jython-bugs — Private bug notification list. Do not post to this list.

[Jython-bugs] [issue2578] professional FPC supplier in China

[Nobody/Anonymous <re...@bu...>]

New submission from Nobody/Anonymous:

0ef7fcae7b10b6ec70c4b9e4f410cdb3

----------
files: unnamed
messages: 11312
nosy: nobody
severity: normal
status: open
title: professional FPC supplier in China
Added file: http://bugs.jython.org/file1577/unnamed

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2578>
_______________________________________

[Jython-bugs] [issue2577] surprising

[Dietrich S. <re...@bu...>]

New submission from Dietrich Schulten:

Greetings! 

You'll be  surprised about what  I've  found on the web, just  take a  look  http://www.crystalaxis.com/quite.php?5455

ds

----------
files: unnamed
messages: 11304
nosy: dschulten
severity: normal
status: open
title: surprising
Added file: http://bugs.jython.org/file1576/unnamed

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2577>
_______________________________________

[Jython-bugs] [issue2576] pip install -r requirements.txt fails with RE recursion limit

[Peter <re...@bu...>]

New submission from Peter:

Discovered and tested under macOS, then reduced to the following simple test case

(1) Install Jython 2.7.0, e.g. as follows:

$ java -jar jython-installer-2.7.0.jar -s -d ~/jython-2.7.0

(2) Update pip to current release, pip 9.0.1

$ ~/jython-2.7.0/bin/jython -m pip install --upgrade pip
Downloading/unpacking pip from https://pypi.python.org/packages/b6/ac/7015eb97dc749283ffdec1c3a88ddb8ae03b8fad0f0e611408f196358da3/pip-9.0.1-py2.py3-none-any.whl#md5=297dbd16ef53bcef0447d245815f5144
  Downloading pip-9.0.1-py2.py3-none-any.whl (1.3MB): 1.3MB downloaded
Installing collected packages: pip
  Found existing installation: pip 1.6.dev1
    Uninstalling pip:
      Successfully uninstalled pip
Successfully installed pip
Cleaning up...

(3) Attempt to run pip with a requirements file, even a trivial example:

$ echo pip > requirements.txt
$ ~/jython-2.7.0/bin/jython -m pip install -r requirements.txt 
Exception:
Traceback (most recent call last):
  File "/Users/xxx/jython-2.7.0/Lib/site-packages/pip/basecommand.py", line 215, in main
    status = self.run(options, args)
  File "/Users/xxx/jython-2.7.0/Lib/site-packages/pip/commands/install.py", line 310, in run
    self.populate_requirement_set(
  File "/Users/xxx/jython-2.7.0/Lib/site-packages/pip/basecommand.py", line 292, in populate_requirement_set
    for req in parse_requirements(
  File "/Users/xx/jython-2.7.0/Lib/site-packages/pip/req/req_file.py", line 89, in parse_requirements
    for line_number, line in lines_enum:
  File "/Users/xxx/jython-2.7.0/Lib/site-packages/pip/req/req_file.py", line 323, in ignore_comments
    for line_number, line in lines_enum:
  File "/Users/xxx/jython-2.7.0/Lib/site-packages/pip/req/req_file.py", line 298, in join_lines
    if COMMENT_RE.match(line):
RuntimeError: maximum recursion depth exceeded (Java StackOverflowError)


I'm assuming this is a Jython problem given pip 9.0.1 works fine for me under CPython and PyPy.

----------
messages: 11299
nosy: pjac
severity: major
status: open
title: pip install -r requirements.txt fails with RE recursion limit
versions: Jython 2.7

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2576>
_______________________________________

[Jython-bugs] [issue2575] Overloaded int and boolean method incorrectly called

[James M. <re...@bu...>]

New submission from James Mudd:

If you import the attached Java class "Overloaded" which defines overloaded "test" and "testVarargs" methods for both int and boolean. When the methods are called in Jython the int versions are always called even when given a boolean argument e.g.

Jython 2.7.1rc1 (, Apr 2 2017, 17:16:05) 
[Java HotSpot(TM) 64-Bit Server VM (Oracle Corporation)] on java1.8.0_121
Type "help", "copyright", "credits" or "license" for more information.
>>> import Overloaded
>>> o = Overloaded()
>>> o.test(True)
u'int:1'
>>> o.test(False)
u'int:0'
>>> o.test(11)
u'int:11'
>>> o.testVarargs(11, 12, 13)
u'ints...:[11, 12, 13]'
>>> o.testVarargs(True, True)
u'ints...:[1, 1]'
>>> o.testVarargs(True, True, False)
u'ints...:[1, 1, 0]'

There is a test around this area test_joverload but it doesn't seem to cover these cases.

This seems like quite a bad bug wondering if this is in fact expected?

----------
components: Core
files: Overloaded.java
messages: 11291
nosy: jamesmudd
severity: normal
status: open
title: Overloaded int and boolean method incorrectly called
type: behaviour
versions: Jython 2.7
Added file: http://bugs.jython.org/file1574/Overloaded.java

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2575>
_______________________________________

[Jython-bugs] [issue2574] test_tarfile failed during CI build

[James M. <re...@bu...>]

New submission from James Mudd:

In this build job https://travis-ci.org/jythontools/jython/jobs/213994211 test_tarfile failed, this test also seems a bit unreliable maybe it should be investigated?

----------
messages: 11268
nosy: jamesmudd
severity: normal
status: open
title: test_tarfile failed during CI build

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2574>
_______________________________________

[Jython-bugs] [issue2573] test_weakref intermittent test failure on linux

[James M. <re...@bu...>]

New submission from James Mudd:

I have observed test_weakref fail a few times while running ant regrtest. The output is:

     [exec] test test_weakref failed -- Traceback (most recent call last):
     [exec]   File "/home/james/git/jython/dist/Lib/test/test_weakref.py", line 1029, in test_weak_values
     [exec]     self.assertEqual(len(list(dict.iterkeys())), 0,
     [exec] AssertionError: deleting the values did not clear the dictionary

The failure appears to be intermittent, I haven't reproduced it when running the test individually, and I have never observed it on Windows.

----------
components: Core
messages: 11255
nosy: jamesmudd
severity: normal
status: open
title: test_weakref intermittent test failure on linux
type: behaviour
versions: Jython 2.7

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2573>
_______________________________________

[Jython-bugs] [issue2572] Network-related tests failing on build bots

[Jeff A. <re...@bu...>]

New submission from Jeff Allen:

These tests currently fail on the Travis & Circle CI build bots:
  test_select_new
  test_socket
  test_ssl
with the exception of the last two on just one Travis configuration.

I'm going to exclude these from the travis build target: if we have green bots, we'll be able to tell when comething breaks that wasn't broken before, e.g. by a commit,  and so will potential contributors of PRs. I'm assigning myself to do that bit.

But the idea of this ticket is to track fixing these failures, so I'll unassign myself in case someone else knows how.

----------
assignee: jeff.allen
components: Core
keywords: test failure causes
messages: 11247
nosy: jeff.allen
severity: normal
status: open
title: Network-related tests failing on build bots
type: crash
versions: Jython 2.7

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2572>
_______________________________________

[Jython-bugs] [issue2571] Error handling in test_codecencodings_tw fails on Java 8

[Jeff A. <re...@bu...>]

New submission from Jeff Allen:

test_codecencodings_tw fails on Java 8 (not on Java 7) like this:

======================================================================
FAIL: test_errorhandle (__main__.Test_Big5)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "...\dist\Lib\test\test_multibytecodec_support.py", line 56, in test_errorhandle
    self.assertEqual(result, expected,
AssertionError: 'abc\x80\x80\xc1\xc4'.decode('big5', 'replace')=u'abc\ufffd\ufffd\u8b10' != u'abc\ufffd\u8b10'

----------------------------------------------------------------------

The same failure affects some other multibyte codecs the tests for which are currently "expected failures" in regrtest.py, because they also fail in other places.

The cause is a change in the behaviour of the built-in codecs. The behaviour in Python 2.7.13 and Jython 2.7 with Java 7 is:

>>> 'abc\x80\x80\xc1\xc4def'.decode('big5', 'replace')
u'abc\ufffd\u8b10def'
>>> 'abc\x80\xc1\xc4def'.decode('big5', 'replace')
u'abc\ufffd\u6514ef'

It seems to read \x80\x80 as the code 8080, which the Java 7 codec reports as UNMAPPED[2], meaning an unmapped code of source length two.

The codec in Java 8 reads \x80 as MALFORMED[1], meaning that the byte does not belong in the code (Big5 leading bytes are in the reange \x81-\xfe), and the length of the malformed text is one. Then it reports that again for the second \x80.

I think the Java 8 policy is more correct, and not ruled out by the Python documentation. Our tests should perhaps allow it.

----------
messages: 11226
nosy: jeff.allen
severity: normal
status: open
title: Error handling in test_codecencodings_tw fails on Java 8

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2571>
_______________________________________

[Jython-bugs] [issue2570] Wrong shebang set for OS X installation of Jython

[Jim B. <re...@bu...>]

New submission from Jim Baker:

OS X is now failing in certain cases, namely yolk, because of a regression on #1112 For now, this may require patching setuptools for Jython itself, as was done earlier; see #2298

To reproduce (and this should be part of any release testing, as a smoke test):

1. Build an installer (in dist/jython-installer.jar):

$ ant all-jars  # will emit warnings like crazy, mostly around doc tags

2. Run the installer. There are many ways to run this, and they should all be ideally tested. More info on possible installer options:

$ java -jar dist/jython-installer.jar --help

However, the CLI is the easiest way to do any testing - especially since the installer has been very stable, with only pip/setuptools integration being added to it in the last 10 or so years:

$ java -jar dist/jython-installer.jar -s -d ~/jython-2.7.1-test-RC1
Performing silent installation
 10 %
 20 %
 30 %
 40 %
 50 %
 60 %
 70 %
 80 %
Generating start scripts ...
Installing pip and setuptools
 90 %
Collecting setuptools
Collecting pip
Installing collected packages: setuptools, pip
Successfully installed pip-9.0.1 setuptools-28.8.0
 100 %
Congratulations! You successfully installed Jython 2.7.1rc1 to directory /Users/jbaker/jython-2.7.1-test-RC1.

3. OK, let's try it out. First, does pip work?

$ cd ~/jython-2.7.1-test-RC1/
jimbaker:jython-2.7.1-test-RC1 jbaker$ bin/pip install yolk
Collecting yolk
/Users/jbaker/jython-2.7.1-test-RC1/Lib/site-packages/pip/_vendor/requests/packages/urllib3/util/ssl_.py:310: SNIMissingWarning: An HTTPS request has been made, but the SNI (Subject Name Indication) extension to TLS is not available on this platform. This may cause the server to present an incorrect TLS certificate, which can cause validation failures. You can upgrade to a newer version of Python to solve this. For more information, see https://urllib3.readthedocs.io/en/latest/security.html#snimissingwarning.
  warnings.warn(
  Using cached yolk-0.4.3.tar.gz
Requirement already satisfied: setuptools in ./Lib/site-packages (from yolk)
Installing collected packages: yolk
  Running setup.py install for yolk ... done
Successfully installed yolk-0.4.3

This SNI warning is due to the fact I'm running on Java 7; if you use Java 8, it goes away, and SNI is used (as in fact is tested in our regrtest).

4. Does setuptools work, in conjunction with yolk?

The reason I chose yolk is that it creates a executable in the bin directory for Jython:

$ ls -ls bin/yolk
8 -rwxr-xr-x  1 jbaker  staff  397 Mar 13 16:31 bin/yolk

However, right now it's not working on OS X, because yolk is being interpreted as a **shell script**:
$ bin/yolk
bin/yolk: line 3: __requires__: command not found
Version: ImageMagick 6.9.3-6 Q16 x86_64 2016-02-28 http://www.imagemagick.org
Copyright: Copyright (C) 1999-2016 ImageMagick Studio LLC
License: http://www.imagemagick.org/script/license.php
... # most output elided but more of the same as the confusing text below!
By default, 'file' is written in the MIFF image format.  To
specify a particular image format, precede the filename with an image
format name and a colon (i.e. ps:image) or specify the image type as
the filename suffix (i.e. image.ps).  Specify 'file' as '-' for
standard input or output.
import: delegate library support not built-in `' (X11) @ error/import.c/ImportImageCommand/1297.
from: can't read /var/mail/pkg_resources
./yolk: line 9: syntax error near unexpected token `('
./yolk: line 9: `    sys.argv[0] = re.sub(r'(-script\.pyw?|\.exe)?$', '', sys.argv[0])'

This bug is because OS X does not properly support recursive shebangs, as noted in #1112 (and also https://www.in-ulm.de/~mascheck/various/shebang/#interpreter-script). To fix, instead of a shebang like the following:

#!/Users/jbaker/jython-2.7.1-test-RC1/bin/jython

We need this:

#!/usr/bin/env /Users/jbaker/jython-2.7.1-test-RC1/bin/jython

With just this small change, yolk will now work:

$ bin/yolk --list
pip 9.0.1 has no metadata
setuptools 28.8.0 has no metadata
wsgiref         - 0.1.2        - active development (/Users/jbaker/jython-2.7.1-test-RC1/Lib)
yolk            - 0.4.3        - active

----------
messages: 11225
milestone: Jython 2.7.1
nosy: zyasoft
priority: urgent
severity: normal
status: open
title: Wrong shebang set for OS X installation of Jython
versions: Jython 2.7

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2570>
_______________________________________

[Jython-bugs] [issue2569] jython launcher fails on Windows if JAVA_HOME is set

[James M. <re...@bu...>]

New submission from James Mudd:

Trying to execute the jython using the exe or python launch script on Windows fails if JAVA_HOME environment variable is set

Microsoft Windows [Version 10.0.14393]
(c) 2016 Microsoft Corporation. All rights reserved.

C:\Users\user\Desktop\dist\bin>jython.exe
Jython 2.7.1rc1 (, Mar 13 2017, 08:51:49)
[Java HotSpot(TM) 64-Bit Server VM (Oracle Corporation)] on java1.8.0_121
Type "help", "copyright", "credits" or "license" for more information.
>>> 
KeyboardInterrupt -------- Works Ok

C:\Users\user\Desktop\dist\bin>set JAVA_HOME="C:\Program Files\Java\jdk1.8.0_121"

C:\Users\user\Desktop\dist\bin>jython.exe
Traceback (most recent call last):
  File "jython.py", line 444, in <module>
    if __name__ == "__main__":
  File "jython.py", line 435, in main
    print command
  File "subprocess.py", line 168, in call
  File "subprocess.py", line 390, in __init__
  File "subprocess.py", line 640, in _execute_child
WindowsError: [Error 5] Access is denied
Failed to execute script jython --------- Fails

Running as admin doesn't help. I think the problem is when JAVA_HOME is set then cmd used to invoke java is the full path C:\Program Files\Java\jdk1.8.0_121\bin\java and python doesn't seem to have permission to launch this. However when JAVA_HOME is not set the cmd used is just java and it relies on java being on the path, it seems python can launch this?

Any ideas for a fix? Or others give it a try and see if it fails for you?

----------
components: Core
messages: 11223
milestone: Jython 2.7.1
nosy: jamesmudd
severity: normal
status: open
title: jython launcher fails on Windows if JAVA_HOME is set
type: behaviour

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2569>
_______________________________________

[Jython-bugs] [issue2568] test_threading occassional failure

[James M. <re...@bu...>]

New submission from James Mudd:

When running the regression tests on linux I have occasionally noticed failures of test_threading with the output:

     [exec] test test_threading failed -- Traceback (most recent call last):
     [exec]   File "/home/james/git/jython/dist/Lib/test/lock_tests.py", line 413, in test_timeout
     [exec]     self.assertTrue(dt >= 0.2, dt)
     [exec] AssertionError: 0.199999809265

It seems to be a issue of precision.

----------
messages: 11221
milestone: Jython 2.7.1
nosy: jamesmudd
severity: normal
status: open
title: test_threading occassional failure
versions: Jython 2.7

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2568>
_______________________________________

[Jython-bugs] [issue2567] System state lost during JSR-223 initialisation

[Jeff A. <re...@bu...>]

New submission from Jeff Allen:

I discovered this trying to re-enable test_jsr223. The ostensible symptom is that regrtest mysteriously becomes verbose:

> dist\bin\jython -m test.regrtest test_binhex test_jsr223 test_binhex
test_binhex
test_jsr223
C:\Users\Jeff\Documents\Eclipse\jython-trunk\dist\Lib\test\regrtest.py:679: RuntimeWarning: Parent module 'test' not found while handling absolute import
  import shutil
test_binhex
C:\Users\Jeff\Documents\Eclipse\jython-trunk\dist\Lib\test\regrtest.py:578: RuntimeWarning: Parent module 'test' not found while handling absolute import
  from test.junit_xml import Tee, write_direct_test
test_binhex (test.test_binhex.BinHexTestCase) ... ok

----------------------------------------------------------------------
Ran 1 test in 0.029s

OK
All 3 tests OK.

But note also the warnings about having lost the 'test' module. Under pdb one may show that, after the JSR-223 test runs, and at the time of these imports, sys.modules and the module list that the import mechanism uses are different:
(Pdb) l
575         if verbose:
576             capture_stdout = None
577         else:
578             capture_stdout = cStringIO.StringIO()
579
580  ->     from test.junit_xml import Tee, write_direct_test
581         try:
582             save_stdout = sys.stdout
583
584             indirect_test = None
585             if junit_xml_dir:
(Pdb) gss = sys.modules['org'].__dict__['python'].__dict__['core'].Py.getSystemState()
(Pdb) id(sys)
198
(Pdb) id(gss)
202
(Pdb) len(sys.modules.keys()) - len(gss.modules.keys())
139
(Pdb)

When test_jsr233 is reduced to just initialising the engine:
    engine = ScriptEngineManager().getEngineByName("python")
we still get this result. (It's nothing in the test itself.)

----------
components: Core
keywords: test failure causes
messages: 11216
nosy: jeff.allen
severity: normal
status: open
title: System state lost during JSR-223 initialisation
type: behaviour

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2567>
_______________________________________

[Jython-bugs] [issue2566] inspect does not recognize code objects from bytecode files

[Andrea B. <re...@bu...>]

Changes by Andrea Bocci <and...@ce...>:


----------
components: Library
nosy: fwyzard
severity: normal
status: open
title: inspect does not recognize code objects from bytecode files
versions: Jython 2.5, Jython 2.7

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2566>
_______________________________________

[Jython-bugs] [issue2565] test_jy_internal deadlock on Windows

[James M. <re...@bu...>]

New submission from James Mudd:

I have just observed another test deadlock in test_jy_internal this time on Windows. Here is the abridged jstack output.

Found one Java-level deadlock:
=============================
"Finalizer":
  waiting for ownable synchronizer 0x00000000eab642c8, (a java.util.concurrent.locks.ReentrantLock$NonfairSync),
  which is held by "MainThread"
"MainThread":
  waiting to lock monitor 0x000000001e423238 (object 0x00000000eb578e28, a java.lang.Class),
  which is held by "Finalizer"

The full output is attached.

----------
components: Core
files: jstack_output.txt
messages: 11195
nosy: jamesmudd
severity: normal
status: open
title: test_jy_internal deadlock on Windows
versions: Jython 2.7
Added file: http://bugs.jython.org/file1562/jstack_output.txt

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2565>
_______________________________________

[Jython-bugs] [issue2564] test_socket_jy fails on Linux

[Stefan R. <re...@bu...>]

New submission from Stefan Richthofer:

Running test_socket_jy I get

test_connect_ex_workout (__main__.SocketConnectTest)
Verify connect_ex states go through EINPROGRESS?, EALREADY*, EISCONN ... ok
test_connect_ex_workout (__main__.SSLSocketConnectTest)
Verify connect_ex states go through EINPROGRESS?, EALREADY*, EISCONN ... FAIL
test_socket_options_defined (__main__.SocketOptionsTest) ... ok

======================================================================
FAIL: test_connect_ex_workout (__main__.SSLSocketConnectTest)
Verify connect_ex states go through EINPROGRESS?, EALREADY*, EISCONN
----------------------------------------------------------------------
Traceback (most recent call last):
  File "Lib/test/test_socket_jy.py", line 140, in test_connect_ex_workout
    self.assertEqual(result[-1], errno.EISCONN)
AssertionError: 115 != 106

----------------------------------------------------------------------
Ran 3 tests in 10.703s

FAILED (failures=1)


After calling

hg update -r 8047

it passes again. So we somehow broke it with that commit.
Something strange: I vaguely remember that directly after that update, after running regrtestes for commit 8048 test_socket_jy was still passing. Any clue?

----------
messages: 11189
nosy: stefan.richthofer
priority: normal
severity: normal
status: open
title: test_socket_jy fails on Linux
type: behaviour
versions: Jython 2.7

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2564>
_______________________________________

[Jython-bugs] [issue2563] Windows: ValueError: invalid encoding: None

[Stefan R. <re...@bu...>]

New submission from Stefan Richthofer:

On Windows 10 I observe this or rather similar output frequently in test_io, test_io_jy and test_univnewlines:

Traceback (most recent call last):
  File "Lib\test\test_io.py", line 2504, in test_read_nonbytes
    t = self.TextIOWrapper(NonbytesStream('a'))
  File "D:\workspace\linux\Jython\ssh\jython\dist\Lib\_io.py", line 1037, in __init__
    raise ValueError("invalid encoding: %r" % encoding)
ValueError: invalid encoding: None

It always names one of these locations:

_io.py line 1037
_pyio.py line 1489

----------
messages: 11166
nosy: stefan.richthofer
priority: normal
severity: normal
status: open
title: Windows: ValueError: invalid encoding: None
type: behaviour
versions: Jython 2.7

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2563>
_______________________________________

[Jython-bugs] [issue2562] Windows: OSError: unlink(): an unknown error occurred

[Stefan R. <re...@bu...>]

New submission from Stefan Richthofer:

I observe this error frequently on Windows. Usually failure isn't stable; it succeeds sometimes.

OSError: unlink(): an unknown error occurred: C:\Users\<name>\AppData\Local\Temp\tmpxyn53w

I observed it e.g. in
test_chdir, test_csv, test_file_newlines, test_tempfile

----------
messages: 11165
nosy: stefan.richthofer
priority: normal
severity: normal
status: open
title: Windows: OSError: unlink(): an unknown error occurred
type: behaviour
versions: Jython 2.7

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2562>
_______________________________________

[Jython-bugs] [issue2561] win32_ver raises exception

[Stefan R. <re...@bu...>]

New submission from Stefan Richthofer:

Jython 2.7.1rc1 (, Mär 3 2017, 13:27:02)
[Java HotSpot(TM) 64-Bit Server VM (Oracle Corporation)] on java1.8.0_121
Type "help", "copyright", "credits" or "license" for more information.
>>> import platform
>>> platform.win32_ver()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "D:\workspace\linux\Jython\ssh\jython\dist\Lib\platform.py", line 600, in win32_ver
    import _winreg
ImportError: No module named _winreg


This is 'caused' by the fix for #2553. However win32_ver also wasn't workable before that. Only difference was that it didn't raise an exception but returned empty strings instead.
Looking at what CPython says, win32_ver would return:

Python 2.7.13 (v2.7.13:a06454b1afa1, Dec 17 2016, 20:53:40) [MSC v.1500 64 bit (AMD64)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import platform
>>> platform.win32_ver()
('10', '10.0.14393', '', u'Multiprocessor Free')

We can easily emulate the first two values. Only the build-type would be hairy to get. It's only reachable via API command, or rather invonveniently via systeminfo cmd-call (hardly possible to parse, because line-captions are localized). So I' suggest to just fill in the first two entries - better than exception or all strings empty.

In long term we might be able to improve this using JNR or by supporting the win32api extension.

----------
messages: 11164
nosy: stefan.richthofer
severity: normal
status: open
title: win32_ver raises exception

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2561>
_______________________________________

[Jython-bugs] [issue2560] StackOverflowError with multiple inheritance, metaclasses, and __isinstancecheck__

[Ryan <re...@bu...>]

New submission from Ryan:

class A(type): pass

class B(type):
    def __instancecheck__(): pass

class C(A, B): pass

print(isinstance(object(), C))


gives:


Traceback (most recent call last):
  File "tst.py", line 8, in <module>
    print(isinstance(object(), C))
RuntimeError: maximum recursion depth exceeded (Java StackOverflowError)


Originally reported as https://github.com/python/typing/issues/395.

----------
messages: 11148
nosy: refi64
severity: normal
status: open
title: StackOverflowError with multiple inheritance, metaclasses, and __isinstancecheck__
type: behaviour
versions: Jython 2.7

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2560>
_______________________________________

[Jython-bugs] [issue2559] test_marshal fails

[James M. <re...@bu...>]

New submission from James Mudd:

When running the regression tests on Ubuntu 16.04 with Java version 1.8.0_121 the test_marshal test fails. The error is below.

[exec] test test_marshal failed -- multiple errors occurred; run in verbose mode for details

The cause of the failure is a StackOverflowError this can be fixed by setting the ThreadStackSize for the JVM larger. I found 3m to be enough so pass -Xss3m to the JVM and the test passes.

As this is a test specifically for this i'm not clear that increasing the ThreadStackSize is the best solution or how that option should be added to the tests.

Maybe this should be resolved for 2.7.1?

----------
messages: 11144
milestone: Jython 2.7.1
nosy: jamesmudd
severity: normal
status: open
title: test_marshal fails

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2559>
_______________________________________

[Jython-bugs] [issue2558] Automatic string to enum coercion

[James M. <re...@bu...>]

New submission from James Mudd:

This is not a bug, but i wasn't sure where to put improvements? It would be nice if when calling a java method which takes a enum automatic type conversion from string could be done if possible. This would make calling enum methods more convenient.

At the moment a call to a method taking a enum with a string fails, even if the enum valueOf(String) method would work e.g

Jython 2.7.1b3 (, Feb 28 2017, 19:53:14) 
[Java HotSpot(TM) 64-Bit Server VM (Oracle Corporation)] on java1.8.0_121
Type "help", "copyright", "credits" or "license" for more information.
>>> import java.time.LocalDate
>>> date = java.time.LocalDate.of(2015, 'DECEMBER', 3)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: of(): 2nd arg can't be coerced to int, java.time.Month
>>> 

Since Jython already does other similar type coercion I think this would be a nice extension, but maybe i'm missing a reason why its a really bad idea?

----------
components: Core
messages: 11142
nosy: jamesmudd
severity: minor
status: open
title: Automatic string to enum coercion
type: behaviour

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2558>
_______________________________________

[Jython-bugs] [issue2557] ongoing pain with platform detection via os.name and sys.platform

[Stefan R. <re...@bu...>]

New submission from Stefan Richthofer:

I am currently implementing a way that will give us fine-grained control of how Jython displays os.name and sys.platform to certain modules.
Actually to certain classes and even certain methods as well.
Opening this issue to keep track of that development and to collect ideas for tagets.

It is based on an idea from JyNI, which we also discussed in #2521.

However it seems that implementing the PyShadowString idea as mentioned in #2521 would break to much code (I did some experiments).
So I added more control by allowing to add 'targets' to PyShadowString. PyShadowString will then only apply its fake-__eq__ implementation if a target is in the traceback. A target can be any regex-String and can be registered with scope on class or scope on method.

This will get clearer once I can come up with an actual implementation. Stay tuned...

----------
assignee: stefan.richthofer
components: Any
messages: 11123
milestone: Jython 2.7.1
nosy: darjus, jeff.allen, stefan.richthofer, zyasoft
priority: normal
severity: normal
status: open
title: ongoing pain with platform detection via os.name and sys.platform
type: behaviour
versions: Jython 2.7

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2557>
_______________________________________

[Jython-bugs] [issue2556] Github mirror is behind the Mercurial one

[James M. <re...@bu...>]

New submission from James Mudd:

Currently the github version of the repo at https://github.com/jythontools/jython is a few commits and days behind the mercurial one https://hg.python.org/jython/ can this be fixed?

Also whats the longer term plan here is the github repo going to become authoritative?

----------
components: None
messages: 11112
nosy: jamesmudd
severity: normal
status: open
title: Github mirror is behind the Mercurial one
type: behaviour

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2556>
_______________________________________

[Jython-bugs] [issue2555] CVE-2016-5699

[James D. <re...@bu...>]

New submission from James Duffy:

The following vulnerability was identified in Python 2.7

https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2016-5699

CRLF injection vulnerability in the HTTPConnection.putheader function in urllib2 and urllib in CPython (aka Python) before 2.7.10 and 3.x before 3.4.4 allows remote attackers to inject arbitrary HTTP headers via CRLF sequences in a URL.

I see the latest jar of Jython doesn't include the fix for this. Is this going to be patched any time?

Thanks!

----------
components: Library
messages: 11111
nosy: jduffy3
severity: normal
status: open
title: CVE-2016-5699
type: security
versions: Jython 2.7

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2555>
_______________________________________

[Jython-bugs] [issue2554] regrtest failures on Windows

[Stefan R. <re...@bu...>]

New submission from Stefan Richthofer:

Running regrtests on Windows 10 I observed some failures that I don't observe on Linux:

test_calendar, test_codecencodings_tw, test_io, test_io_jy, test_strptime, test_time, test_univnewlines

fail especially on Windows for me, using Java 8 + current Jython trunk version.

----------
messages: 11110
nosy: stefan.richthofer
priority: normal
severity: normal
status: open
title: regrtest failures on Windows
type: behaviour
versions: Jython 2.7

_______________________________________
Jython tracker <re...@bu...>
<http://bugs.jython.org/issue2554>
_______________________________________