Fix multiplying a list by an integer (list *= int): detect the
integer overflow when the new allocated length is close to the
maximum size. Issue reported by Jordan Limor.
list_resize() now checks for integer overflow before multiplying the
new allocated length by the list item size (sizeof(PyObject*)).
Previously, checkbuttons in different parent widgets could have the same
short name and share the same state if arguments "name" and "variable" are
not specified. Now they are globally unique.
Fix command line parsing: reject "-X int_max_str_digits" option with
no value (invalid) when the PYTHONINTMAXSTRDIGITS environment
variable is set to a valid limit.
Only call tkinter.tk and its follow-up code in _init_tk_type when requires('gui')
does not raise. This function can be called as an unintended side-effect of
calling other idlelib code as part of tests on macOS without a GUI enabled.
The main problem was that an unluckily timed task cancellation could cause
the semaphore to be stuck. There were also doubts about strict FIFO ordering
of tasks allowed to pass.
The Semaphore implementation was rewritten to be more similar to Lock.
Many tests for edge cases (including cancellation) were added.
They were undertested, and since #96954 might involve a
rewrite of this part of the code we want to ensure that
there won't be any behavioral change.
Co-authored-by: Carl Friedrich Bolz-Tereick <cfbolz@gmx.de>
This reverts commit 0587810698.
Reason: This broke buildbots (some warnings added by that commit are turned to errors in the SSL buildbot).
Repro: ./python Lib/test/ssltests.py
- Improve error message when parameter without a default follows one with a default
- Show same error message when positional-only params precede the default/non-default sequence
Warn on loop initialization, when setting the wakeup fd disturbs a previously set wakeup fd, and on loop closing, when upon resetting the wakeup fd, we find it has been changed by someone else.
Previously codeop.compile_command() emitted compiler warnings (SyntaxWarning or
DeprecationWarning) and raised a SyntaxError for incomplete input containing
a potentially incorrect code. Now it always returns None for incomplete input
without emitting any warnings.
* fix: annotate `pstats.FunctionProfile.ncalls` as `str`
This change aligns the type annotation of `pstats.FunctionProfile.ncalls` with its runtime type.
The test file, a modified version of Lib/test/audiodata/pluck-pcm24.wav, was provided by Andrea Celletti on the bug tracker.
Co-authored-by: Zackery Spytz <zspytz@gmail.com>
A regression would still absolutely fail and even a flaky pass isn't
harmful as it'd fail most of the time across our N system test runs.
Windows has a low resolution timer and CI systems are prone to odd
timing so this just gives more leeway to avoid flakiness.
This makes tokenizer.c:valid_utf8 match stringlib/codecs.h:decode_utf8.
It also fixes an off-by-one error introduced in 3.10 for the line number when the tokenizer reports bad UTF8.
This removes the unused `name` variable in the block where `ArgumentTypeError` is handled.
`ArgumentTypeError` errors are handled by showing just the string of the exception; unlike `ValueError`, the name (`__name__`) of the function is not included in the error message.
Fixes#96548
This doesn't happen naturally, but is allowed by the ASDL and compiler.
We don't want to change ASDL for backward compatibility reasons
(#57645, #92987)
Converting a large enough `int` to a decimal string raises `ValueError` as expected. However, the raise comes _after_ the quadratic-time base-conversion algorithm has run to completion. For effective DOS prevention, we need some kind of check before entering the quadratic-time loop. Oops! =)
The quick fix: essentially we catch _most_ values that exceed the threshold up front. Those that slip through will still be on the small side (read: sufficiently fast), and will get caught by the existing check so that the limit remains exact.
The justification for the current check. The C code check is:
```c
max_str_digits / (3 * PyLong_SHIFT) <= (size_a - 11) / 10
```
In GitHub markdown math-speak, writing $M$ for `max_str_digits`, $L$ for `PyLong_SHIFT` and $s$ for `size_a`, that check is:
$$\left\lfloor\frac{M}{3L}\right\rfloor \le \left\lfloor\frac{s - 11}{10}\right\rfloor$$
From this it follows that
$$\frac{M}{3L} < \frac{s-1}{10}$$
hence that
$$\frac{L(s-1)}{M} > \frac{10}{3} > \log_2(10).$$
So
$$2^{L(s-1)} > 10^M.$$
But our input integer $a$ satisfies $|a| \ge 2^{L(s-1)}$, so $|a|$ is larger than $10^M$. This shows that we don't accidentally capture anything _below_ the intended limit in the check.
<!-- gh-issue-number: gh-95778 -->
* Issue: gh-95778
<!-- /gh-issue-number -->
Co-authored-by: Gregory P. Smith [Google LLC] <greg@krypto.org>
* gh-68163: Correct conversion of Rational instances to float
Also document that numerator/denominator properties are instances of Integral.
Co-authored-by: Mark Dickinson <dickinsm@gmail.com>
Link to #93884
* Test with some large negative and positive values(out of range of a longlong,i.e.[-2\*\*63, 2\*\*63-1])
* Test with objects of non-int type
Automerge-Triggered-By: GH:mdickinson
Co-authored-by: Jelle Zijlstra <jelle.zijlstra@gmail.com>
Co-authored-by: Martin Panter <vadmium@users.noreply.github.com>
Co-authored-by: Terry Jan Reedy <tjreedy@udel.edu>
Integer to and from text conversions via CPython's bignum `int` type is not safe against denial of service attacks due to malicious input. Very large input strings with hundred thousands of digits can consume several CPU seconds.
This PR comes fresh from a pile of work done in our private PSRT security response team repo.
Signed-off-by: Christian Heimes [Red Hat] <christian@python.org>
Tons-of-polishing-up-by: Gregory P. Smith [Google] <greg@krypto.org>
Reviews via the private PSRT repo via many others (see the NEWS entry in the PR).
<!-- gh-issue-number: gh-95778 -->
* Issue: gh-95778
<!-- /gh-issue-number -->
I wrote up [a one pager for the release managers](https://docs.google.com/document/d/1KjuF_aXlzPUxTK4BMgezGJ2Pn7uevfX7g0_mvgHlL7Y/edit#). Much of that text wound up in the Issue. Backports PRs already exist. See the issue for links.
* gh-96132: Add some comments and minor fixes missed in the original PR
* Update Doc/using/cmdline.rst
Co-authored-by: Kumar Aditya <59607654+kumaraditya303@users.noreply.github.com>
Co-authored-by: Kumar Aditya <59607654+kumaraditya303@users.noreply.github.com>
⚠️⚠️ Note for reviewers, hackers and fellow systems/low-level/compiler engineers ⚠️⚠️
If you have a lot of experience with this kind of shenanigans and want to improve the **first** version, **please make a PR against my branch** or **reach out by email** or **suggest code changes directly on GitHub**.
If you have any **refinements or optimizations** please, wait until the first version is merged before starting hacking or proposing those so we can keep this PR productive.
datetime.isoformat generates the tzoffset with colons, but there
was no format code to make strftime output the same format.
for simplicity and consistency the %:z formatting behaves mostly
as %z, with the exception of adding colons. this includes the
dynamic behaviour of adding seconds and microseconds only when
needed (when not 0).
this fixes the still open "generate" part of this issue:
https://github.com/python/cpython/issues/69142
Co-authored-by: Kumar Aditya <59607654+kumaraditya303@users.noreply.github.com>
* inspect.getsource: avoid stat on file in linecache
The check for os.path.exists() on source file is postponed in
inspect.getsourcefile() until needed avoiding an expensive filesystem
stat call and PEP 302 module loader check is moved last for performance
since it is an uncommon case.
Editors don't agree that `test_source_encoding.py` was valid koi8-r, making it
hard to edit that file without the editor breaking it in some way (see gh-96272).
Only one test actually relied on the koi8-r encoding and it was a duplicate of a
test from the deprecated `imp` module's `test_imp`, so here we replace
`test_pep263` with `test_import_encoded_module` stolen from `test_imp` and
set `test_source_encoding.py`'s encoding to utf-8 to make editing it easier
going forward.
find_unused_port() has an inherent race condition, but we can't use
bind_port() as that uses .getsockname() which this test is exercising.
Try binding to unused ports a few times before failing.
Signed-off-by: Ross Burton <ross.burton@arm.com>
* gh-93503: Add APIs to set profiling and tracing functions in all threads in the C-API
* Use a separate API
* Fix NEWS entry
* Add locks around the loop
* Document ignoring exceptions
* Use the new APIs in the sys module
* Update docs
Tests for IsolatedAsyncioTestCase.debug() rely on the runner be closed
in __del__. It makes tests depending on the GC an unreliable on other
implementations. It is better to close the runner explicitly even if
currently there is no a public API for this.
The clearing of the temporary directory is not working on some platforms and
leaving behind files.
This has been updated to use the pattern in test_cmd_line.py [1] using the
special TESTFN rather than a test directory.
[1] https://github.com/python/cpython/blob/main/Lib/test/test_cmd_line.py#L559
* Treat tp_weakref and tp_dictoffset like other opaque slots for multiple inheritance.
* Document Py_TPFLAGS_MANAGED_DICT and Py_TPFLAGS_MANAGED_WEAKREF in what's new.
- Limited API needs to be enabled per source file
- Some builds don't support Limited API, so Limited API tests must be skipped on those builds
(currently this is `Py_TRACE_REFS`, but that may change.)
- `Py_LIMITED_API` must be defined before `<Python.h>` is included.
This puts the hoop-jumping in `testcapi/parts.h`, so individual
test files can be relatively simple. (Currently that's only
`vectorcall_limited.c`, imagine more.)
When a task catches CancelledError and raises some other error,
the other error should not silently be suppressed.
Any scenario where a task crashes in cleanup upon cancellation
will now result in an ExceptionGroup wrapping the crash(es)
instead of propagating CancelledError and ignoring the side errors.
NOTE: This represents a change in behavior (hence the need to
change several tests). But it is only an edge case.
Co-authored-by: Thomas Grainger <tagrain@gmail.com>
When loading a source file from disk, there is a separate UTF-8 validator
distinct from the one in `unicode_decode_utf8`. This exercises that code path
with the same set of invalid inputs as we use for testing the "other" UTF-8
decoder.
If kernel fips is enabled, we get permission error upon doing
`import crypt`. So, if kernel fips is enabled, disable the
unallowed hashing methods.
Python 3.9.1 (default, May 10 2022, 11:36:26)
[GCC 10.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import crypt
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/usr/lib/python3.9/crypt.py", line 117, in <module>
_add_method('MD5', '1', 8, 34)
File "/usr/lib/python3.9/crypt.py", line 94, in _add_method
result = crypt('', salt)
File "/usr/lib/python3.9/crypt.py", line 82, in crypt
return _crypt.crypt(word, salt)
PermissionError: [Errno 1] Operation not permitted
Signed-off-by: Shreenidhi Shedi <sshedi@vmware.com>
* Make sure that tp_dictoffset is correct with Py_TPFLAGS_MANAGED_DICT is set.
* Avoid traversing managed dict twice when subclassing class with Py_TPFLAGS_MANAGED_DICT set.
Automate WASM build with a new Python script. The script provides
several build profiles with configure flags for Emscripten flavors
and WASI. The script can detect and use Emscripten SDK and WASI SDK from
default locations or env vars.
``configure`` now detects Node arguments and creates HOSTRUNNER
arguments for Node 16. It also sets correct arguments for
``wasm64-emscripten``.
Co-authored-by: Brett Cannon <brett@python.org>
Have pathlib use `os.path.join()` to join arguments to the `PurePath` initialiser, which fixes a minor bug when handling relative paths with drives.
Previously:
```python
>>> from pathlib import PureWindowsPath
>>> a = 'C:/a/b'
>>> b = 'C:x/y'
>>> PureWindowsPath(a, b)
PureWindowsPath('C:x/y')
```
Now:
```python
>>> PureWindowsPath(a, b)
PureWindowsPath('C:/a/b/x/y')
```
We only statically initialize for core code and builtin modules. Extension modules still create
the tuple at runtime. We'll solve that part of interpreter isolation separately.
This change includes generated code. The non-generated changes are in:
* Tools/clinic/clinic.py
* Python/getargs.c
* Include/cpython/modsupport.h
* Makefile.pre.in (re-generate global strings after running clinic)
* very minor tweaks to Modules/_codecsmodule.c and Python/Python-tokenize.c
All other changes are generated code (clinic, global strings).
#91242 replaced the Windows chm help file with a copy
of the html docs. This PR replaces the IDLE code that
fetches the Windows local help url passed to os.startfile.
Co-authored-by: Steve Dower
Under certain build conditions, test_check_c_globals fails. This fix takes the same approach as we took for gh-84236 (via gh-20095). We'll be removing use of distutils in the c-analyzer at some point. Until then we'll hide the warning filter.
An unrecognized format character in PyUnicode_FromFormat() and
PyUnicode_FromFormatV() now sets a SystemError.
In previous versions it caused all the rest of the format string to be
copied as-is to the result string, and any extra arguments discarded.
gh-93243
This PR is required to reduce diffs of the following porting (no need to either maintain documentation and tests consistent with each porting step, or try to port everything and remove smtpd in a single PR).
Automerge-Triggered-By: GH:warsaw
Have `pathlib.WindowsPath.is_mount()` call `ntpath.ismount()`. Previously it raised `NotImplementedError` unconditionally.
https://bugs.python.org/issue42777
Adds a regression test for an re slowdown observed by rjsmin.
Uses multiprocessing to kill the test after SHORT_TIMEOUT.
Co-authored-by: Oleg Iarygin <dralife@yandex.ru>
Co-authored-by: Christian Heimes <christian@python.org>
* Add test for inheriting explicit __dict__ and weakref.
* Restore 3.10 behavior for multiple inheritance of C extension classes that store their dictionary at the end of the struct.
The API documentation for [findtext](https://docs.python.org/3/library/xml.etree.elementtree.html#xml.etree.ElementTree.Element.findtext) states that this function gives back an empty string on "no text content." With the previous implementation, this would give back a empty string even on text content values such as 0 or False. This patch attempts to resolve that by only giving back an empty string if the text attribute is set to `None`. Resolves#91447.
Automerge-Triggered-By: GH:gvanrossum
If one selects whole lines, as the sidebar makes easy, do not
add an extra line. Only move the end of a selection to the
beginning of the next line when not already at the beginning
of a line. (Also improve the surrounding code.)
Move `Select All` above `Cut` as it is used with `Cut` and `Copy` but not `Paste`. Add a separator between `Replace` and `Go to Line` to separate items that belong to the 'Edit-find' (above) and 'Edit-show' (below) IDLE github project topics.
When keyword argument name is an instance of a str subclass with
overloaded methods __eq__ and __hash__, the former code could not find
the name of an extraneous keyword argument to report an error, and
_PyArg_UnpackKeywords() returned success without setting the
corresponding cell in the linearized arguments array. But since the number
of expected initialized cells is determined as the total number of passed
arguments, this lead to reading NULL as a keyword parameter value, that
caused SystemError or crash or other undesired behavior.
This is the last precursor to storing tp_subclasses (and tp_weaklist) on the interpreter state for static builtin types.
Here we add per-type storage on PyInterpreterState, but only for the static builtin types. This involves the following:
* add PyInterpreterState.types
* move PyInterpreterState.type_cache to it
* add a "num_builtins_initialized" field
* add a "builtins" field (a static array big enough for all the static builtin types)
* add _PyStaticType_GetState() to look up a static builtin type's state
* (temporarily) add PyTypeObject.tp_static_builtin_index (to hold the type's index into PyInterpreterState.types.builtins)
We will be eliminating tp_static_builtin_index in a later change.
Previously, when using `functools.wrap` around them (and inherit their docstrings), sphinx renders the docstrings badly and raises warnings about wrong indent.
* gh-95218: Move tests for importlib.resources into test_importlib.resources.
* Also update makefile
* Include test_importlib/resources in code ownership rule.
This PR partially reverts gh-24421 (PR) and fixes the remaining concerns
given in gh-93044 (issue):
- keyword arguments are passed as positional arguments to factory()
- if an argument is not passed to sqlite3.connect(), its default value
is passed to factory()
Co-authored-by: Serhiy Storchaka <storchaka@gmail.com>
This appears to be a typo. It causes try_protocol_combo to try to turn
on SSL 3.0 when testing PROTOCOL_SSLv23 (aka PROTOCOL_TLS), which
doesn't make any sense. Fix it to be PROTOCOL_SSLv3.
Without this, try_protocol_combo is actually setting
context.minimum_version to SSLv3 when called as
try_protocol_combo(ssl.PROTOCOL_TLS, ssl.PROTOCOL_TLS, True)
One would think this causes a no-ssl3 OpenSSL build to fail, but OpenSSL
forgot to make SSL_CTX_set_min_proto_version(SSL3_VERSION) does not
notice no-ssl3, so this typo has gone undetected. But we should still
fix the typo because, presumably, a future version of OpenSSL will
remove SSL 3.0 and do so more thoroughly, at which point this will
break.
* gh-94949: Disallow parsing parenthesised ctx manager with old feature_version
* 📜🤖 Added by blurb_it.
* Allow it with feature_version=(3, 9) as well
Co-authored-by: blurb-it[bot] <43283697+blurb-it[bot]@users.noreply.github.com>
Support for bytes broke sometime between Python 3.2 and 3.6 and has been broken ever since. Trying to bring back supports is surprisingly difficult in the face of -b and checking for keys in sys.path_importer_cache. Since the support was broken for so long, trying to overcome the difficulty of bringing back the support has been deemed not worth it.
Co-authored-by: Eryk Sun <eryksun@gmail.com>
Co-authored-by: Brett Cannon <brett@python.org>
The case where there are more than (1 << 15) lines was not covered.
I don't know if increasing test coverage requires a blurb -- let me know if it does.
Automerge-Triggered-By: GH:brandtbucher
When binding a unix socket to an empty address on Linux, the socket is
automatically bound to an available address in the abstract namespace.
>>> s = socket.socket(socket.AF_UNIX, socket.SOCK_STREAM)
>>> s.bind("")
>>> s.getsockname()
b'\x0075499'
Since python 3.9, the socket is bound to the one address:
>>> s.getsockname()
b'\x00'
And trying to bind multiple sockets will fail with:
Traceback (most recent call last):
File "/home/nsoffer/src/cpython/Lib/test/test_socket.py", line 5553, in testAutobind
s2.bind("")
OSError: [Errno 98] Address already in use
Added 2 tests:
- Auto binding empty address on Linux
- Failing to bind an empty address on other platforms
Fixes f6b3a07b7d (bpo-44493: Add missing terminated NUL in sockaddr_un's length (GH-26866)
This makes calls to co_lnotab to exercise this code, as well
as generating synthetically large code to exercise the corner
cases where line numbers need multiple bytes.
Automerge-Triggered-By: GH:brandtbucher
The current status quo is that private attribute names are not
mangled when a class is matched. I've added a test to
document/legimize this behaviour.
Co-authored-by: Brandt Bucher <brandtbucher@gmail.com>
This is a quick-and-dirty way to run the C++ tests.
It can definitely be improved in the future, but it should fail when things go wrong.
- Run test functions on import (yes, this can definitely be improved)
- Fudge setuptools metadata (name & version) to make the extension installable
- Install and import the extension in test_cppext
Koki Saito
Ofey Chan
Gregory P. Smith
Łukasz Langa
Guido van Rossum
Cheryl Sabella
Shin-myoung-serp
Victor Stinner
Jakob Weigert
Kumar Aditya
Pablo Galindo Salgado
Serhiy Storchaka
Cyker Way
Michael Curran
Dennis Sweeney
Terry Jan Reedy
Michael Droettboom
Steve Dower
Brandt Bucher
Batuhan Taskaya
Nikita Sobolev
wim glenn
serge-sans-paille
Lysandros Nikolaou
Michel Hidalgo
Vinay Sajip
adphrost
Ruan Comelli
Yusuke Kadowaki
Benjamin Peterson
Irit Katriel
Christian Heimes
Dong-hee Na
Hendrik Makait
Zachary Ware
finefoot
Mark Shannon
philg314
Raymond Hettinger
Jelle Zijlstra
Jonathon Reinhart
Jason R. Coombs
Shantanu
Pamela Fox
Mark Dickinson
Sergey B Kirpichev
Charlie Zhao
Gregory P. Smith
Anh71me
Filip Łajszczak
Alexandru Mărășteanu
Daniel Giger
Anthony Sottile
TW
Kirill
Daniel Giger
Carl Friedrich Bolz-Tereick
Pankaj Pandey
Kristján Valur Jónsson
Harry
Sam Ezeh
Ross Burton
Kevin Kirsche
Jeong YunWon
Duncan Grisby
Petr Viktorin
Ken Jin
Shreenidhi Shedi
Erlend E. Aasland
Barney Gale
Yury Selivanov
Eric Snow
David Bonner
Matthias Görgens
JuniorJPDJ
Sion Kang
da-woods
Oleg Iarygin
Fantix King
Barry Warsaw
Miro Hrončok
Jo, Yunjin
Stéphane Bidoul
Brandt Bucher
Eugene Triguba
Jacob Walls
Robert O'Shea
Raghunandan Bhat
Jingxuan He
Thomas Grainger
Roman Novak
Ethan Furman
Inada Naoki
Erlend Egeberg Aasland
Stanislav Zmiev
David Benjamin
Florian Bruhin
Paul Moore
Nir Soffer
Itay
Max Zhenzhera