Stefan Behnel added the comment:
Here's another idea. There could be two implementations of slice, one that
uses Python object indices (as currently) and one that has Py_ssize_t indices
(and properties for the start/stop/step attributes). Similar to what Unicode
strings do, the slice type
Berker Peksag added the comment:
Thanks for the report, but the signature is correct.
From https://docs.python.org/devguide/documenting.html#information-units
The signature should include the parameters, enclosing optional parameters
in brackets.
See also
Stefan Behnel added the comment:
Thanks for the review, Serhiy. There isn't really a reason not to 'normalise'
the Python step object to None when 1 is passed from C code, so I updated the
patch to do that.
Regarding on-demand creation of the Python values, the problem is that C code
that
Serhiy Storchaka added the comment:
There could be two implementations of slice, one that
uses Python object indices (as currently) and one that has Py_ssize_t
indices (and properties for the start/stop/step attributes).
Agree, this idea LGTM. Single boolean flag should be enough to switch
New submission from Stefan Behnel:
The median tends to give a better idea about benchmark results than an average
as it inherently ignores outliers.
--
components: Benchmarks
files: show_median.patch
keywords: patch
messages: 231239
nosy: pitrou, scoder
priority: normal
severity:
Serhiy Storchaka added the comment:
In case of even number of samples the median value is calculated as arithmetic
mean of two middle samples.
med_base = (base_times[len(base_times)//2] +
base_times[(len(base_times)+1)//2]) / 2
--
nosy: +serhiy.storchaka
stage: - patch review
Changes by Serhiy Storchaka storch...@gmail.com:
--
Removed message: http://bugs.python.org/msg231240
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue22881
___
Serhiy Storchaka added the comment:
In case of even number of samples the median value is calculated as arithmetic
mean of two middle samples.
med_base = (base_times[len(base_times)//2] +
base_times[(len(base_times)-1)//2]) / 2
--
___
Python
Serhiy Storchaka added the comment:
Oh, socket.gaierror was not made an alias of OSError, only subclass of it.
Well, we can apply the same patch to all releases. Does it fix tests?
--
___
Python tracker rep...@bugs.python.org
Stefan Behnel added the comment:
Fair enough, patch updated.
--
Added file: http://bugs.python.org/file37207/show_median.patch
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue22881
___
Stefan Behnel added the comment:
I reran the benchmarks with the fast path in _PyEval_SliceIndex() and the
results are not conclusive. There is no clear indication that it improves the
performance.
The problem is that most slices have no step, many slices are open ended on at
least one side
New submission from Ludovic Gasc:
Hi,
To install easily Python 3.4.2 on Linux, I use Pythonz:
http://saghul.github.io/pythonz/
I've discovered that, depends on the packages already installed on Linux, I
don't have the same Python each time after compilation on servers.
Some features could
New submission from Serhiy Storchaka:
There are several references to PyInt in outdated comments in .c and .h files:
$ find * -name '*.[ch]' -exec egrep -n --color -- 'PyInt\b' '{}' +
Include/longobject.h:34:/* It may be useful in the future. I've added it in the
PyInt - PyLong
Serhiy Storchaka added the comment:
Indeed, PySequence_GetSlice() is used only in few performance non-critical
places in the stdlib, PySequence_Set/DelSlice() are not used at all. So looks
as this way doesn't speed up any code.
As for faster_PyEval_SliceIndex.patch see issue17576. May be such
Changes by Serhiy Storchaka storch...@gmail.com:
--
assignee: - haypo
dependencies: +Add a function to know about interpreter shutdown
type: - behavior
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue22599
Serhiy Storchaka added the comment:
I'm +0 for is_finalizing.
There is a typo in the docstring.
--
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue22696
___
New submission from Kevin Orr:
When one uses a file object returned by `FileType.__call__` as a context
manager, `sys.stdin`'s or `sys.stdout`'s `__exit__` will be triggered upon exit
of the context, in turn calling their `close` method.
Perhaps the issue is that `sys.stdin` and `sys.stdout`
Serhiy Storchaka added the comment:
I like the idea. I have not make faultfinding review yet, but one thing is
questionable to me.
Is it necessary to get rid of the END_FINALLY opcode?
Currently Python code
try:
stmt1
finally:
stmt2
is compiled to:
SETUP_FINALLY
Kevin Orr added the comment:
Hmm, not sure why I thought I needed `inline code formatting`. It's all
monospace anyway!
Anyway, I'm thinking that adding this somewhere in argparse.py:
class _WrappedIO(object):
def __init__(self, fileobj):
self._file = fileobj
def __exit__(tp,
R. David Murray added the comment:
To be clear: this was/is the python2 documentation style. In python3 we've
switched to using keyword argument notation in functions that support
specifying optional arguments via keyword (which is many more of the C
functions than in python2).
--
R. David Murray added the comment:
I don't think it is a good idea for us to try to track which packages are
needed and what they are named by the various downstreams. We could document
the optional modules by project name, though. Note also that the devguide does
point to the umbrella
R. David Murray added the comment:
Perhaps instead it could return an io object built from the file descriptor
with closefd=False.
--
nosy: +r.david.murray
versions: +Python 3.4, Python 3.5 -Python 2.7, Python 3.6
___
Python tracker
New submission from Stephen Farris:
The dumbdbm module uses an unchecked call to eval() in the _update method,
which is called in response to a call to dumbdbm.open(), and is used to load
the index from the directory file. This poses a security vulnerability because
it allows an attacker to
Stefan Behnel added the comment:
why you pass through fast path only if -1= Py_SIZE(v) =1?
It's usually enough, it's the fastest fast path, and it doesn't even need
error handling.
Any slicing where the slice calculation time matters is going to be of
fairly limited size, often involving the
Ludovic Gasc added the comment:
I'm not agree with you: I've lost a lot of time to find all packages I need for
that, I've merged information from several blog posts.
I imagine that I'm not alone with this problem.
If you move this information in devguide, most persons don't find this.
Andrew Svetlov added the comment:
For Ubuntu/debian you can just run:
$ sudo apt-get build-dep python3
--
nosy: +asvetlov
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue22882
___
Ludovic Gasc added the comment:
$ sudo apt-get build-dep python3
Good to know, I will test for my next deployment.
The only problem with that, if you use a old Debian with a Python3 that need
less dependencies that actual version.
But you can add a line for that in documentation if we have
Nick Coghlan added the comment:
The Interface classes are actually designed to cover any association of an IP
address with a specific network, not just host interface addresses. We knew it
was a slight misnomer when we chose it, but network and broadcast addresses
weren't considered special
New submission from Robert Collins:
TestProgram leaves defaultTestLoader.errors dirty. This primarily affects tests
but it would be good hygiene to clear it at the end of TestProgram.
--
components: Library (Lib)
messages: 231261
nosy: rbcollins
priority: normal
severity: normal
New submission from José Alberto Goncalves:
Summary:
Python 3.4's venv works fine in Windows, and pip works fine when installing
both pure Python libraries and extension modules. However, when the virtual
environment is under a path with non-ASCII characters, attempting to install a
package
New submission from José Alberto Goncalves:
Summary:
Python 3.4's venv works fine in Windows, and pip works fine when installing
both pure Python libraries and extension modules. However, when the virtual
environment is under a path with non-ASCII characters, attempting to install a
package
Changes by Ezio Melotti ezio.melo...@gmail.com:
--
resolution: - duplicate
stage: - resolved
status: open - closed
superseder: - ensurepip and distutils' build_scripts fails on Windows when
path to Python contains accented characters
___
Python
koobs added the comment:
I don't have the environment to test here. Can you run a custom build on the
buildbots?
--
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue17293
___
R. David Murray added the comment:
The fact that it cost you a lot of time buttresses my point. Any documentation
we write is likely to get stale; it is only the downstream that knows when
these things change and could maintain such a document. And the easy answer is
as Andrew indicated,
Changes by R. David Murray rdmur...@bitdance.com:
--
versions: -Python 3.6
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue22882
___
___
Changes by R. David Murray rdmur...@bitdance.com:
--
nosy: +r.david.murray
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue22885
___
___
Changes by R. David Murray rdmur...@bitdance.com:
--
title: [patch] Linux packages you need to compile Python with all dependencies
- Document Linux packages you need to compile Python with all dependencies
___
Python tracker rep...@bugs.python.org
New submission from Kevin Burke:
It would be nice to be able to set a timeout for DNS lookups in the event of a
DNS server failure.
1. Set your DNS to something like 123.123.123.123 or any other host that is not
listening for DNS queries on port 53.
2. Run requests.get('http://jsonip.com',
eryksun added the comment:
On Windows, shouldn't copy_scripts use UTF-8 instead of os.fsencode (MBCS)? The
Python launcher executes the shebang line on Windows, and it defaults to UTF-8
if a script doesn't have a BOM. See line 1105 in maybe_handle_shebang:
New submission from Serhiy Storchaka:
StringIO.StringIO is pickleable and unpickleable on 2.7 but is not unpickleable
on 3.x.
Python 2.7:
import pickle, StringIO
pickle.dumps(StringIO.StringIO('abc'), 2)
Serhiy Storchaka added the comment:
Here is a patch which fixes multiple reversible mappings.
Mappings for StringIO and cStringIO is removed at all because
cStringIO.StringIO is not pickleable on 2.7 at all and pickled
StringIO.StringIO is not compatible with 3.x (issue22890).
--
Changes by Serhiy Storchaka storch...@gmail.com:
--
stage: needs patch - patch review
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue18473
___
___
42 matches
Mail list logo
