Nick Coghlan added the comment:
I'm in the process of updating the LBYL support to use a
"rmtree.avoids_symlink_attacks" function attribute rather than the
"rmtree_is_safe" module level attribute.
As I said in the hmac.secure_compare function discussion, the words
Nick Coghlan added the comment:
Éric - there's almost certainly going to be a PEP for 3.4 about doing this kind
of feature advertisement in a cleaner and more consistent way.
--
___
Python tracker
<http://bugs.python.org/i
Nick Coghlan added the comment:
-1. The entire Python API is already accessible from C, so expansions to the
public C API need more justification than "it's exposed in Python".
If new instances of this type end up being used in performance critical third
party C extensions, t
Nick Coghlan added the comment:
I created #14945 as a suggestion to add a simple page to
http://docs.python.org/dev/using/index.html that will provide a central
reference to the module documentation for modules with officially supported
behaviour when used with the "-m&qu
Nick Coghlan added the comment:
Found it: #11260. I've left it open, since the original suggestion in that
issue is related to actually documenting the -m behaviour of the smptd module -
it was just the issue *discussion* that ended up covering the more general
question of how such co
Nick Coghlan added the comment:
Same goes for idle and 2to3. These may just be cross-references to the relevant
module documentation rather than completely new text.
--
title: Setup & Usage documentation for pydoc -> Setup & Usage documentation for
pydoc,
Nick Coghlan added the comment:
Additional candidates after grepping the docs:
python -m site
python -m sysconfig
python -m pickle
python -m pickletools
python -m compileall
python -m test
--
___
Python tracker
<http://bugs.python.org/issue14
Nick Coghlan added the comment:
I realised that with the addition of types.new_class(), that's fairly easy to
invoke from C (no harder than any other Python function, anyway). Thus, no need
to duplicate the functionality directly in the C API.
--
resolution: -> rejected
stat
Nick Coghlan added the comment:
Given Antoine's other change to the build process to fix the bootstrapping
problem, I'm OK with leaving it up to anyone hacking on _bootstrap.py to
remember that they need to run make if they want the interpreter to see any of
their changes.
Un
Nick Coghlan added the comment:
Tinkering with os.path.join, that traceback means that "name" is a str
instance, while "path" is a bytes instance.
The culprit actually appears to be the fact that the type returned by
os.listdir (et al) when handed a file descriptor is a
Nick Coghlan added the comment:
+1 for changing the API for any cases where the filesystem path is also used to
determine the return type - given that we stuffed this up for the rmtree
implementation, I expect end users would be prone to making exactly the same
mistake.
Retaining the
Nick Coghlan added the comment:
Don't forget to point people to os.fsencode() if they actually wanted bytes, or
os.fsdecode() if they already have bytes and want to convert them to text.
--
___
Python tracker
<http://bugs.python.org/is
New submission from Nick Coghlan :
As seen in #4489, the traceback when mixing str and bytes in os.path.join is
rather cryptic and hard to decipher if you've never encountered it before:
>>> import os.path
>>> os.path.join(b'', '')
Traceback (mo
Nick Coghlan added the comment:
Yeah, that should do it - I just hadn't looked at the structure of the code to
see how annoying it will be to separate the check out from the if statement.
Good use case for PEP 409, too :)
--
___
Python tr
Nick Coghlan added the comment:
I'm not comfortable adding a dependency on inspect to the contextlib module,
and I think it would be undesirable to have too many cases where the early
validation is silently skipped.
While I like the idea of adding early signature validation to these AP
Nick Coghlan added the comment:
Close in favour of #6422 - that one at least has a patch :)
--
resolution: -> duplicate
superseder: -> timeit called from within Python should allow autoranging
___
Python tracker
<http://bugs.python.org/
Nick Coghlan added the comment:
In #5442, I proposed leaving the architecture of the module alone, and simply
exposing the main module functionality as a high level helper function:
def measure(stmt="pass", setup="pass", timer=default_timer,
repeat=de
Nick Coghlan added the comment:
Oops, that link reference should have been to #5441.
--
___
Python tracker
<http://bugs.python.org/issue6422>
___
___
Python-bug
Changes by Nick Coghlan :
--
resolution: accepted ->
___
Python tracker
<http://bugs.python.org/issue6422>
___
___
Python-bugs-list mailing list
Unsubscri
Changes by Nick Coghlan :
--
status: open -> closed
___
Python tracker
<http://bugs.python.org/issue5441>
___
___
Python-bugs-list mailing list
Unsubscri
Nick Coghlan added the comment:
The from clause is intended for replacing previous exceptions with *new*
exceptions, not editing the attributes of existing ones which may already
have a different __cause__ set. So - 1 from me, even for 3.4. A patch to
the docs explaining that this is not
Nick Coghlan added the comment:
The purpose of the from clause in general is to change the exception *type*
(for example, from KeyError -> AttributeError or vice-versa), or to add
additional information without losing access to any previous information (like
the original traceback). This
Nick Coghlan added the comment:
Correctly avoiding symlink attacks, which is the whole reason the POSIX *at
variants and the dir_fd parameters were added, is not trivial in general.
os.fwalk and shutil.rmtree went through many iterations before reaching a state
where they should successfully
Nick Coghlan added the comment:
My current opinion is that this should be a PEP for 3.4, to make sure we flush
out all the corner cases and other details correctly.
--
versions: +Python 3.4 -Python 3.3
___
Python tracker
<http://bugs.python.
Nick Coghlan added the comment:
For that matter, with the relevant codecs restored in 3.2, a transform() helper
could probably be added to six (or a new project on PyPI) to prototype the
approach.
--
___
Python tracker
<http://bugs.python.
Nick Coghlan added the comment:
Setting as a release blocker for 3.4 - this is important.
--
priority: normal -> release blocker
stage: commit review ->
___
Python tracker
<http://bugs.python.org/
New submission from Nick Coghlan :
As discussed on python-ideas, swapping out the sys.std* streams can be fraught
with peril. This can make writing code that wants to support an "--encoding"
option for pipeline processing difficult.
The proposal [1] is that TextIOWrapper
Nick Coghlan added the comment:
Yep - note that PEP 409 was updated to say that the the implementation
discussion has been superceded by PEP 415. It may be worth making that note
more prominent, though...
--
___
Python tracker
<h
Nick Coghlan added the comment:
If you don't want the exception context set *at all*, just use a pass statement
to get out of the exception before trying the fallback.
try:
return windows_module.getch()
except NameError:
pass # No exception cha
Nick Coghlan added the comment:
Correction, your try block is overbroad and will suppress errors in the getch
implementation. This is better:
try:
_getch = windows_module.getch
except NameError:
_ getch = fallback_module.getch
_getch()
So I'm sticking wi
Nick Coghlan added the comment:
The os.path.exists() docs already cover all the gory details of when it may be
false due to limited permissions. The is* docs refer to this by their use of
the word "existing", but that's probably too subtle.
I suggest adding an extra sentence
Nick Coghlan added the comment:
Indeed. However, the current alternatives (based on detach() and fileNo()) are
also problematic - using detach() breaks the corresponding sys.__std*__ entry,
while using fileNo() means you now have two independent IO stacks using the
same underlying descriptor
Nick Coghlan added the comment:
Alternatively, perhaps it would make sense to have a "reopen()" method on file
objects that covers the necessary dance to reopen with the correct flags?
That would solve more problems than just this one (possibly including making it
possible
Nick Coghlan added the comment:
I've thought of a different approach to this which should be less hazardous to
doctests: include the truncation warning in the *traceback* section, rather
than in the header. doctests already ignore the details of that section because
checking it i
Nick Coghlan added the comment:
Just letting people know I'm working on a patch that updates the class
constructors to give meaningful error messages for malformed addresses, instead
of assuming that users can figure it out just by looking at the address. I'm
also updating the tes
Nick Coghlan added the comment:
There's actually one semantic change in that last patch: IPv4Address (et al)
will now accept leading zeroes in those cases where they're *not* ambiguous
(i.e. values less than 8, which have identical representations in both decimal
and octa
Nick Coghlan added the comment:
Also, I noted the provisional API status in the NEWS message, but I wonder if
it would make more sense to leave that kind of note for the commit messages.
--
___
Python tracker
<http://bugs.python.org/issue14
New submission from Nick Coghlan :
The pkgutil import emulation is insane and permits modules identifiers to
contain paths.
Identified in #15230 (reporting some very surprising behaviour from
runpy.run_module).
--
messages: 164806
nosy: brett.cannon, ncoghlan
priority: normal
Nick Coghlan added the comment:
I would add a simple note to the exceptions section:
"Note, when "shell=True", OSError will be raised by the child only if the
selected shell itself cannot be found. To determine if the shell failed to find
the requested application, it is nec
Nick Coghlan added the comment:
Firstly, I think you've identified a real bug with __package__ not being set
correctly when using runpy.run_path (and I have updated this issue title
accordingly).
I have also created a separate bug report (#15272) for the bizarre behaviour
you identifi
Changes by Nick Coghlan :
--
assignee: -> ncoghlan
___
Python tracker
<http://bugs.python.org/issue15230>
___
___
Python-bugs-list mailing list
Unsubscri
Nick Coghlan added the comment:
You may also have identified a bug with pkgutil's import emulation failing to
clean up sys.modules correctly when an import fails.
--
___
Python tracker
<http://bugs.python.org/is
Nick Coghlan added the comment:
Sorry, that's not accurate - you have enough code in b.py to make the relative
import work by convincing the interpreter it's actually being done in a package.
So what you're seeing in that regard is the fact that runpy is not any kind of
sand
Nick Coghlan added the comment:
Many of Serhiy's tweaks were also micro-optimisations, but I committed them
mainly because I found them easier to read.
--
___
Python tracker
<http://bugs.python.org/is
Nick Coghlan added the comment:
My guess would be that they're intended as a micro-optimisation (instance
lookups should be faster than class attribute lookups). You do lose on the
memory front though, since they make the instances bigger.
Without a micro benchmark of any kind, I&
Nick Coghlan added the comment:
FWIW, if we *were* going to micro-optimise ipaddress, one of the first places
I'd start is using __slots__ to try to bring the instance size down (since any
application using the module is likely to end up creating a *lot* of these
objects). There are
Nick Coghlan added the comment:
Yup, it's broken (doesn't handle NotImplemented correctly) and Raymond
isn't keen on fixing it. I should add a link to the relevant tracker issue
in a comment, though.
--
___
Python tracker
<http
Nick Coghlan added the comment:
As another in-the-standard-library uses case: my additions to the ipaddress
test suite are really crying out for parameterised test support.
My current solution is adequate for coverage and debugging purposes (a custom
assert applied to multiple values in a
Changes by Nick Coghlan :
--
title: total_ordering -> functools.total_ordering fails to handle
NotImplemented correctly
___
Python tracker
<http://bugs.python.org/issu
Nick Coghlan added the comment:
The PyPy and 3.3 behaviour are actually correct according to the spec, but it's
*really* unclear in PEP 302.
sys.meta_path accepts finder objects. These are explicitly documented as
returning "None" from find_module() to indicate "try the n
Nick Coghlan added the comment:
Perhaps the porting section in the 3.3 What's New?
--
___
Python tracker
<http://bugs.python.org/issue15292>
___
___
Pytho
New submission from Nick Coghlan :
There's some lingering weirdness in runpy due to its reliance on pkgutil's
import emulation (see #15272).
For 3.3, I'd like to explicitly switch runpy over to using importlib. This was
supposed to happen automatically, but the
Nick Coghlan added the comment:
I've taken 3.2 and 2.7 off the list - no doubt someone, somewhere is relying on
this particular piece of missing input validation, so it's not worth risking
breakage in a point release.
I think it's worth fixing for 3.3, though.
--
nosy
New submission from Nick Coghlan :
In writing a post looking at a potentially different way of handling codec
pipelines that is source compatible across both Python 2 and 3, I noticed that
the bytes-bytes codec aliases are all missing but are still listed in the
documentation as being
Nick Coghlan added the comment:
While you're right about that, the docs shouldn't be claiming they're available
as long as that is the case.
Updated the issue to make it clear this is a problem with the current docs and
the change in the nature of the encode() and decode() met
Nick Coghlan added the comment:
FWIW it's, I've been thinking further about this recently and I think
implementing this feature as builtin methods is the wrong way to approach it.
Instead, I propose the addition of codecs.encode and codecs.decode methods that
are type neutral (l
Nick Coghlan added the comment:
Given the scope of the issue, a concurrent programming howto that lists some of
the available options and the caveats associated with each of them seems
worthwhile.
--
nosy: +ncoghlan
___
Python tracker
<h
Nick Coghlan added the comment:
Indeed, pydoc relies on pkgutil.walk_packages to work out what to document, and
that's broken currently due to the reliance on a non-standard importer API that
isn't in PEP 302 (not even as an optional extension, like the get_filename()
used by runpy
Nick Coghlan added the comment:
*sigh* And, of course, there's no meaningful regression test defined for
pkgutil.walk_packages, which is why the test suite didn't pick this up :(
Oh well, at least I have a clear place to start.
--
assignee: -
New submission from Nick Coghlan :
To set the stage for fixing the regression reported in #15343, I'm going
through and clearly marking the parts of pkgutil which should no longer be used
now that the default import system is PEP 302 compliant.
--
assignee: ncoghlan
compo
Nick Coghlan added the comment:
Unassigning for the moment - I'm busy fighting fires in pkgutil (which is
actually broken due to the import changes). If someone else has time to flesh
out the method and data attribute documentation for the ipaddress classes, that
would be
Nick Coghlan added the comment:
The pkgutil import emulation is being deprecated in 3.3
However, I'll use this as useful input for any new walk_packages tests added
while resolving #15343
--
resolution: -> wont fix
stage: needs patch -> committed/rejected
status: ope
Nick Coghlan added the comment:
I'll add a regression test for this as part of my purge of any internal usage
of the pkgutil import emulation.
--
___
Python tracker
<http://bugs.python.org/is
Nick Coghlan added the comment:
OK, the way I plan to tackle this is to actually redirect the pkgutil APIs to
importlib (rather than using the private import emulation).
The reason I'm going to tackle it like this is that there are some assumptions
pkgutil deals with that importlib do
Nick Coghlan added the comment:
Oh, special - importlib.find_loader() currently expects every module to have a
__loader__ attribute, but there's at least one that doesn't: __main__.
Guess I'll be tweaking it to handle that case more grace
Nick Coghlan added the comment:
Ah, __main__, thou art such a special beast.
Here's my current plan:
1. __main__ will be initialised with __loader__ set to BuiltinImporter. This
covers cases like the interactive prompt or execution of stdin where there is
no source code availabl
Nick Coghlan added the comment:
OK, I think runpy and __main__.__loader__ are all sorted now. The relevant
runpy and command line script tests all check for the expected values, and I
added a couple of new tests to cover the "-c" and "reading from stdin" cases.
-
Changes by Nick Coghlan :
--
assignee: -> ncoghlan
___
Python tracker
<http://bugs.python.org/issue15272>
___
___
Python-bugs-list mailing list
Unsubscri
Nick Coghlan added the comment:
OK, this one is trickier than I thought - the exact behaviour depends on how
you traverse the code, and I believe a PEP 302 importer is technically allowed
to accept "/" in module names. (Unless there's a module names "must be valid
i
New submission from Nick Coghlan :
Created to record the egregious hack of relying on the test_runpy
infrastructure in order to test pkgutil.walk_packages.
It gets the job done, but is a really messy way of going about it. Worth
cleaning up by factoring the support code out to a helper module
Changes by Nick Coghlan :
--
dependencies: +"pydoc -w " writes out page with empty "Package
Contents" section
___
Python tracker
<http://bug
Nick Coghlan added the comment:
Also, a second test case should be added to cover the zipimporter.zipimporter
component.
--
___
Python tracker
<http://bugs.python.org/issue15
Changes by Nick Coghlan :
--
priority: low -> normal
___
Python tracker
<http://bugs.python.org/issue15358>
___
___
Python-bugs-list mailing list
Unsubscri
Nick Coghlan added the comment:
Fixing this has uncovered another issue: the old import emulation in PEP 302
ignored encoding cookies, thus merrily decoding everything as utf-8 in
"get_source()". importlib is smarter about this, which means the pydoc tests
started failing as the
Nick Coghlan added the comment:
Main change in that last commit is really the one to make pydoc ignore just
about *any* exception from get_source(). This should make it more robust
against buggy loaders, too.
--
___
Python tracker
<h
Nick Coghlan added the comment:
And a manual check confirms the higher level issue is also fixed. (I believe
there's already a meta issue somewhere about the lack of automated tests for
pydoc's emitted HTML)
--
resolution: -> fixed
status: o
Nick Coghlan added the comment:
I'm pretty sure I got them all while fixing #15343
--
resolution: -> fixed
stage: -> committed/rejected
status: open -> closed
___
Python tracker
<http://bugs.pyth
Nick Coghlan added the comment:
Right, this is a separate bug in pkgutil. Specifically, when it goes to import
a package in order to check it for submodules, it invokes the global import
system via __import__() rather than constraining the import to the path
argument supplied to
Nick Coghlan added the comment:
At the very least, the pkgutil docs need to state clearly that walk_packages
only works properly with sys.path entries, and the constraint feature may not
descend into packages correctly if an entry is shadowed by a sys.modules entry
or an entry earlier on
New submission from Nick Coghlan :
test_runpy doesn't currently check that code can be correctly executed from
namespace packages
--
components: Tests
messages: 165613
nosy: ncoghlan
priority: normal
severity: normal
status: open
title: test_runpy should include namespace package
New submission from Nick Coghlan :
test_cmd_line_script doesn't currently check that namespace packages and
submodules of namespace packages can be executed via the -m switch
--
components: Library (Lib)
messages: 165614
nosy: ncoghlan
priority: normal
severity: normal
status:
Nick Coghlan added the comment:
The walk_packages tests should also be enhanced to ensure correctly handling of
namespace packages
--
___
Python tracker
<http://bugs.python.org/issue15
Nick Coghlan added the comment:
Note that the problem described in #14787 will limit the scope of what can be
tested.
--
___
Python tracker
<http://bugs.python.org/issue15
Nick Coghlan added the comment:
I just realised this is going to behave strangely with namespace packages as
well: the __import__ step will pick up *every* portion of the namespace
package, not just those defined in the identified subset of sys.path
Nick Coghlan added the comment:
At least we know the hash randomisation is working :)
Spurious changes when freezing modules seems like a legitimate reason to fix it
- the import benchmarks would probably give the compiler enough of a workout to
highlight if the sorting is excessively time
Nick Coghlan added the comment:
That sounds good, and thanks for working on this - if we can get the tests
cleaned up for the 3.3 release, that will make ongoing maintenance a lot easier.
--
___
Python tracker
<http://bugs.python.org/issue15
Nick Coghlan added the comment:
I'd be more inclined to tighten up the check for the "can't mix" message to
something like:
valid_types = all(isinstance(s, (str, bytes, bytearray)) for s in (a, ) + p)
if valid_types:
# Must have a mixture of text and binary data
Nick Coghlan added the comment:
*grumble*grumble*os-x-and-its-crazy-symlink-as-tmp-dir*grumble*grumble*
I'm fairly sure I've tripped over this kind of thing before, so I believe I
know how to fix it (add a realpath() call when figuring out the expected path
value).
The buildbots w
Nick Coghlan added the comment:
One trap to watch out for: it's best to call os.path.realpath() on any
temporary directories created, since the interpreter tends to do that
internally.
I'd previously dealt with this in script_helper.temp_dir, but it came up again
when I tighte
Nick Coghlan added the comment:
Looks good to me
--
___
Python tracker
<http://bugs.python.org/issue15377>
___
___
Python-bugs-list mailing list
Unsubscribe:
Nick Coghlan added the comment:
Closing again, since the OS X buildbot was happy with the change on 3.2. If
trunk fails (which it shouldn't), that should be filed as a new issue (as it
would almost certainly relate to the move from the pkgutil emulation to
importlib).
--
s
Nick Coghlan added the comment:
OK, I just pushed a tweak to the test that will hopefully resolve the problem.
If the tweak works, though, it does suggest we have a bit of weirdness in
importlib: it means pkgutil.get_importer("") is returning an importer on POSIX
systems, but None
New submission from Nick Coghlan :
As noted in #15314, one of the pkgutil tests was failing on Windows because
pkgutil.get_importer("") was returning None. On my Fedora system it returns
FileFinder(".").
I've now tweaked that particular test to use a bogus path st
Changes by Nick Coghlan :
--
stage: -> test needed
type: -> behavior
___
Python tracker
<http://bugs.python.org/issue15384>
___
___
Python-bugs-list
New submission from Nick Coghlan :
Directory setup: linkdir is a symlink to realdir
$ python3 -c "import sys; sys.path.insert(0, 'linkdir'); import runpy;
print(runpy.run_module('foo')['__file__'])"
/home/ncoghlan/devel/play/realdir/foo.py
$ ../py3k/
Changes by Nick Coghlan :
--
resolution: -> works for me
status: open -> closed
___
Python tracker
<http://bugs.python.org/issue15385>
___
___
Python-bugs-
Nick Coghlan added the comment:
Buildbots are green (at least as far as this issue goes), so closing again.
I created #15384 for the surprising behaviour of FileFinder on Windows.
#15385 records my investigation into the odd symlink behaviour on 3.2 (it was
just a discrepancy between the
New submission from Nick Coghlan :
I haven't worked out how yet, but importlib.machinery is managing to bypass the
replacement of importlib._bootstrap with _frozen_importlib:
Python 3.3.0b1 (default:8bf691d0b004+, Jul 15 2012, 23:20:06)
[GCC 4.7.0 20120507 (Red Hat 4.7.0-5)] on linux
Changes by Nick Coghlan :
--
title: Still getting two copies of importlib._boostrap -> Still getting two
copies of importlib._bootstrap
___
Python tracker
<http://bugs.python.org/issu
Nick Coghlan added the comment:
Well, that's weird. A manual test on my Windows 7 gaming rig indicates that the
old test *should* have worked on Windows as well.
I have no idea how to investigate this further. David, are you able to check
what pkgutil.get_importer("") retur
3301 - 3400 of 6301 matches
Mail list logo
