[issue21872] LZMA library sometimes fails to decompress a file
Esa Peuha added the comment:
This code
import _lzma
with open('22h_ticks_bad.bi5', 'rb') as f:
infile = f.read()
for i in range(8191, 8195):
decompressor = _lzma.LZMADecompressor()
first_out = decompressor.decompress(infile[:i])
first_len = len(first_out)
last_out = decompressor.decompress(infile[i:])
last_len = len(last_out)
print(i, first_len, first_len + last_len, decompressor.eof)
prints this
8191 36243 45480 True
8192 36251 45473 False
8193 36253 45475 False
8194 36260 45480 True
It seems to me that this is a subtle bug in liblzma; if the input stream to the
incremental decompressor is broken at the wrong place, the internal state of
the decompressor is corrupted. For this particular file, it happens when the
break occurs after reading 8192 or 8193 bytes, and lzma.py happens to use a
buffer of 8192 bytes. There is nothing wrong with the compressed file, since
lzma.py decompresses it correctly if the buffer size is set to almost any other
value.
--
nosy: +Esa.Peuha
___
Python tracker
<http://bugs.python.org/issue21872>
___
___
Python-bugs-list mailing list
Unsubscribe:
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue21876] os.rename(src, dst) does nothing when src and dst files are hard-linked
Esa Peuha added the comment: This looks like a documentation bug. Functions in module os are usually just thin wrappers around the underlying OS functions, and POSIX states that doing nothing is the correct thing to do here. (It is arguably a bug in early Unix implementations that got mistakenly codified as part of POSIX, and it is certainly inconsistent with the POSIX requirement that the mv command *must* remove the source file in this case, but there is nothing Python can do about that.) -- nosy: +Esa.Peuha ___ Python tracker <http://bugs.python.org/issue21876> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue19325] _osx_support imports many modules
Esa Peuha added the comment: Actually Lib/_osx_support.py directly imports only five modules: os, re and sys at top level, and contextlib and tempfile in _read_output(). These last two aren't really needed at all, and there's no point even trying to avoid importing os and sys, so the only real problem is re which imports lots of other modules. It might be possible to avoid using re (the regular expressions don't look very complicated) but I noticed that some functions are only ever called from distutils, so maybe they should be moved from _osx_support to distutils._osx_support before doing anything else. I'm willing to do that if it's considered a good idea. -- nosy: +Esa.Peuha ___ Python tracker <http://bugs.python.org/issue19325> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue19335] codeop misclassifies incomplete code with 'nonlocal'
Esa Peuha added the comment: > Someone needs to compare _maybe_compile to the equivalent C code used by the > real interpreter. Well, _maybe_compile() just calls the built-in function compile() internally, so I'm not sure what sort of comparison you want... -- nosy: +Esa.Peuha title: codeop misclassifies incomple code with 'nonlocal' -> codeop misclassifies incomplete code with 'nonlocal' ___ Python tracker <http://bugs.python.org/issue19335> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue18650] intermittent test_pydoc failure on 3.4.0a1
Esa Peuha added the comment: The problem seems to be /Users/pyi/Library/Python/3.4/lib/python/site-packages/setuptools-0.9.8-py3.4.egg according to the error page. What is that file? The fact that the second run of the test succeeds implies that it could be a temporary file which is no longer present when the test is run again. -- nosy: +Esa.Peuha ___ Python tracker <http://bugs.python.org/issue18650> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue19275] test_site is failing on AMD64 Snow Leopard
Esa Peuha added the comment: Just by looking at the code, there still seems to be a problem: test_site.py was changed in 0b6052f2a8ee to test that the collections module isn't imported, but due to a bug it (re)tests the re module instead (re_mods needs to be changed to collection_mods). -- ___ Python tracker <http://bugs.python.org/issue19275> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue19275] test_site is failing on AMD64 Snow Leopard
Esa Peuha added the comment: Apparently related to issue 19205 and issue 19209. Looks like _osx_support imports re which imports copyreg. -- nosy: +Esa.Peuha ___ Python tracker <http://bugs.python.org/issue19275> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue19246] GC does not really free up memory in console
Esa Peuha added the comment: > So best guess is that Microsoft's allocators have gotten fatally fragmented, > but I don't know how to confirm/refute that. Let's test this in pure C. Compile and run the attached uglyhack.c on win32; if it reports something significantly less than 100%, it's probably safe to conclude that this has nothing to do with Python. -- nosy: +Esa.Peuha Added file: http://bugs.python.org/file32110/uglyhack.c ___ Python tracker <http://bugs.python.org/issue19246> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue12853] global name 'r' is not defined in upload.py
Esa Peuha added the comment: > so I don't know where the "global name 'r' is not defined" message came from. It came from Python 2.7. There are two separate bugs here, one in 3.x and the other in 2.7: the 3.x bug has to do with bytes/string separation, while the 2.7 bug is that result was changed to r in baf1a482b57d. -- nosy: +Esa.Peuha ___ Python tracker <http://bugs.python.org/issue12853> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue19202] Additions to docs
Esa Peuha added the comment: How would you give a single definition of reduce() that helps people to understand both 2-arg and 3-arg variants? The way it is implemented in C is impossible to duplicate in pure Python; the best you could do is a hack that works unless someone *tries* to break it, but that would just be confusing. -- ___ Python tracker <http://bugs.python.org/issue19202> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue19202] Additions to docs
New submission from Esa Peuha: Here are some additions to documentation of a few functions: all, any: alternative definitions using functools.reduce enumerate: alternative definition using zip and itertools.count sum: equivalent definition using functools.reduce and operator.add functools.reduce: equivalent definitions, use operator.add in example itertools.accumulate: point to functools.reduce itertools.filterfalse: point to filter -- assignee: docs@python components: Documentation files: doc.diff keywords: patch messages: 199281 nosy: Esa.Peuha, docs@python priority: normal severity: normal status: open title: Additions to docs type: enhancement versions: Python 3.4 Added file: http://bugs.python.org/file32013/doc.diff ___ Python tracker <http://bugs.python.org/issue19202> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue19165] Change formatter warning to DeprecationWarning in 3.5
Esa Peuha added the comment: The wanted patch seems to be very simple: change PendingDeprecationWarning to DeprecationWarning on line 24 of Lib/formatter.py but only for Python 3.5 (i.e. no need to do anything until 3.4 is released). On a related note, should formatter be added to PEP 4? -- nosy: +Esa.Peuha ___ Python tracker <http://bugs.python.org/issue19165> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue19182] Socket leak in logging.handlers
New submission from Esa Peuha: Running test_logging produces the following: /home/peuha/python/cpython/Lib/logging/handlers.py:550: ResourceWarning: unclosed self.retryTime = now + self.retryPeriod /home/peuha/python/cpython/Lib/logging/handlers.py:550: ResourceWarning: unclosed self.retryTime = now + self.retryPeriod This turns out to be a bug in SocketHandler.makeSocket; the attachment should fix it. -- components: Library (Lib) files: logging.handlers.diff keywords: patch messages: 199074 nosy: Esa.Peuha priority: normal severity: normal status: open title: Socket leak in logging.handlers type: behavior Added file: http://bugs.python.org/file31973/logging.handlers.diff ___ Python tracker <http://bugs.python.org/issue19182> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
[issue19148] Minor issues with Enum docs
New submission from Esa Peuha: I noticed a couple of issues with the Enum documentation. First, there are some markup bugs which should be fixed by the attached file. Second, I think "If the only change desired is no aliases allowed" is very bad English and should be reworded; I would go with "If you only want to disallow aliases" but maybe other people have better ideas. -- assignee: docs@python components: Documentation files: enum.diff keywords: patch messages: 198869 nosy: Esa.Peuha, docs@python priority: normal severity: normal status: open title: Minor issues with Enum docs type: enhancement versions: Python 3.4 Added file: http://bugs.python.org/file31947/enum.diff ___ Python tracker <http://bugs.python.org/issue19148> ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
