Issue1069433
This issue tracker has been migrated to GitHub,
and is currently read-only.
For more information,
see the GitHub FAQs in the Python's Developer Guide.
Created on 2004-11-19 14:33 by pieterb, last changed 2022-04-11 14:56 by admin. This issue is now closed.
Messages (7) | |||
---|---|---|---|
msg23218 - (view) | Author: PieterB (pieterb) | Date: 2004-11-19 14:33 | |
On my WinXP SP 1 system with Python2.4c1.msi installed the testcase test_univnewlines.py fails: Traceback (most recent call last): File "C:\Python24\Lib\test\test_univnewlines.py", line 8, in -toplevel- raise test_support.TestSkipped, \ TestSkipped: This Python does not have universal newline support I assume Python on WinXP should have universal newline support. This also causes test_subprocess.py to fail at line 334. Furthermore my test_descr also fails, but I don't know if it's related: ... Testing methods... Testing special operators... Traceback (most recent call last): File "C:\Python24\Lib\test\test_descr.py", line 4065, in -toplevel- test_main() File "C:\Python24\Lib\test\test_descr.py", line 4008, in test_main specials() File "C:\Python24\Lib\test\test_descr.py", line 1875, in specials print Letter('w') PicklingError: Can't pickle <class '__main__.Letter'>: attribute lookup __main__.Letter failed |
|||
msg23219 - (view) | Author: Scott David Daniels (scott_daniels) * | Date: 2004-11-20 01:01 | |
Logged In: YES user_id=493818 Not replicated on Win2K Service Pack 4. All named tests pass here. |
|||
msg23220 - (view) | Author: Alyssa Coghlan (ncoghlan) * | Date: 2004-11-22 12:19 | |
Logged In: YES user_id=1038590 Win XP Pro, SP 1 here: 240 tests OK, 50 tests skipped (all expected skips) The OP needs to check their path settings and their file associations. If 2.4c1 is installed without registering the extensions, then the regression tests should be run using: c: cd \Python24\Lib\test \Python24\python.exe regrtest.py If the .py file association is to a previous version (e.g. 2.2), then "regrtest.py" will invoke that version. Similarly, if a previous version is on the path, "python regrtest.py" will also invoke the incorrect version. |
|||
msg23221 - (view) | Author: Alyssa Coghlan (ncoghlan) * | Date: 2004-11-22 13:51 | |
Logged In: YES user_id=1038590 The three tests mentioned (test_univnewlines, test_subprocess & test_descr) also all work correctly for me when invoked directly, rather than through regrtest.py. |
|||
msg23222 - (view) | Author: PieterB (pieterb) | Date: 2004-11-22 22:13 | |
Logged In: YES user_id=458461 I ran the tests through the IDLE IDE. All tests succeed when I run regrtest.py from a command prompt or from with Cygwin 1.5.11. (240 tests OK, 50 tests skipped as expected) I don't know if running the tests and/or using universal newlines support within IDLE is/will be supported. Please close this bug if it's inappropriate. For more information Python is not in the PATH and TIX_LIBRARY=C:\Python24\tcl\tix8.1 and TK_LIBRARY=C:\Python24\tcl\tk8.4 From Python shell within IDLE: >>> import sys >>> dir(sys.stdin) ['_RPCProxy__attributes', '_RPCProxy__getattributes', '_RPCProxy__getmethods', '_RPCProxy__methods', '__doc__', '__getattr__', '__init__', '__module__', 'encoding', 'oid', 'sockio'] >>> print sys.stdin <idlelib.rpc.RPCProxy instance at 0x00B20508> From a Python shell run from a command prompt: >>> import sys >>> dir(sys.stdin) ['__class__', '__delattr__', '__doc__', '__getattribute__', '__hash__', '__init__', '__iter__', '__new__', '__reduce__', '__reduce_ex__', '__repr__', '__setattr__', '__str__', 'close', 'closed', 'encoding', 'fileno', 'flush', 'isatty', 'mode', 'name', 'newlines', 'next', 'read', 'readinto', 'readline', 'readlines', 'seek', 'softspace', 'tell', 'truncate', 'write', 'writelines', 'xreadlines'] >>> print sys.stdin <open file '<stdin>', mode 'r' at 0x0095E020> PS: I changed the summary because this problem is IDLE specific. |
|||
msg23223 - (view) | Author: Alyssa Coghlan (ncoghlan) * | Date: 2004-11-23 09:07 | |
Logged In: YES user_id=1038590 Ah, that makes a big difference - running through IDLE I get the same three errors (failures in test_descr and test_subprocess, skip on test_univnewlines). I'd be inclined to keep the bug, but change the category to IDLE and assign it to Kurt (SF user 'kbk'). IDLE's his baby so he can make the call as to whether this is a bug, or unfortunate but expected behaviour. |
|||
msg23224 - (view) | Author: Kurt B. Kaiser (kbk) * | Date: 2006-07-27 02:33 | |
Logged In: YES user_id=149084 I don't expect IDLE to be able to pass all the Python tests. It does pretty well, considering it's really two instances of Python communicating with each other. test_descr: there is a special test which fails which causes IDLE to try to pickle an embedded class, which is illegal, to communicate with its GUI. The test_subprocess passes on 2.5. The universal newline support isn't included in IDLE because as an editor, IDLE preserves the line endings in the original file. Any edits are done using the 'native' line endings for that file. This allows IDLE files on disc to be moved between different platforms. If any failing tests are identified in the future which affect IDLE's functionality or the correct execution of 'real' Python code, please post a bug on the tracker. |
History | |||
---|---|---|---|
Date | User | Action | Args |
2022-04-11 14:56:08 | admin | set | github: 41197 |
2004-11-19 14:33:08 | pieterb | create |