAILS : (I posted it also there

[issue24894] iso-8859-11 missing from codecs table

Prashant Tyagi added the comment: That was mistake in previous patch,so here is new patch for the same. -- Added file: http://bugs.python.org/file40557/issue_24894.patch ___ Python tracker

[issue25223] Statically or dynamically linked to the VC++runtime ? or how Python install fails on Vista despite the VC redist packages - api-ms-win-crt-runtime-l1-1-0.dll

Changes by acx01bc : -- title: Statically or dynamically linked to the VC++runtime ? or how Python install fails o Vista despite the VC redist packages - api-ms-win-crt-runtime-l1-1-0.dll -> Statically or dynamically linked to the VC++runtime ? or how Python

[issue25210] Special-case NoneType() in do_richcompare()

STINNER Victor added the comment: R. David Murray wrote: "There are other places NoneType shows up (eg: attribute error). Why should this case be special? Do we want to fix them all?" I think you missed the last sentence of the initial message :-) We are simply discussing a comment :-D

[issue25205] setattr accepts invalid identifiers

R. David Murray added the comment: I did however make the same mistake without checking the docs or the behavior. But the fact that I didn't look at it doesn't make the current documentation wrong :) What change is it that you think would be beneficial? --

[issue25194] Register of Financial Interests for core contributors

R. David Murray added the comment: Antoine: think that's exactly what Nick has in mind, exposing that kind of information. -- ___ Python tracker ___

[issue18967] Find a less conflict prone approach to Misc/NEWS

Brett Cannon added the comment: Yes, a text box or something along those lines. I assume the selection label question is to specify a section of the change log for something to go into. That's actually not necessary as the proper section is implied by the Components label. And we could allow

[issue25143] 3.5 install fails poorly on Windows XP

Tim Golden added the comment: I'd just bail as early as poss. from the installer. If it's possible to detect Windows versions, stick something like "The last version to support WinXP is 3.4". If that's too tricky, perhaps something "Your system may be unsupported. Please try an earlier

[issue25220] Enhance and refactor test.regrtest (convert regrtest.py to a package)

R. David Murray added the comment: This may be opening a can of worms, but I wonder if what we should really do is re-engineer regrtest from the ground up, keeping the existing regrtest around until we are satisfied with its replacement...and maybe some of the options wouldn't make the

[issue25220] Enhance and refactor test.regrtest (convert regrtest.py to a package)

STINNER Victor added the comment: "This may be opening a can of worms, but I wonder if what we should really do is re-engineer regrtest from the ground up," It's not a full reengineering. My patch takes the current code and split it into smaller files. It's not a new implementation or

[issue25222] 3.5.0 regression - Fatal Python error: Cannot recover from stack overflow

New submission from Bernie Hackett: While running PyMongo's test suite against python 3.5.0 the interpreter inconsistently aborts when we test encoding a recursive data structure like: evil = {} evil['evil'] = evil The test that triggers this was added to test the use of Py_EnterRecursiveCall

[issue25222] 3.5.0 regression - Fatal Python error: Cannot recover from stack overflow

Changes by STINNER Victor : -- nosy: +haypo ___ Python tracker ___ ___

[issue18967] Find a less conflict prone approach to Misc/NEWS

Terry J. Reedy added the comment: Larry: commit messages are for other developers, which news entries are ultimately for users. Some developers want (insist on) the freedom to make the two different, with different details. I am using that for Idle patches since NEWS entries become 'What

[issue25085] Windows x86-64 embeddable zip file contains test directorys

Terry J. Reedy added the comment: Thanks, re-reading I see the new exclusion test for this issue: if name in {'test', 'tests'} and p.parts[-3].lower() == 'lib': -- ___ Python tracker

[issue23329] _ssl cannot be compiled with LibreSSL anymore (on OpenBSD 5.5) because of ALPN

Changes by STINNER Victor : -- resolution: fixed -> status: closed -> open ___ Python tracker ___

[issue25164] Windows default installation path is inconsistent between per-user and system-wide installation

Tim Golden added the comment: I don't feel that strongly, but my preference would be "python35[-whatever]" rather than the version with the spaces & the dots. Both for ease of use and for some kind of continuity with the c:\pythonxy we've had for many years. --

[issue25194] Register of Financial Interests for core contributors

Antoine Pitrou added the comment: What would be the rules for how to fill this up? For example, I work for Continuum, which develops Conda and Anaconda, but I'm not paid to work on CPython (I work on Numba). However, perhaps someone might want to ask for CPython work that aligns with

[issue18967] Find a less conflict prone approach to Misc/NEWS

R. David Murray added the comment: I believe auto-filling it from the commit message (for those who don't put in the effort to treat the two audiences differently :) was something we discussed, but it would be a separate enhancement after the news box is added to the tracker. Which I might

[issue25222] 3.5.0 regression - Fatal Python error: Cannot recover from stack overflow

Changes by A. Jesse Jiryu Davis : -- nosy: +emptysquare ___ Python tracker ___ ___

[issue25220] Enhance and refactor test.regrtest (convert regrtest.py to a package)

Roundup Robot added the comment: New changeset eaf9a99b6bb8 by Victor Stinner in branch 'default': Issue #25220: Create Lib/test/libregrtest/ https://hg.python.org/cpython/rev/eaf9a99b6bb8 -- nosy: +python-dev ___ Python tracker

[issue25225] Idle doc: redo Syntax Colors section

Terry J. Reedy added the comment: Current: 3. Syntax colors The coloring is applied in a background “thread,” so you may occasionally see uncolorized text. To change the color scheme, edit the [Colors] section in config.txt. Python syntax colors: Keywords orange Strings

[issue25225] Idle doc: redo Syntax Colors section

Terry J. Reedy added the comment: Possible alternative, with additions and corrections: 3. Syntax colors The coloring is applied in a background “thread,” so you may occasionally see uncolorized text. To change the color scheme, edit the [Colors] section in config.txt.

[issue25225] Idle doc: redo Syntax Colors section

Terry J. Reedy added the comment: The alternative above matches the format of help.txt, except that the latter has the colors lined up. The excessive addition of whitespace here was an argument against replacing help.txt with help.html (#16893). I consider the proposed condensation to be

[issue18967] Find a less conflict prone approach to Misc/NEWS

Nick Coghlan added the comment: I think Larry's "split news" and "merge news" ideas are still useful as a bridging mechanism to help us get from the status quo to a tracker based solution. It would just mean that, at some point in the future, the commit hook would change to be a post-commit

[issue25164] Windows default installation path is inconsistent between per-user and system-wide installation

Steve Dower added the comment: Good question on the extra "Python" directory - I don't recall exactly why that was there, but I suspect it was for consistency with something - probably {userbase} from sysconfig. Also, thanks for catching the user scheme difference. We shouldn't really be

[issue24657] CGIHTTPServer module discard continuous '/' letters from params given by GET method.

Xiang Zhang added the comment: I think this is a bug. According to the rfcs, "/" is a reserved character in query component and continuous "/" in query component may be invalid and how to deal with it depends on the server. But encoded "/", %2F, acts as data and should be preserved. And

[issue25225] Idle doc: redo Syntax Colors section

New submission from Terry J. Reedy: I plan to make the following changes to the Syntax colors section: * Change name to Text and syntax colors, as not all colors are code-syntax related. * Make it a subsection of the previous Editing and navigation section, which is about editor and shell

[issue25222] 3.5.0 regression - Fatal Python error: Cannot recover from stack overflow

Bernie Hackett added the comment: I used sys.setrecursionlimit(250) - the default appears to be 1000 on all my test machines - and that reduced the occurrence of the abort but didn't completely solve the problem. There must be something more going on here. --

[issue25224] Replace Idle's README.txt with annotated file list

Roundup Robot added the comment: New changeset 5a60fdd9af8c by Terry Jan Reedy in branch '2.7': Issue 25224: Augment Idle doc feature list and no-subprocess section https://hg.python.org/cpython/rev/5a60fdd9af8c New changeset 0f20f3fe7ab4 by Terry Jan Reedy in branch '3.4': Issue 25224: Augment

[issue25223] Statically or dynamically linked to the VC++runtime ? or how Python install fails on Vista despite the VC redist packages - api-ms-win-crt-runtime-l1-1-0.dll

acx01bc added the comment: With dependency walker I see that : - in 3.4.3, Python34.exe and also Python.34.dll have been dynamically linked with msvcr100.dll (platform target : v100 in Visual C++ project properties), - while in 3.5.0 (the one you download on python.org main page) Python35.exe

[issue25223] Statically or dynamically linked to the VC++runtime ? or how Python install fails on Vista despite the VC redist packages - api-ms-win-crt-runtime-l1-1-0.dll

Changes by Mark Lawrence : -- components: +Windows nosy: +paul.moore, steve.dower, tim.golden, zach.ware ___ Python tracker ___

[issue25211] Error message formatting errors in int object unit-test script

Martin Panter added the comment: I will commit your patch to 2.7. Here is a my alternative proposal for 3.4+, dropping the Frm class and using subTest() instead. I kept some error messages where I thought they added clarity, but dropped most because I thought they were redundant with the test

[issue16701] Docs missing the behavior of += (in-place add) for lists.

Martin Panter added the comment: New patch mentioning __index__() -- Added file: http://bugs.python.org/file40560/seq-inplace.v2.patch ___ Python tracker

[issue25164] Windows default installation path is inconsistent between per-user and system-wide installation

eryksun added the comment: Also, why does the per-user install path have a seemingly pointless "Python" base directory? I expected it to install directly into FOLDERID_UserProgramFiles, to be consistent with installing to FOLDERID_ProgramFiles. Also, I doubt anyone cares, but the roaming

[issue25211] Error message formatting errors in int object unit-test script

Roundup Robot added the comment: New changeset e2f1f69d0618 by Martin Panter in branch '2.7': Issue #25211: Fix error message code in test_long; patch from s-wakaba https://hg.python.org/cpython/rev/e2f1f69d0618 -- nosy: +python-dev ___ Python

[issue25222] 3.5.0 regression - Fatal Python error: Cannot recover from stack overflow

Martin Panter added the comment: I saw this error recently on some 3.4 builtbots: http://buildbot.python.org/all/builders/AMD64%20Windows10%203.4/builds/74/steps/test/logs/stdio (failed last 4/5 builds) [306/390] test_json Fatal Python error: Cannot recover from stack overflow. Current

[issue25223] Statically or dynamically linked to the VC++runtime ? or how Python install fails on Vista despite the VC redist packages - api-ms-win-crt-runtime-l1-1-0.dll

Steve Dower added the comment: None of them are statically linked, and we looked into that for 3.5 and it had other issues, so we went fully dynamically linked. This is either an issue specific to your machine, or a Vista issue (we didn't get a lot of testing done on Vista). Installing the

[issue25194] Register of Financial Interests for core contributors

Nick Coghlan added the comment: Right, I'm in a similar situation at Red Hat to the one Antoine's in at Continuum Analytics - I'm not personally part of Red Hat's Python maintenance team (I used to work on internal Engineering tools, and now work on Fedora's overall developer experience), but

[issue22820] RESTART line with no output

Roundup Robot added the comment: New changeset e7bf0727f7df by Terry Jan Reedy in branch '2.7': Issue #22820: Explain need for *print* when running file from Idle editor. https://hg.python.org/cpython/rev/e7bf0727f7df New changeset 824cb25448ab by Terry Jan Reedy in branch '3.4': Issue #22820:

[issue22820] RESTART line with no output

Terry J. Reedy added the comment: For anyone who reads the doc, the added explanation should help. -- assignee: docs@python -> terry.reedy resolution: -> fixed stage: -> resolved status: open -> closed versions: +Python 2.7, Python 3.6 ___ Python