, argc=2, argv=0x7fffffffde38, init=, fini=, rtld_fini=, stack_end=0x7fffffffde28) at ../csu/libc-start.c:291 #18 0x00000000005d5f29 in _start () -------- $ ./python --version Python 3.6.1 $ gdb --args ./python ./segfault.py GNU gdb (Ubuntu 7.11.1-0ubuntu1~16.04) 7.11.1 (gdb) r Starting program: ./python ./segfault.py [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1". dictionary changed size during iteration dictionary changed size during iteration dictionary changed size during iteration dictionary changed size during iteration dictionary changed size during iteration Program received signal SIGSEGV, Segmentation fault. _PyObject_Alloc (ctx=0x0, elsize=28, nelem=1, use_calloc=0) at Objects/obmalloc.c:1258 1258 if ((pool->freeblock = *(block **)bp) != NULL) { (gdb) bt #0 _PyObject_Alloc (ctx=0x0, elsize=28, nelem=1, use_calloc=0) at Objects/obmalloc.c:1258 #1 _PyObject_Malloc (ctx=0x0, nbytes=28) at Objects/obmalloc.c:1437 #2 0x0000000000490732 in _PyLong_New (size=1) at Objects/longobject.c:196 #3 PyLong_FromLong (ival=) at Objects/longobject.c:254 #4 0x0000000000542fda in _PyEval_EvalFrameDefault (f=, throwflag=) at Python/ceval.c:3060 #5 0x000000000053d671 in PyEval_EvalFrameEx (throwflag=0, f=0x7ffff7fa1648) at Python/ceval.c:718 #6 _PyFunction_FastCall (co=, args=, nargs=1, globals=globals at entry=0x7ffff7f452d0) at Python/ceval.c:4880 #7 0x000000000053e521 in fast_function (kwnames=0x0, nargs=, stack=, func=0x7ffff7eae510) at Python/ceval.c:4915 #8 call_function (pp_stack=pp_stack at entry=0x7fffffffd920, oparg=oparg at entry=1, kwnames=kwnames at entry=0x0) at Python/ceval.c:4819 #9 0x0000000000542c17 in _PyEval_EvalFrameDefault (f=, throwflag=) at Python/ceval.c:3284 #10 0x000000000053e015 in PyEval_EvalFrameEx (throwflag=0, f=0x93a368) at Python/ceval.c:718 #11 _PyEval_EvalCodeWithName (_co=_co at entry=0x7ffff7ec26f0, globals=globals at entry=0x7ffff7f5e078, locals=locals at entry=0x7ffff7ec26f0, args=args at entry=0x0, argcount=argcount at entry=0, kwnames=kwnames at entry=0x0, kwargs=0x8, kwcount=0, kwstep=2, defs=0x0, defcount=0, kwdefs=0x0, closure=0x0, name=0x0, qualname=0x0) at Python/ceval.c:4128 #12 0x000000000053ee43 in PyEval_EvalCodeEx (closure=0x0, kwdefs=0x0, defcount=0, defs=0x0, kwcount=0, kws=0x0, argcount=0, args=0x0, locals=locals at entry=0x7ffff7ec26f0, globals=globals at entry=0x7ffff7f5e078, _co=_co at entry=0x7ffff7ec26f0) at Python/ceval.c:4149 #13 PyEval_EvalCode (co=co at entry=0x7ffff7ec26f0, globals=globals at entry=0x7ffff7f452d0, locals=locals at entry=0x7ffff7f452d0) at Python/ceval.c:695 #14 0x000000000042740f in run_mod (arena=0x7ffff7f5e078, flags=0x7fffffffdc00, locals=0x7ffff7f452d0, globals=0x7ffff7f452d0, filename=0x7ffff7e844f8, mod=0x933c28) at Python/pythonrun.c:980 #15 PyRun_FileExFlags (fp=0x93a350, filename_str=, start=, globals=0x7ffff7f452d0, locals=0x7ffff7f452d0, closeit=1, flags=0x7fffffffdc00) at Python/pythonrun.c:933 #16 0x000000000042763c in PyRun_SimpleFileExFlags (fp=0x93a350, filename=, closeit=1, flags=0x7fffffffdc00) at Python/pythonrun.c:396 #17 0x000000000043b975 in run_file (p_cf=0x7fffffffdc00, filename=0x8f82a0 L"./segfault.py", fp=0x93a350) at Modules/main.c:338 #18 Py_Main (argc=argc at entry=2, argv=argv at entry=0x8f7010) at Modules/main.c:809 #19 0x000000000041dc52 in main (argc=2, argv=) at ./Programs/python.c:69 ---------- components: Interpreter Core files: segfault.py messages: 294548 nosy: jimwright priority: normal severity: normal status: open title: Garbage Collector can cause Segfault whilst iterating dictionary items type: crash versions: Python 3.5, Python 3.6 Added file: http://bugs.python.org/file46905/segfault.py _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Fri May 26 08:57:08 2017 From: report at bugs.python.org (Ben Wainwright) Date: Fri, 26 May 2017 12:57:08 +0000 Subject: [New-bugs-announce] [issue30485] Element.findall(path, dict) doesn't insert null namespace Message-ID: <1495803428.74.0.976679303383.issue30485@psf.upfronthosting.co.za> New submission from Ben Wainwright: The findall method for ElementTree.Element handles namespace prefixes by tokenising the path and inserting the full namespace in braces based on entries in a dictionary. Unfortunately, this does not work for a namespace without a prefix, so if you have files containing namespaces with and without prefixes, you still need to manually add the namespace url for the unprefixed path. The function xpath_tokenizer checks to see if tokens contain a colon and only adds in the namespace url in that instance. This could be changed to add the url if their is a colon, or if there is not, and the empty string key is present in the namespaces dictionary. ---------- components: Library (Lib), XML messages: 294549 nosy: ben.wainwright priority: normal severity: normal status: open title: Element.findall(path, dict) doesn't insert null namespace type: behavior versions: Python 3.6 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Fri May 26 12:17:01 2017 From: report at bugs.python.org (Antoine Pitrou) Date: Fri, 26 May 2017 16:17:01 +0000 Subject: [New-bugs-announce] [issue30486] Allow setting cell value Message-ID: <1495815421.45.0.489335851706.issue30486@psf.upfronthosting.co.za> New submission from Antoine Pitrou: There are use cases for setting a cell value. One such use case is for (un)pickling recursive closures (see heroic workaround here: https://github.com/cloudpipe/cloudpickle/pull/90/files#diff-d2a3618afedd4e124c532151eedbae09R74 ). Other use cases may include tinkering around and general education value. There also doesn't seem to be, AFAICS, any counter-indication to being able to do so. It's already possible in C using PyCell_Set(), which is a public API. It just lacks an API in Python land. For example `cell_contents` could become a read/write property... ---------- components: Interpreter Core messages: 294554 nosy: benjamin.peterson, gvanrossum, ncoghlan, pitrou, rhettinger priority: normal severity: normal stage: needs patch status: open title: Allow setting cell value type: enhancement versions: Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Fri May 26 12:49:48 2017 From: report at bugs.python.org (Caleb Hattingh) Date: Fri, 26 May 2017 16:49:48 +0000 Subject: [New-bugs-announce] [issue30487] DOC: automatically create a venv and install Sphinx when running make Message-ID: <1495817388.46.0.640927890572.issue30487@psf.upfronthosting.co.za> New submission from Caleb Hattingh: Under guidance from zware during Pycon sprints, I've changed the Doc/ Makefile to automatically create a virtual environment and install Sphinx, all as part of the `make html` command. ---------- assignee: docs at python components: Documentation messages: 294556 nosy: cjrh, docs at python, willingc, zach.ware priority: normal pull_requests: 1909 severity: normal status: open title: DOC: automatically create a venv and install Sphinx when running make versions: Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Fri May 26 13:42:41 2017 From: report at bugs.python.org (Max) Date: Fri, 26 May 2017 17:42:41 +0000 Subject: [New-bugs-announce] [issue30488] Documentation for subprocess.STDOUT needs clarification Message-ID: <1495820561.33.0.197241379069.issue30488@psf.upfronthosting.co.za> New submission from Max: The documentation states that subprocess.STDOUT is: Special value that can be used as the stderr argument to Popen and indicates that standard error should go into the same handle as standard output. However, when Popen is called with stdout=None, stderr=subprocess.STDOUT, stderr is not redirected to stdout and continues to be sent to stderr. To reproduce the problem: $ python >/dev/null -c 'import subprocess;\ subprocess.call(["ls", "/404"],stderr=subprocess.STDOUT)' and observe the error message appearing on the console (assuming /404 directory does not exist). This was reported on SO 5 years ago: https://stackoverflow.com/questions/11495783/redirect-subprocess-stderr-to-stdout. The SO attributed this to a documentation issue, but arguably it should be considered a bug because there seems to be no reason to make subprocess.STDOUT unusable in this very common use case. ---------- components: Interpreter Core messages: 294560 nosy: max priority: normal severity: normal status: open title: Documentation for subprocess.STDOUT needs clarification type: behavior versions: Python 2.7, Python 3.3, Python 3.4, Python 3.5, Python 3.6, Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Fri May 26 14:27:27 2017 From: report at bugs.python.org (Santiago Castro) Date: Fri, 26 May 2017 18:27:27 +0000 Subject: [New-bugs-announce] [issue30489] Add CmdLineTest to standard library Message-ID: <1495823247.09.0.958014555003.issue30489@psf.upfronthosting.co.za> New submission from Santiago Castro: I see that you use some helpers to test command line utilities (https://hg.python.org/cpython/file/default/Lib/test/test_cmd_line_script.py) that would be useful to Python programmers. Please, consider to add them to the standard library (or somewhere) to reuse them. ---------- messages: 294565 nosy: Santiago Castro priority: normal severity: normal status: open title: Add CmdLineTest to standard library type: enhancement versions: Python 2.7, Python 3.3, Python 3.4, Python 3.5, Python 3.6, Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Fri May 26 16:59:34 2017 From: report at bugs.python.org (pfreixes) Date: Fri, 26 May 2017 20:59:34 +0000 Subject: [New-bugs-announce] [issue30490] Allow pass an exception to the Event.set method Message-ID: <1495832374.38.0.182460539379.issue30490@psf.upfronthosting.co.za> New submission from pfreixes: Having the Event as the way to synchronize 1:N coroutines, the none happy path should be able to be expressed making possible call the `set_exception` for each future related to each waiter. As an example the following code trying to implement a way to avoid the dogpile effect for a DNS cache. If the coro that holds the event fails, the original exception is also broadcasted to the waiters. if key in throttle_dns_events: yield from throttle_dns_events[key].wait() else: throttle_dns_events[key] = Event(loop=loop) try: addrs = yield from \ resolver.resolve(host, port, family=family) cached_hosts.add(key, addrs) throttle_dns_events[key].set() except Exception as e: # any DNS exception, independently of the implementation # is set for the waiters to raise the same exception. throttle_dns_events[key].set(exc=e) raise finally: throttle_dns_events.pop(key) ---------- components: asyncio messages: 294572 nosy: pfreixes, yselivanov priority: normal severity: normal status: open title: Allow pass an exception to the Event.set method versions: Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Sat May 27 02:43:00 2017 From: report at bugs.python.org (Nathaniel Smith) Date: Sat, 27 May 2017 06:43:00 +0000 Subject: [New-bugs-announce] [issue30491] Add a lightweight mechanism for detecting un-awaited coroutine objects Message-ID: <1495867380.86.0.533205071954.issue30491@psf.upfronthosting.co.za> New submission from Nathaniel Smith: A common problem when working with async functions is to attempt to call them but forget the 'await', which eventually leads to a 'Warning: coroutine ... was never awaited' (possibly buried in the middle of a bunch of traceback shrapnel caused by follow-on errors). This can be confusing, lead to spuriously passing tests, and generally isn't a great experience for the dev who makes the mistake. To improve this, I'd like to do things like have a test harness reliably detect when this has happened inside a test case, or have trio's main loop check occasionally to see if this has happened and immediately raise a useful error. (Unfortunately, it *doesn't* work to promote the warning to an error using the usual warnings filter machinery, because the warning is raised inside a __del__ method, which suppresses exception propagation.) In principle this is possible with sys.setcoroutinewrapper, but that adds non-trivial runtime overhead to every async function invocation, which is something we'd like to avoid. (It's OK for a "debug mode", but "modes" are also a poor dev experience: they certainly have their place, but whenever possible it's much nicer to give a proper error in the first place instead of relying on the dev to realize they need to enable debug mode and then remember how to do it.) Therefore I propose that CPython keep a thread-local counter of how many coroutine objects currently exist in a created-but-not-yet-started state -- so corofn.__call__ would increment this counter, and the first call to coroobj.__next__/send/throw would decrement it. And there's some way to access it e.g. via a magic function in the sys module. Then test harnesses can assert that this is zero, trio could occasionally poll this from its run loop and assert that it's zero, etc., with very low overhead. (This is a slight modification of the version I discussed with Yury and others at PyCon last week; the previous request was to keep a count of how many times the "coroutine '...' was never awaited" warning had been emitted. The problem with the original idea is that the warning message doesn't fire until the garbage collector has collected the coroutine object, and that might not happen at a convenient time if we're using PyPy, or if there are cycles, or in a test where the missing 'await' eventually leads to an exception whose traceback pins the coroutine object in memory just long enough for the warning to be detected on the *next* test and confuse everyone. Thanks to Matthias Bussonnier for spurring the new idea, and see discussion here: https://github.com/python-trio/trio/issues/79#issuecomment-304364010) ---------- components: asyncio messages: 294584 nosy: giampaolo.rodola, haypo, ncoghlan, njs, yselivanov priority: normal severity: normal status: open title: Add a lightweight mechanism for detecting un-awaited coroutine objects versions: Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Sat May 27 14:28:10 2017 From: report at bugs.python.org (Gregory P. Smith) Date: Sat, 27 May 2017 18:28:10 +0000 Subject: [New-bugs-announce] [issue30492] 'make clinic' does not work for out of tree builds / clinic.py is not in the devguide Message-ID: <1495909690.05.0.0866994683674.issue30492@psf.upfronthosting.co.za> New submission from Gregory P. Smith: I never do in-tree builds anymore because they make me feel dirty and leave build artifacts cluttering up my source tree. make clinic does not work for out of tree builds. To reproduce: ~$ git clone ...cpython repo... ~$ mkdir build && cd build ~/build$ ../cpython/configure --with-pydebug ~/build$ make clinic ./python -E ./Tools/clinic/clinic.py --make ./python: can't open file './Tools/clinic/clinic.py': [Errno 2] No such file or directory Makefile:545: recipe for target 'clinic' failed make: *** [clinic] Error 2 Per https://docs.python.org/3/howto/clinic.html it looks like I can just manually run clinic.py on the files I have modified. Ideally the build system would take care of that for me. But all I really want is a single command that keeps me up to date. python3 Tools/clinic/clinic.py appears to be that command. Even if it isn't fixed for out of tree builds, use of clinic needs to be in the devguide. ---------- assignee: docs at python components: Argument Clinic, Documentation messages: 294604 nosy: docs at python, gregory.p.smith, larry priority: normal severity: normal stage: needs patch status: open title: 'make clinic' does not work for out of tree builds / clinic.py is not in the devguide type: compile error versions: Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Sat May 27 16:45:48 2017 From: report at bugs.python.org (Lee Cannon) Date: Sat, 27 May 2017 20:45:48 +0000 Subject: [New-bugs-announce] [issue30493] Increase coverage of base64 Message-ID: <1495917948.16.0.130856040345.issue30493@psf.upfronthosting.co.za> New submission from Lee Cannon: Added tests to ensure a85decode correctly handles the 'y' character within the byte like object when foldspaces is true. While attempting to increase coverage I found two lines of code within base64.py that appear to never execute: - line 472 - line 501 Line 472: The only time the TypeError except block (line 467) executes is when _b85dec[c] resolves to None for one of the c (characters) in chunk. Then within the except block the c (characters) within chunk are iterated through until the None is hit then a ValueError is raised (line 470). Making it impossible for line 472 to execute. Line 501: s is assigned a value in the expression `s = input.read(MAXBINSIZE)` (line 494), due to this len(s) is only ever not equal to MAXBINSIZE if the end of input has been reached and nothing was read or the last chunk of data was read from input which has a len less than MAXBINSIZE but non-zero. Therefore the only way to enter into the while loop (line 497) is if there are no more characters left to be read from input, yet all that happens in the while loop is the read function is called on input with a non zero parameter (`MAXBINSIZE-len(s)` must be greater than zero for the while to have executed) then when it returns nothing (it always will) execution breaks out of the while loop without modifiying s. ---------- components: Tests messages: 294611 nosy: leecannon priority: normal severity: normal status: open title: Increase coverage of base64 type: enhancement versions: Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Sat May 27 20:54:59 2017 From: report at bugs.python.org (aokaywe) Date: Sun, 28 May 2017 00:54:59 +0000 Subject: [New-bugs-announce] [issue30494] python.exe crashed when open long path in windows 7 Message-ID: <1495932899.34.0.384023708799.issue30494@psf.upfronthosting.co.za> New submission from aokaywe: 1. os.makedirs(ur'\\?\c:\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????') 2. open(ur'\\?\c:\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\?????\t.txt','w') you can see python was crashed. ---------- components: Windows files: python.png messages: 294618 nosy: aokaywe, paul.moore, steve.dower, tim.golden, zach.ware priority: normal severity: normal status: open title: python.exe crashed when open long path in windows 7 versions: Python 2.7 Added file: http://bugs.python.org/file46907/python.png _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Sat May 27 23:22:20 2017 From: report at bugs.python.org (Terry J. Reedy) Date: Sun, 28 May 2017 03:22:20 +0000 Subject: [New-bugs-announce] [issue30495] IDLE: modernize textview module Message-ID: <1495941740.56.0.621655501188.issue30495@psf.upfronthosting.co.za> New submission from Terry J. Reedy: First get issue #. ---------- assignee: terry.reedy components: IDLE messages: 294622 nosy: csabella, terry.reedy priority: normal severity: normal stage: patch review status: open title: IDLE: modernize textview module type: enhancement versions: Python 3.6, Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Sun May 28 09:28:54 2017 From: report at bugs.python.org (Stefan Seefeld) Date: Sun, 28 May 2017 13:28:54 +0000 Subject: [New-bugs-announce] [issue30496] Incomplete traceback with `exec` Message-ID: <1495978134.83.0.774241176124.issue30496@psf.upfronthosting.co.za> New submission from Stefan Seefeld: The following code is supposed to catch and report errors encountered during the execution of a (python) script: ``` import traceback import sys try: env = {} with open('script') as f: exec(f.read(), env) except: type_, value_, tb = sys.exc_info() print (traceback.print_tb(tb)) ``` However, depending on the nature of the error, the traceback may contain the location of the error *within* the executed `script` file, or it may only report the above `exec(f.read(), env)` line. The attached tarball contains both the above as well as a 'script' that exhibit the problem. Is this a bug or am I missing something ? Are there ways to work around this, i.e. determine the correct (inner) location of the error ? (I'm observing this with both Python 2.7 and Python 3.5) ---------- files: pyerror.tgz messages: 294645 nosy: stefan priority: normal severity: normal status: open title: Incomplete traceback with `exec` type: behavior Added file: http://bugs.python.org/file46908/pyerror.tgz _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Sun May 28 14:44:26 2017 From: report at bugs.python.org (Steven Myint) Date: Sun, 28 May 2017 18:44:26 +0000 Subject: [New-bugs-announce] [issue30497] Line number of docstring in AST Message-ID: <1495997066.31.0.794903121268.issue30497@psf.upfronthosting.co.za> New submission from Steven Myint: Since #29463, it is no longer obvious how to get the line number of a docstring in the AST: import ast x = ast.parse('''\ def foo(): """This is a docstring.""" ''') # In Python 3.6, the docstring and line number would be: print(x.body[0].body[0].value.s) print(x.body[0].body[0].value.lineno) # In Python 3.7, I don't know what the equivalent would be. print(x.body[0].docstring) # Line number? We use this feature in pyflakes (https://github.com/PyCQA/pyflakes/issues/271). ---------- components: Interpreter Core messages: 294654 nosy: myint priority: normal severity: normal status: open title: Line number of docstring in AST versions: Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Sun May 28 15:41:20 2017 From: report at bugs.python.org (Brett Cannon) Date: Sun, 28 May 2017 19:41:20 +0000 Subject: [New-bugs-announce] [issue30498] Run Python's slowest tests in the first 3/4 of tests when using -r Message-ID: <1496000480.86.0.388224961805.issue30498@psf.upfronthosting.co.za> New submission from Brett Cannon: If we guaranteed that the slowest tests in the test suite started early enough in the test run to complete before the final test finished, we could shave off several minutes worth of time in cases where the slowest tests are the hold-up. So perhaps if we randomly inserted the 10 slowest tests in the first 3/4 of the complete test list we could make sure they finish early enough? ---------- components: Tests messages: 294658 nosy: brett.cannon priority: normal severity: normal status: open title: Run Python's slowest tests in the first 3/4 of tests when using -r versions: Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Sun May 28 17:30:18 2017 From: report at bugs.python.org (Julien Palard) Date: Sun, 28 May 2017 21:30:18 +0000 Subject: [New-bugs-announce] [issue30499] Deprecated note in set documentation Message-ID: <1496007018.0.0.712296031489.issue30499@psf.upfronthosting.co.za> New submission from Julien Palard: While translating doc to french, I found a surprising comment in stdtypes.html about sets, stating: > To support searching for an equivalent frozenset, the elem set is temporarily mutated during the search and then restored. During the search, the elem set should not be read or mutated since it does not have a meaningful value. So I searched the equivalent in the code, at least to help me understand it to translate it in a proper way, but found it has been deleted in 51ced7afe72d861b46f069d757f2787f13253d32. This commit is in branches 2.7 and 3.1 to 3.6 (and master). ---------- assignee: docs at python components: Documentation messages: 294662 nosy: docs at python, mdk, rhettinger priority: normal severity: normal status: open title: Deprecated note in set documentation versions: Python 2.7, Python 3.5, Python 3.6 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Mon May 29 00:04:12 2017 From: report at bugs.python.org (Nam Nguyen) Date: Mon, 29 May 2017 04:04:12 +0000 Subject: [New-bugs-announce] [issue30500] urllib connects to a wrong host Message-ID: <1496030652.64.0.90102082916.issue30500@psf.upfronthosting.co.za> New submission from Nam Nguyen: Reported by Orange Tsai: ========== Hi, Python Security Team import urllib from urlparse import urlparse url = 'http://127.0.0.1#@evil.com/' print urlparse(url).netloc # 127.0.0.1 print urllib.urlopen(url).read() # will access evil.com I have tested on the latest version of Python 2.7.13. ========== ---------- components: Library (Lib) messages: 294667 nosy: Nam.Nguyen priority: normal pull_requests: 1933 severity: normal status: open title: urllib connects to a wrong host type: security versions: Python 2.7, Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Mon May 29 02:58:36 2017 From: report at bugs.python.org (Serhiy Storchaka) Date: Mon, 29 May 2017 06:58:36 +0000 Subject: [New-bugs-announce] [issue30501] Produce optimized code for boolean conditions Message-ID: <1496041116.54.0.344455983628.issue30501@psf.upfronthosting.co.za> New submission from Serhiy Storchaka: The peephole optimizer optimizes some boolean conditions. For example in "if not a:" it replaces UNARY_NOT+POP_JUMP_IF_FALSE with POP_JUMP_IF_TRUE, and in "if a and b:" it makes checking the boolean value of a only once. But it is unable to optimize more complex conditions, like "if not a and b:". Proposed patch makes the compiler producing optimized code for conditions. It supports expressions with arbitrary complexity containing the "not" operator, the "and" and "or" binary operators, the "if" ternary operator, and chained comparisons, used as conditions in the ternary operator, in the "if", "while" and "assert" statements, and in generator expressions and comprehensions. It would be possible to remove the part of the peepholer optimizer, but it is needed also for optimizing the "and" and "or" operators in non-boolean context. Doing this optimization in the compiler is possible but too cumbersome, it requires 3 times more code that in the proposed patch. May be I'll find the more general solution in future. ---------- components: Interpreter Core messages: 294674 nosy: benjamin.peterson, brett.cannon, ncoghlan, rhettinger, serhiy.storchaka, yselivanov priority: normal severity: normal stage: patch review status: open title: Produce optimized code for boolean conditions type: performance versions: Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Mon May 29 03:19:13 2017 From: report at bugs.python.org (Christian Heimes) Date: Mon, 29 May 2017 07:19:13 +0000 Subject: [New-bugs-announce] [issue30502] Fix buffer handling of OBJ_obj2txt Message-ID: <1496042353.91.0.558550525758.issue30502@psf.upfronthosting.co.za> New submission from Christian Heimes: Frawser Tweedle from Red Hat's identity management team found an issue in PyCA cryptography's handling of buffers for OpenSSL OBJ_obj2txt(). Cryptography fails to handle long OIDs as used by Active Directory. https://github.com/pyca/cryptography/pull/3612/ https://bugzilla.redhat.com/show_bug.cgi?id=1455755 CPython's ssl module doesn't handle buffer allocation for OBJ_obj2txt() correctly, too. A default buffer size of 255+1 makes the bug less likely to occur, though. We should fix the problem anyway. ---------- assignee: christian.heimes components: SSL messages: 294679 nosy: christian.heimes priority: critical severity: normal status: open title: Fix buffer handling of OBJ_obj2txt type: behavior versions: Python 2.7, Python 3.5, Python 3.6, Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Mon May 29 04:22:55 2017 From: report at bugs.python.org (=?utf-8?b?SMOla29uIEjDpmdsYW5k?=) Date: Mon, 29 May 2017 08:22:55 +0000 Subject: [New-bugs-announce] [issue30503] It should be possible to use a module name with the same name as a package name Message-ID: <1496046175.66.0.187733028461.issue30503@psf.upfronthosting.co.za> New submission from H?kon H?gland: I have the following folder structure: . ??? aaa ??? ??? bbb ??? ??? ??? ccc.py ??? ??? ??? __init__.py ??? ??? bbb.py ??? ??? __init__.py ??? __init__.py ??? t.py ./t.py: import sys sys.path = ['.'] import aaa.bbb print(aaa.bbb.get_name()) ./aaa/bbb.py: def get_name(): return "aaa/bbb" however, when I run the main script: $ python -B t.py Traceback (most recent call last): File "t.py", line 5, in print(aaa.bbb.get_name()) AttributeError: module 'aaa.bbb' has no attribute 'get_name' The reason is that there is also a package with name 'aaa.bbb' (i.e. file "./aaa/bbb/__init__.py") and python will see this package before it sees my module "./aaa/bbb.py" and will never load the module. If this is correct, than this is a bad design in my opinion. I should be possible to use a module with the same name as a package. Thanks for considering this issue, and let me know if I can help improve Python at this point. Note: I asked the question first at stackoverflow.com: https://stackoverflow.com/q/44227763/2173773 ---------- components: Interpreter Core messages: 294682 nosy: hakonhagland priority: normal severity: normal status: open title: It should be possible to use a module name with the same name as a package name type: enhancement _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Mon May 29 04:59:06 2017 From: report at bugs.python.org (Antoine Pitrou) Date: Mon, 29 May 2017 08:59:06 +0000 Subject: [New-bugs-announce] [issue30504] Allow inspecting buffering attribute of IO objects Message-ID: <1496048346.63.0.715963034085.issue30504@psf.upfronthosting.co.za> New submission from Antoine Pitrou: It would be useful to be able to inspect the buffering attribute of buffered and text I/O objects, especially for debugging. I would expect e.g.: >>> sys.stdout.buffering 1 # line-buffered >>> sys.stdout.buffer.buffering 8192 ---------- components: IO messages: 294683 nosy: benjamin.peterson, pitrou, stutzbach priority: normal severity: normal stage: needs patch status: open title: Allow inspecting buffering attribute of IO objects type: enhancement versions: Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Mon May 29 05:16:49 2017 From: report at bugs.python.org (Oren Ben-Kiki) Date: Mon, 29 May 2017 09:16:49 +0000 Subject: [New-bugs-announce] [issue30505] Performance of typing._ProtocolMeta._get_protocol_attrs and isinstance Message-ID: <1496049409.93.0.192071382303.issue30505@psf.upfronthosting.co.za> New submission from Oren Ben-Kiki: In 3.6.0, invocation of isinstance calls typing._ProtocolMeta._get_protocol_attrs. This creates a set of all attributes in all base classes, loops on these attributes to check they exist, and discards the set. It is very slow. My program uses isinstance to allow for flexibility in parameter types in certain key functions. I realize that using isinstance is frowned upon, but it seems to make sense in my case. As a result, >95% of its run-time is inside typing._ProtocolMeta._get_protocol_attrs (!). I have created a simple wrapper around isinstance which caches its result with a Dict[Tuple[type, type], bool]. This solved the performance problem, but introduced a different problem - type checking. I use mypy and type annotations, and my code cleanly type-checks (with the occasional # type: ignore). If I switch to using my own isinstance function, then mypy's type inference no longer treats it as special, so it starts complaining about all uses of values protected by if isinstance(value, SomeType): ... I propose that either the private typing._ProtocolMeta.__subclasscheck__ (which invokes _get_protocol_attrs), or the public isinstance, would be modified to cache their results. I can create a PR for either approach, if this is acceptable. ---------- components: Library (Lib) messages: 294686 nosy: orenbenkiki priority: normal severity: normal status: open title: Performance of typing._ProtocolMeta._get_protocol_attrs and isinstance type: performance versions: Python 3.6 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Mon May 29 07:03:04 2017 From: report at bugs.python.org (Jim Fasarakis-Hilliard) Date: Mon, 29 May 2017 11:03:04 +0000 Subject: [New-bugs-announce] [issue30506] Replace 'list' with 'array' in array.remove and array.index Message-ID: <1496055784.02.0.613673148776.issue30506@psf.upfronthosting.co.za> New submission from Jim Fasarakis-Hilliard: Changes the error message for array.remove and array.index to say 'array.(x): x not in array' instead of 'array.(x): x not in list'. This is a splinter issue of issue13349. ---------- components: Library (Lib) messages: 294689 nosy: Jim Fasarakis-Hilliard priority: normal severity: normal status: open title: Replace 'list' with 'array' in array.remove and array.index type: behavior versions: Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Mon May 29 10:49:29 2017 From: report at bugs.python.org (Jim Fasarakis-Hilliard) Date: Mon, 29 May 2017 14:49:29 +0000 Subject: [New-bugs-announce] [issue30507] Elements reports it is a list on Element.remove Message-ID: <1496069369.64.0.23526837777.issue30507@psf.upfronthosting.co.za> New submission from Jim Fasarakis-Hilliard: Another splinter issue from issue13349. Currently, Element reports it's a list when remove is called on it: from xml.etree.ElementTree import Element Element('').remove(Element('')) ValueError: list.remove(x): x not in list >From what I understand, this was done in order for it to conform with the error reporting performed from the pure python implementation of Element. (side note: These also differ regarding the type of value supplied to .remove, the C implementation only wants instances of Element) The message, imo, is confusing and should be changed to Element.remove(x): x not in Element. ---------- components: Library (Lib) messages: 294697 nosy: Jim Fasarakis-Hilliard priority: normal severity: normal status: open title: Elements reports it is a list on Element.remove type: behavior versions: Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Mon May 29 22:14:06 2017 From: report at bugs.python.org (Miguel Grinberg) Date: Tue, 30 May 2017 02:14:06 +0000 Subject: [New-bugs-announce] [issue30508] "Task exception was never retrieved" reported for a canceled task Message-ID: <1496110446.47.0.115157551922.issue30508@psf.upfronthosting.co.za> New submission from Miguel Grinberg: I am seeing a strange issue that occurs when a task that is awaiting an asyncio.wait_for() is cancelled. I created a simple example that I think demonstrates the issue, even though it isn't exactly how it manifests on my application. When I run the attached script never-retrieved.py I get the following error: Task exception was never retrieved future: exception=ZeroDivisionError('division by zero',)> Traceback (most recent call last): File "/usr/lib/python3.5/asyncio/tasks.py", line 239, in _step result = coro.send(None) File "never-retrieved.py", line 5, in crash a = 1/0 ZeroDivisionError: division by zero You can see in the script that the future was cancelled, but the cancel() call was made after the task finished in a zero division error. I think the cancel() call should update the internal state of the future so that the "exception was never retrieved" error does not appear. My application has a more complex setup that I have been unable to reproduce with a simple example. I have a task that is waiting on asyncio.wait_for(fut, timeout), with fut subsequently waiting on a websocket server's receive function. When the websocket client closes the connection, a bunch of cancellations happen, but this future inside the wait_for call crashes before wait_for gets to call cancel() on it. Even though I need to investigate this crash, the fact is that wait_for did cancel this future, but because it already ended in an error the "never retried" error is reported anyway. ---------- components: asyncio files: never-retrieved.py messages: 294732 nosy: Miguel Grinberg, yselivanov priority: normal severity: normal status: open title: "Task exception was never retrieved" reported for a canceled task type: behavior versions: Python 3.5, Python 3.6 Added file: http://bugs.python.org/file46911/never-retrieved.py _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Tue May 30 01:56:40 2017 From: report at bugs.python.org (Serhiy Storchaka) Date: Tue, 30 May 2017 05:56:40 +0000 Subject: [New-bugs-announce] [issue30509] Optimize calling type slots Message-ID: <1496123800.73.0.877253409287.issue30509@psf.upfronthosting.co.za> New submission from Serhiy Storchaka: In excellent Peter Cawley's article "Why are slots so slow?" [1] analysed causes why `a + b` is slower than `a.__add__(b)` for custom __add__ and provided suggestions for optimizing type slot calls. `a + b` and `a.__add__(b)` execute the same user code, `a + b` should have smaller overhead of bytecode interpreting, but it was 2 times slower than `a.__add__(b)`. In the article `a + b` has been made 16% faster than `a.__add__(b)`. In 3.7 the difference between two ways is smaller, but `a + b` still is 25-30% slower than `a.__add__(b)`. After analyzing the article and comparing it with the current code I have found that virtually all proposed optimization steps already applied in 3.7 by Victor! The difference is only in details. The proposed patch tweaks the code and makes `a + b` only 12% slower than `a.__add__(b)`. There is similar effect for other type slot calls. [1] https://www.corsix.org/content/why-are-slots-so-slow ---------- components: Interpreter Core files: type-slot-calls.diff keywords: patch messages: 294739 nosy: haypo, pitrou, serhiy.storchaka priority: normal severity: normal stage: patch review status: open title: Optimize calling type slots type: performance versions: Python 3.7 Added file: http://bugs.python.org/file46912/type-slot-calls.diff _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Tue May 30 06:35:54 2017 From: report at bugs.python.org (Hassan El Karouni) Date: Tue, 30 May 2017 10:35:54 +0000 Subject: [New-bugs-announce] [issue30510] c_bool type not supported for BigEndianStructure on little-endian machine Message-ID: <1496140554.49.0.81782531481.issue30510@psf.upfronthosting.co.za> New submission from Hassan El Karouni: When trying to create a BigEndianStructure containing a c_bool field (on a windows machine), i get the following traceback: File "/home/elkarouh/anaconda2/lib/python2.7/ctypes/_endian.py", line 24, in _other_endian raise TypeError("This type does not support other endian: %s" % typ) TypeError: This type does not support other endian: I fixed the problem by patching __init__.py under the ctypes directory by adding the following two lines: c_bool.__ctype_le__ = c_bool.__ctype_be__ = c_bool _check_size(c_bool) ---------- components: ctypes messages: 294747 nosy: Hassan El Karouni priority: normal severity: normal status: open title: c_bool type not supported for BigEndianStructure on little-endian machine type: compile error versions: Python 2.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Tue May 30 09:46:18 2017 From: report at bugs.python.org (Alex Gaynor) Date: Tue, 30 May 2017 13:46:18 +0000 Subject: [New-bugs-announce] [issue30511] shutil.make_archive should not need to chdir (alternatively: make shutil.make_archive thread-safe) Message-ID: <1496151978.87.0.20600653197.issue30511@psf.upfronthosting.co.za> New submission from Alex Gaynor: Currently shutil.make_archive uses os.chdir, however there's no need for that. Everything that's done could be equally accomplished with path manipulation: https://github.com/python/cpython/blob/master/Lib/shutil.py#L773-L779 We should switch to using path manipulation in order to make shutil.make_archive thread safe. (Flag: This is probably a good bug for someone with Python skills interested in contributing to CPython!) ---------- components: Library (Lib) messages: 294756 nosy: Alex Gaynor, tarek priority: normal severity: normal status: open title: shutil.make_archive should not need to chdir (alternatively: make shutil.make_archive thread-safe) versions: Python 2.7, Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Tue May 30 10:50:43 2017 From: report at bugs.python.org (Thomas Klausner) Date: Tue, 30 May 2017 14:50:43 +0000 Subject: [New-bugs-announce] [issue30512] CAN Socket support for NetBSD Message-ID: <1496155843.0.0.360939407267.issue30512@psf.upfronthosting.co.za> New submission from Thomas Klausner: NetBSD recently got support for CAN. The existing Linux support for CAN partially detects this, but then fails to build. The attached patch against 3.6.1 (from Manuel Bouyer ) fixes this problem and adds proper NetBSD support for CAN sockets. ---------- components: Build files: can.diff keywords: patch messages: 294764 nosy: wiz priority: normal severity: normal status: open title: CAN Socket support for NetBSD type: compile error versions: Python 3.6 Added file: http://bugs.python.org/file46915/can.diff _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Tue May 30 13:01:19 2017 From: report at bugs.python.org (sam-s) Date: Tue, 30 May 2017 17:01:19 +0000 Subject: [New-bugs-announce] [issue30513] getrusage returns platform-dependent value Message-ID: <1496163679.04.0.847547358971.issue30513@psf.upfronthosting.co.za> New submission from sam-s: `resource.getrusage(resource.RUSAGE_SELF).ru_maxrss` returns the peak memory usage, in *bytes* on BSD (including Mac OS X) and in *kilobytes* on Linux. This means that to get a cross-platform value, the user has to check `sys.platform`, which is fairly inconvenient. It would seem like a good idea to return a platform-independent value (e.g., multiply the Linux return value by 1000 - or is it 1024?!) Please see also * https://bugs.python.org/issue20468 * https://stackoverflow.com/a/7669482/850781 Thank you! ---------- components: Library (Lib) messages: 294768 nosy: sam-s priority: normal severity: normal status: open title: getrusage returns platform-dependent value type: behavior versions: Python 2.7, Python 3.3, Python 3.4, Python 3.5, Python 3.6, Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Tue May 30 13:24:00 2017 From: report at bugs.python.org (Grzegorz Grzywacz) Date: Tue, 30 May 2017 17:24:00 +0000 Subject: [New-bugs-announce] [issue30514] test_poplib replace asyncore Message-ID: <1496165040.07.0.652591907468.issue30514@psf.upfronthosting.co.za> New submission from Grzegorz Grzywacz: sub-issue of issue28533 ---------- components: Tests messages: 294770 nosy: grzgrzgrz3 priority: normal severity: normal status: open title: test_poplib replace asyncore versions: Python 3.6, Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Tue May 30 13:26:28 2017 From: report at bugs.python.org (sam-s) Date: Tue, 30 May 2017 17:26:28 +0000 Subject: [New-bugs-announce] [issue30515] unittest: assertAlmostEqual rounding error Message-ID: <1496165188.27.0.848186644048.issue30515@psf.upfronthosting.co.za> New submission from sam-s: I get this `unittest` failure: ``` Traceback (most recent call last): File "zzz.py", line 348, in test_opposite self.assertAlmostEqual(a, b, places=2) AssertionError: 1.1036640046288428 != 1.0986122886681098 within 2 places ``` This seems incorrect: 1.1036640046288428 rounded off to 2 places is 1.10 and so is 1.0986122886681098. E.g., in Lisp: ``` > (round 1.1036640046288428d0 0.01) 110 ; 0.0036640292157588336d0 > (round 1.0986122886681098d0 0.01) 110 ; -0.001387686744974238d0 ``` ---------- components: Library (Lib) messages: 294772 nosy: sam-s priority: normal severity: normal status: open title: unittest: assertAlmostEqual rounding error versions: Python 2.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Tue May 30 17:22:13 2017 From: report at bugs.python.org (=?utf-8?q?Ren=C3=A9_Hern=C3=A1ndez_Remedios?=) Date: Tue, 30 May 2017 21:22:13 +0000 Subject: [New-bugs-announce] [issue30516] Documentation for datetime substract operation incorrect? Message-ID: <1496179333.45.0.413102703331.issue30516@psf.upfronthosting.co.za> New submission from Ren? Hern?ndez Remedios: In the documentation for the supported arithmetic operations for a datetime object, there is the following note, among other: datetime2 = datetime1 - timedelta Comment: Computes the datetime2 such that datetime2 + timedelta == datetime1. As for addition, the result has the same tzinfo attribute as the input datetime, and no time zone adjustments are done even if the input is aware. This isn?t quite equivalent to datetime1 + (-timedelta), because -timedelta in isolation can overflow in cases where datetime1 - timedelta does not. While reading the source code for __sub__ operation I found in the first few lines: Line 1885: def __sub__(self, other): "Subtract two datetimes, or a datetime and a timedelta." if not isinstance(other, datetime): if isinstance(other, timedelta): return self + -other return NotImplemented Is the documentation in contradiction with the actual implementation? ---------- messages: 294787 nosy: Ren? Hern?ndez Remedios priority: normal severity: normal status: open title: Documentation for datetime substract operation incorrect? versions: Python 3.6 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Tue May 30 22:08:08 2017 From: report at bugs.python.org (Max) Date: Wed, 31 May 2017 02:08:08 +0000 Subject: [New-bugs-announce] [issue30517] Enum does not recognize enum.auto as unique values Message-ID: <1496196488.98.0.634063519366.issue30517@psf.upfronthosting.co.za> New submission from Max: This probably shouldn't happen: import enum class E(enum.Enum): A = enum.auto B = enum.auto x = E.B.value print(x) # print(E(x)) # E.A The first print() is kinda ok, I don't really care about which value was used by the implementation. But the second print() seems surprising. By the same token, this probably shouldn't raise an exception (it does now): import enum @enum.unique class E(enum.Enum): A = enum.auto B = enum.auto C = object() and `dir(E)` shouldn't skip `B` in its output (it does now). ---------- components: Library (Lib) messages: 294804 nosy: max priority: normal severity: normal status: open title: Enum does not recognize enum.auto as unique values type: behavior versions: Python 3.6 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Tue May 30 22:54:11 2017 From: report at bugs.python.org (John Jackson) Date: Wed, 31 May 2017 02:54:11 +0000 Subject: [New-bugs-announce] [issue30518] Import type aliases from another module Message-ID: <1496199251.51.0.430123628937.issue30518@psf.upfronthosting.co.za> New submission from John Jackson: I have a 'base' module where I define some type aliases, such as: from typing import List, Tuple Block = [int, Tuple[int]] Blocks = List[Block] Tags = List[str] I would like to import these aliases into other modules so that the 'base' module provides the definitive alias. Currently, if I attempt to import the type aliases using: from base_module import Tags, Blocks Pycharm shows no error, but when I attempt to execute the code I get the error: ImportError: cannot import name 'Tags' I see that there has been some discussion related to this in 2015, but I can't find any documentation saying that something like this has been implemented. ---------- components: Build messages: 294807 nosy: Paragape priority: normal severity: normal status: open title: Import type aliases from another module type: enhancement versions: Python 3.5 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Tue May 30 22:54:13 2017 From: report at bugs.python.org (Andre Wolokita) Date: Wed, 31 May 2017 02:54:13 +0000 Subject: [New-bugs-announce] [issue30519] Add daemon argument to Timer Message-ID: <1496199253.36.0.456621970472.issue30519@psf.upfronthosting.co.za> New submission from Andre Wolokita: Currently in order to have daemonic Timer objects one must instantiate the class, set daemonic status through the property (Timer.daemon=True), and then start the Timer. It would be nice to have the ability to set the daemonic status of the Timer class during instantiation, similar to what is possible with the Thread superclass. This is a trivial enhancement to implement: simply add the daemon keyword argument to the Timer constructor, defaulted to None, and pass it on to the Thread constructor in the call to super().__init__. ---------- messages: 294808 nosy: awolokita priority: normal pull_requests: 1958 severity: normal status: open title: Add daemon argument to Timer type: enhancement versions: Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Wed May 31 03:50:48 2017 From: report at bugs.python.org (Antoine Pitrou) Date: Wed, 31 May 2017 07:50:48 +0000 Subject: [New-bugs-announce] [issue30520] loggers can't be pickled Message-ID: <1496217048.71.0.370871054359.issue30520@psf.upfronthosting.co.za> New submission from Antoine Pitrou: Loggers could simply be pickled and unpickled by name, but pickle currently tries the hard way: >>> import pickle >>> import logging >>> log = logging.getLogger('foo') >>> pickle.dumps(log) Traceback (most recent call last): File "", line 1, in pickle.dumps(log) TypeError: can't pickle _thread.RLock objects ---------- components: Library (Lib) messages: 294818 nosy: pitrou, vinay.sajip priority: normal severity: normal status: open title: loggers can't be pickled type: enhancement versions: Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Wed May 31 04:02:16 2017 From: report at bugs.python.org (Louie Lu) Date: Wed, 31 May 2017 08:02:16 +0000 Subject: [New-bugs-announce] [issue30521] IDLE: Add navigate bar and replace current goto dialog Message-ID: <1496217736.64.0.123972960126.issue30521@psf.upfronthosting.co.za> New submission from Louie Lu: Propose a sublime-like navigate bar to replace current goto dialog. Navigate bar current support two different mode: prefix ':' for goto line, and prefix '@' for goto symbol. When user use shortcut Ctrl+G to open goto line, it will prompt out a navigate bar on the top of the editor.text with prefix ':', user can enter valid lineno, and it will highlight the line on the editor. After return, it will close the navigate bar and clear the highlight, user cursor will be set to the target lineno. ---------- assignee: terry.reedy components: IDLE messages: 294819 nosy: louielu, terry.reedy priority: normal severity: normal status: open title: IDLE: Add navigate bar and replace current goto dialog type: enhancement versions: Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Wed May 31 04:09:10 2017 From: report at bugs.python.org (Antoine Pitrou) Date: Wed, 31 May 2017 08:09:10 +0000 Subject: [New-bugs-announce] [issue30522] Allow replacing a StreamHandler's stream Message-ID: <1496218150.64.0.987173470763.issue30522@psf.upfronthosting.co.za> New submission from Antoine Pitrou: Right now it is probably possible using something like: handler.acquire() try: handler.flush() handler.stream = new_stream finally: handler.release() but it would be nice to have an officially-supported method to do that. Context: if I reconfigure sys.stderr, I'd like to update all handlers that have a reference to the old sys.stderr. ---------- messages: 294820 nosy: pitrou, vinay.sajip priority: normal severity: normal status: open title: Allow replacing a StreamHandler's stream type: enhancement versions: Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Wed May 31 05:22:37 2017 From: report at bugs.python.org (STINNER Victor) Date: Wed, 31 May 2017 09:22:37 +0000 Subject: [New-bugs-announce] [issue30523] unittest: add --list-tests option to only display the list of test names, don't run tests Message-ID: <1496222557.85.0.438733952805.issue30523@psf.upfronthosting.co.za> New submission from STINNER Victor: Attached pull requests adds a --list-tests option to only display the list of test names, don't run tests. It's a first step to implement the issue #29512: "regrtest refleak: implement bisection feature". ---------- components: Tests messages: 294825 nosy: ezio.melotti, haypo, michael.foord, rbcollins priority: normal severity: normal status: open title: unittest: add --list-tests option to only display the list of test names, don't run tests type: enhancement versions: Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Wed May 31 08:01:15 2017 From: report at bugs.python.org (Martijn Pieters) Date: Wed, 31 May 2017 12:01:15 +0000 Subject: [New-bugs-announce] [issue30524] iter(classmethod, sentinel) broken for Argument Clinic class methods? Message-ID: <1496232075.99.0.36248414101.issue30524@psf.upfronthosting.co.za> New submission from Martijn Pieters: I'm not sure where exactly the error lies, but issue 27128 broke iter() for Argument Clinic class methods. The following works in Python 3.5, but not in Python 3.6: from datetime import datetime from asyncio import Task next(iter(datetime.now, None)) next(iter(Task.all_tasks, None)) In 3.6 StopIteration is raised: >>> next(iter(datetime.now, None)) Traceback (most recent call last): File "", line 1, in StopIteration >>> next(iter(Task.all_tasks, None)) Traceback (most recent call last): File "", line 1, in StopIteration (In 3.5 a `datetime.datetime` and `set` object are produced, respectively) The only thing these two methods have in common is that they are class methods with no arguments, parsed out by the Argument Clinic generated code (so using _PyArg_Parser). What appears to have changed is that iter() was switched from using PyObject_Call to _PyObject_FastCall, see https://github.com/python/cpython/commit/99ee9c70a73ec2f3db68785821a9f2867c3f637f ---------- messages: 294835 nosy: mjpieters priority: normal severity: normal status: open title: iter(classmethod, sentinel) broken for Argument Clinic class methods? versions: Python 3.6, Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Wed May 31 08:21:52 2017 From: report at bugs.python.org (Alex Gaynor) Date: Wed, 31 May 2017 12:21:52 +0000 Subject: [New-bugs-announce] [issue30525] Expose SCTs on TLS connections Message-ID: <1496233312.72.0.638169189351.issue30525@psf.upfronthosting.co.za> New submission from Alex Gaynor: CT (https://www.certificate-transparency.org/) is starting to become a thing! It'd be great if we exposed SCTs (whether from TLS extensions, OCSP, or embedded in the certificate) for TLS connections. This would allow higher level protocols to begin acting on them. This would need to be an OpenSSL 1.1.0+ feature. ---------- assignee: christian.heimes components: SSL messages: 294838 nosy: alex, christian.heimes, dstufft, janssen priority: normal severity: normal status: open title: Expose SCTs on TLS connections versions: Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Wed May 31 12:15:46 2017 From: report at bugs.python.org (Antoine Pitrou) Date: Wed, 31 May 2017 16:15:46 +0000 Subject: [New-bugs-announce] [issue30526] Allow setting line_buffering on existing TextIOWrapper Message-ID: <1496247346.69.0.574083557121.issue30526@psf.upfronthosting.co.za> New submission from Antoine Pitrou: Currently, if you want to change the line buffering behaviour of standard streams, you can create a new stream and assign it to sys.{stdout,stderr,stdin}. Unfortunately, it is common for references to the old streams to be stored in various places (such as logging configuration, or third-party libraries). Replacing them all is probably a hopeless cause. It would be much better if one could simply write `sys.stdout.line_buffering = True` and be done with it. ---------- components: IO messages: 294850 nosy: benjamin.peterson, pitrou, stutzbach priority: normal severity: normal status: open title: Allow setting line_buffering on existing TextIOWrapper type: enhancement versions: Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Wed May 31 12:35:05 2017 From: report at bugs.python.org (Christoph Reiter) Date: Wed, 31 May 2017 16:35:05 +0000 Subject: [New-bugs-announce] [issue30527] PyMemoryView_FromBuffer memory leak Message-ID: <1496248505.52.0.454625386469.issue30527@psf.upfronthosting.co.za> New submission from Christoph Reiter: I'm using the following code PyObject * get_memoryview (PyObject *self) { Py_buffer view; ... // this takes a ref on self if (PyBuffer_FillInfo (&view, self, buffer, length, 0, 0) < 0) return NULL; // this returns a object return PyMemoryView_FromBuffer (&view); } The problem is that when I call release() on the returned memory object the buffer does not get release and as a result the exporter leaks. Am I missing something or is this a bug? ---------- messages: 294851 nosy: lazka priority: normal severity: normal status: open title: PyMemoryView_FromBuffer memory leak _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Wed May 31 13:15:31 2017 From: report at bugs.python.org (Hristo Venev) Date: Wed, 31 May 2017 17:15:31 +0000 Subject: [New-bugs-announce] [issue30528] ipaddress.IPv{4, 6}Network.reverse_pointer is broken Message-ID: <1496250931.44.0.236459920081.issue30528@psf.upfronthosting.co.za> New submission from Hristo Venev: `ipaddress.IPv4Network('127.0.0.0/16').reverse_pointer = '0/16.0.0.127.in-addr.arpa'` is definitely wrong. I think it should be '0.127.in-addr.arpa'. A funnier case, `ipaddress.IPv6Network('2001:db8::/32').reverse_pointer = '2.3./.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.8.b.d.0.1.0.0.2.ip6.arpa'`. For the case where no single reverse pointer exists (e.g. `127.0.0.0/13`), I think it should be `None`. ---------- components: Library (Lib) messages: 294854 nosy: h.venev priority: normal severity: normal status: open title: ipaddress.IPv{4,6}Network.reverse_pointer is broken type: behavior versions: Python 3.5 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Wed May 31 16:24:52 2017 From: report at bugs.python.org (Serhiy Storchaka) Date: Wed, 31 May 2017 20:24:52 +0000 Subject: [New-bugs-announce] [issue30529] Incorrect error messages for invalid whitespaces in f-string subexpressions Message-ID: <1496262292.77.0.0898281487446.issue30529@psf.upfronthosting.co.za> New submission from Serhiy Storchaka: Python parser supports only ' ', '\t', '\x0c' and '\r' as whitespaces. The parser of f-strings raises incorrect error messages for subexpressions consistent only from whitespaces, if they contain whitespaces out this set. >>> eval("\xa0") Traceback (most recent call last): File "", line 1, in File "", line 1 ? ^ SyntaxError: invalid character in identifier >>> eval("f'''{\xa0}'''") Traceback (most recent call last): File "", line 1, in File "", line 1 SyntaxError: f-string: empty expression not allowed ---------- components: Interpreter Core messages: 294869 nosy: eric.smith, serhiy.storchaka priority: normal severity: normal status: open title: Incorrect error messages for invalid whitespaces in f-string subexpressions type: behavior versions: Python 3.6, Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Wed May 31 17:11:25 2017 From: report at bugs.python.org (Mariano Anaya) Date: Wed, 31 May 2017 21:11:25 +0000 Subject: [New-bugs-announce] [issue30530] Descriptors HowTo: Example on function.__get__ needs update Message-ID: <1496265085.29.0.159074921199.issue30530@psf.upfronthosting.co.za> New submission from Mariano Anaya: The example on which tries to create a method from the function's __get__ doesn't work for Python 3.5+. It would give this error: TypeError: method expected 2 arguments, got 3 The 3rd parameter, needs to be removed. ---------- assignee: docs at python components: Documentation messages: 294871 nosy: Mariano Anaya, docs at python priority: normal pull_requests: 1967 severity: normal status: open title: Descriptors HowTo: Example on function.__get__ needs update versions: Python 3.5, Python 3.6, Python 3.7 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Wed May 31 17:29:51 2017 From: report at bugs.python.org (Thomas Shinnick) Date: Wed, 31 May 2017 21:29:51 +0000 Subject: [New-bugs-announce] [issue30531] Windows 10 3.6.1 install disallows pip installs of packages Message-ID: <1496266191.79.0.841695217582.issue30531@psf.upfronthosting.co.za> New submission from Thomas Shinnick: Installed using python-3.6.1-amd64.exe then tried to pip install a package. Repeatedly had to manually intervene to open up permissions on python install subdirs, after getting these errors one at a time from pip: PermissionError: [WinError 5] Access is denied: 'c:\\program files\\python 3.6\\Lib\\site-packages\\fontTools' PermissionError: [Errno 13] Permission denied: 'c:\\program files\\python 3.6\\Scripts\\pylupdate5.exe' PermissionError: [WinError 5] Access is denied: 'c:\\program files\\python 3.6\\share' Gave up and set "full control" on \python 3.6\ dir and contents. It seems permissions default to "read-only" subdirs which makes pip installs impossible. Install and pip used under same userid and same session. I thought I'd clicked yes to any question "install for all users". ---------- components: Installation messages: 294872 nosy: Thomas Shinnick priority: normal severity: normal status: open title: Windows 10 3.6.1 install disallows pip installs of packages type: behavior versions: Python 3.6 _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Wed May 31 18:08:53 2017 From: report at bugs.python.org (Chris Bradley) Date: Wed, 31 May 2017 22:08:53 +0000 Subject: [New-bugs-announce] [issue30532] email.policy.SMTP.fold() mangles long headers Message-ID: <1496268533.13.0.885025594185.issue30532@psf.upfronthosting.co.za> New submission from Chris Bradley: When email.policy.SMTP.fold() attempts to wrap a line in which a chunk that does not contain a space is longer than the maximum line length, two things go wrong: - The second line does not begin with a space, creating an invalid header. - The space before the long chunk is stripped, causing the line break to occur in an invalid place. In the attached test case, email.policy.SMTP.fold() is called on a line of the style: Content-Disposition: attachment; filename="" does not contain any spaces and increases in length with each pass. Everything works correctly when 'filename=""' fits onto one line. However, once it is longer than the maximum line length, the following happens: - the space after the semicolon is dropped - the line break often splits up the 'filename' keyword - the second line loses the initial whitespace ---------- files: test_long_mime_headers.py messages: 294875 nosy: chrisb at emergence.com priority: normal severity: normal status: open title: email.policy.SMTP.fold() mangles long headers versions: Python 3.5 Added file: http://bugs.python.org/file46917/test_long_mime_headers.py _______________________________________ Python tracker _______________________________________ From report at bugs.python.org Wed May 31 18:52:48 2017 From: report at bugs.python.org (Carl Meyer) Date: Wed, 31 May 2017 22:52:48 +0000 Subject: [New-bugs-announce] [issue30533] missing feature in inspect module: getmembers_static Message-ID: <1496271168.24.0.639482608296.issue30533@psf.upfronthosting.co.za> New submission from Carl Meyer: The inspect module contains a getattr_static() function, for accessing an arbitrary attribute on a Python object without risking descriptor or __getattr__ code execution. This is useful for introspection tools that don't want to trigger any side effects. The inspect module also contains a getmembers() function, which returns a mapping of names to values for all the object's members. This function could also be very useful to introspection tools, except that internally it uses normal getattr, thus reintroduces the risk of arbitrary code execution. It would be useful to have an equivalent to getmembers() that is descriptor-safe. This could be done either by introducing a new getmembers_static(), or possibly by adding a `getattr` optional keyword argument to getmembers, that would take a getattr-equivalent callable to use in fetching attributes from the object. (The latter option might render some internal assumptions of getmembers() incorrect, needs experimentation.) ---------- components: Library (Lib) messages: 294876 nosy: carljm priority: normal severity: normal status: open title: missing feature in inspect module: getmembers_static type: enhancement versions: Python 3.7 _______________________________________ Python tracker _______________________________________