Ryan Smith-Roberts added the comment:
I notified secur...@python.org and waited for the go-ahead (from Guido I think)
before opening this bug. If today is the first that the PSRT is hearing about
this, then the issue is broader than just the bugtracker
Ryan Smith-Roberts added the comment:
The send part of the test doesn't matter, since what's being tested happens
before any reads. The MSG multiplier should be removed completely, since none
of the other tests do that.
Patch attached.
--
Added file:
http://bugs.python.org/file33573
Ryan Smith-Roberts added the comment:
I believe the feature you starred resolves this enhancement issue, in which
case my patches are obsolete. And yes, the 'as' syntax makes a lot more sense.
Here's how I hope it works (the arg parsing wrapper remains unchanged):
foo: int
...
_impl
Ryan Smith-Roberts added the comment:
Yes, that makes more sense. The specific syntax is nearly irrelevant compared
to the feature existing at all. You still haven't indicated whether your
starred feature matches the *result* that I outlined. Does
Ryan Smith-Roberts added the comment:
I saw the fix for this in the commit stream. Closing.
--
resolution: - fixed
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue20252
Changes by Ryan Smith-Roberts r...@lab.net:
--
status: open - closed
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue20252
___
___
Python-bugs-list
Ryan Smith-Roberts added the comment:
Perhaps the test is sending an infeasibly large message. If you remove the
'*2048' does it pass? (I set up a FreeBSD 9.2 amd64 VM but all tests are
passing here).
--
___
Python tracker rep...@bugs.python.org
New submission from Ryan Smith-Roberts:
The docs indicate that you can do:
class foo_converter(CConverter):
...
c_default = 'bar()'
py_default = 'Bar'
except that CConverter.__init__() unconditionally overwrites these values.
Patch attached.
--
components: Build
files
New submission from Ryan Smith-Roberts:
If a custom converter declares both default and converter class attributes, you
get this exception (test file attached):
Exception raised during parsing:
Traceback (most recent call last):
File Tools/clinic/clinic.py, line 1445, in parse
Changes by Ryan Smith-Roberts r...@lab.net:
--
components: +Build
type: - crash
versions: +Python 3.4
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue20300
Changes by Ryan Smith-Roberts r...@lab.net:
--
nosy: +Nikratio
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue20299
___
___
Python-bugs-list
Ryan Smith-Roberts added the comment:
No, I'm happy. It isn't really news anyway, since the patch just brings reality
inline with the docs :).
--
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue20299
Ryan Smith-Roberts added the comment:
I have reviewed this as best I am able. I'll be honest that a lot of clinic.py
makes my eyes cross; I'm used to webdev templating, which is inverted from AC
(flow control inside the template). Not complaining, it's a complicated
subject. I do like
Ryan Smith-Roberts added the comment:
socketmodule has three builtins which use PyLong_AsUnsignedLong on their
arguments and would benefit from these converters, but only if they raise
OverflowError. So I vote for #2.
I don't think it's unreasonable to continue to have locally-defined
New submission from Ryan Smith-Roberts:
Find test file attached. The exception:
Exception raised during parsing:
Traceback (most recent call last):
...
File proj/python/trunk.hg/Tools/clinic/clinic.py, line 1535, in render
self.parse_argument(data.parse_arguments)
File proj/python
Ryan Smith-Roberts added the comment:
The use case is primarily to minimize code churn for the derby, but since
you're the one (heroically) doing the code review it's really your call. I
whipped up a quick patch for this feature, and even if you remove c_name from
__init__ I think it's still
Changes by Ryan Smith-Roberts r...@lab.net:
Added file:
http://bugs.python.org/file33494/argument_clinic_ensure_legal_cleanup.patch
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue20227
Ryan Smith-Roberts added the comment:
Here's sendmsg with only nested bracket optional args. If Rietveld doesn't like
this patch I may cry.
--
Added file:
http://bugs.python.org/file33496/argument_clinic_socketmodule_v4.patch
___
Python tracker rep
Ryan Smith-Roberts added the comment:
After all our discussions I'm closing this with resolution don't do that then.
--
resolution: - wont fix
status: open - closed
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue20232
Ryan Smith-Roberts added the comment:
Forgot to linewrap a paragraph.
--
Added file:
http://bugs.python.org/file33489/argument_clinic_socketmodule_v2.patch
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue20174
Ryan Smith-Roberts added the comment:
I've realized this is basically insoluble without PEP 457, because any other
solution involves changing behavior (None being silently accepted where it
would previously raise an exception). If that's OK, then per-function defensive
programming is probably
Ryan Smith-Roberts added the comment:
sha1 is Vajrasky's module. I have a much more complicated example (and Vajrasky
and I just both posted on python-dev separately about this issue):
sockobj.sendmsg(buffers[, ancdata[, flags[, address]]])
-
sockobj.sendmsg(buffers, ancdata=None, flags=0
Ryan Smith-Roberts added the comment:
No, that still leaves address broken since neither None nor an empty tuple are
acceptable.
--
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue20232
Ryan Smith-Roberts added the comment:
Tweaked the argument list for functions using a NULL default. Kludgy but
doesn't lie to the user.
--
Added file:
http://bugs.python.org/file33490/argument_clinic_socketmodule_v3.patch
___
Python tracker rep
Ryan Smith-Roberts added the comment:
Here's the socketmodule patch. I aggressively imported text from the docs for
the docstrings, along with matching parameter names, given how far the old
docstrings have drifted over time. The Windows-specific code is untested, but
otherwise the tests pass
Ryan Smith-Roberts added the comment:
Even more imperative than my version. Excellent!
--
status: open - closed
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue20202
Changes by Ryan Smith-Roberts r...@lab.net:
--
nosy: +rmsr
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue20227
___
___
Python-bugs-list mailing
New submission from Ryan Smith-Roberts:
recvfrom_into fails to check that the supplied buffer object is big enough for
the requested read and so will happily write off the end.
I will attach patches for 3.4 and 2.7, I'm not familiar with the backporting
procedure to go further but all
Changes by Ryan Smith-Roberts r...@lab.net:
Added file:
http://bugs.python.org/file33453/recvfrom_into_buffer_overflow_2.7.patch
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue20246
Changes by Ryan Smith-Roberts r...@lab.net:
--
keywords: +patch
Added file:
http://bugs.python.org/file33452/recvfrom_into_buffer_overflow_3.4.patch
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue20246
Ryan Smith-Roberts added the comment:
Georg Brandl wrote:
Although now is a good time to ensure sensible argument names (I
usually look at the docs to find the documented ones), so that
switching the function to keyword arg support is basically just a
removal of '/'
I started doing
New submission from Ryan Smith-Roberts:
The docs indicate their arguments are positional-only. Trivial patch attached.
--
assignee: docs@python
components: Documentation
files: docs_socket_recvmsg_args.patch
keywords: patch
messages: 208067
nosy: docs@python, rmsr
priority: normal
Ryan Smith-Roberts added the comment:
Never mind, it's actually recvfrom{,_into}. Sigh.
--
status: open - closed
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue20248
New submission from Ryan Smith-Roberts:
type - types
--
assignee: docs@python
components: Documentation
files: argument_clinic_howto_y-hash.patch
keywords: patch
messages: 208076
nosy: docs@python, larry, rmsr
priority: normal
severity: normal
status: open
title: Argument Clinic howto
Ryan Smith-Roberts added the comment:
Thank you for including the doc change in your other patch. As the grammar was
off, instead of just fixing it I naturally rewrote it. Use this patch or just
remove the extraneous for you as you prefer.
--
keywords: +patch
Added file:
http
Ryan Smith-Roberts added the comment:
Just discovered that the bugtracker mail was all going into my spam filter,
yay. Didn't notice your reply until just now.
The functools patch was just a quickie to get it out of my mental queue, and
I'm still working on socketmodule. This is my first time
Changes by Ryan Smith-Roberts r...@lab.net:
--
type: - crash
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue20231
___
___
Python-bugs-list mailing
New submission from Ryan Smith-Roberts:
A signature of the form
a: object = None
b: object
results in b being uninitialized in the generated C code prior to the
PyArgs_ParseTuple call. If ParseTuple does not set a value for b (as it is an
optional argument), b then contains a garbage pointer
New submission from Ryan Smith-Roberts:
When one specifies a default of NULL, the generated signature doc indicates a
default of None. However, if the user actually supplies None, perhaps as a
placeholder in a positional-only call, C code which only checks for NULL will
fail inappropriately
Ryan Smith-Roberts added the comment:
Accepts only a writable bytearray, when the function needs to mutate-in-place.
The rw_buffer support is for PyPy I guess.
--
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue20201
New submission from Ryan Smith-Roberts:
socket.sendto is apparently even weirder than addch or range: the optional
argument is in the *middle*. Attempting this configuration gets me:
Function sendto has an unsupported group configuration. (Unexpected state 5)
An expected unexpected
Ryan Smith-Roberts added the comment:
Taking a crack at this.
socketmodule.h: just a comment, skipping.
_functools.c: kind-of skipping, because the relevant functions are not normally
directly called (pickle protocol __reduce__ and __setstate__ on partial,
__call__ on cmp wrapper). Will add
New submission from Ryan Smith-Roberts:
The custom converter declaration is wrapped with [clinic input] when it should
be [python input].
Patch attached.
--
files: argument_clinic_docfix.patch
keywords: patch
messages: 207724
nosy: rmsr
priority: normal
severity: normal
status: open
Changes by Ryan Smith-Roberts r...@lab.net:
--
keywords: +patch
Added file: http://bugs.python.org/file33372/argument_clinic_functools.patch
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue20174
Ryan Smith-Roberts added the comment:
I'm such a goofball, how about some bug metadata
--
components: +Build
type: - compile error
versions: +Python 3.4
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue20201
New submission from Ryan Smith-Roberts:
Following the howto's instructions on the new syntax for the 'w*' argument, one
gets an exception:
...
File Tools/clinic/clinic.py, line 1780, in converter_init
elif types == (bytearray_type | rwuffer_type):
NameError: global name 'rwuffer_type
New submission from Ryan Smith-Roberts:
Almost didn't catch that PyBuffer_Release isn't needed. Also we need a
verb/adjective for applying AC. Cure?
zlibmodule has already been cured.
--
assignee: docs@python
components: Documentation
messages: 207728
nosy: docs@python, larry, rmsr
47 matches
Mail list logo
