]> git.ipfire.org Git - thirdparty/Python/cpython.git/log
thirdparty/Python/cpython.git
11 days agogh-151942: Fix all Sphinx nitpicks in the 2.5 What's New (#153120)
Stan Ulbrych [Fri, 10 Jul 2026 10:08:31 +0000 (12:08 +0200)] 
gh-151942: Fix all Sphinx nitpicks in the 2.5 What's New (#153120)

12 days agogh-153201: Make TSan CI mandatory (#153491)
Kumar Aditya [Fri, 10 Jul 2026 09:36:26 +0000 (15:06 +0530)] 
gh-153201: Make TSan CI mandatory (#153491)

12 days agogh-145633: Allow PyFloat_Pack8 & PyFloat_UnPack* to fail in future CPython versions...
Petr Viktorin [Fri, 10 Jul 2026 08:35:35 +0000 (10:35 +0200)] 
gh-145633: Allow PyFloat_Pack8 & PyFloat_UnPack* to fail in future CPython versions (GH-153440)

12 days agogh-153062: Fix a crash iterating itertools.tee on the free-threaded build (gh-153063)
tonghuaroot (童话) [Fri, 10 Jul 2026 02:38:58 +0000 (10:38 +0800)] 
gh-153062: Fix a crash iterating itertools.tee on the free-threaded build (gh-153063)

itertools.tee branches share a linked list of teedataobject cells.  On the free-threaded build, iterating one branch from multiple threads, or iterating sibling branches concurrently, raced on the shared cells and each branch's position, corrupting refcounts and crashing.

Lock each teedataobject while reading, extending, or clearing it, and snapshot each branch's position under the tee object's own lock, revalidating before advancing, so the two locks are never nested.  Concurrent iteration of one tee is undefined and may raise RuntimeError as documented, but no longer crashes.

The free-threading snapshot and revalidation add per-element overhead on the default build, where the GIL already serializes access. Guard that path under Py_GIL_DISABLED so the default build keeps the original iteration and the free-threaded path is unchanged.

Co-authored-by: Neil Schemenauer <nas-github@arctrix.com>
12 days agogh-152132: Test Py_CompileString() in test_capi.test_run (#153470)
Victor Stinner [Fri, 10 Jul 2026 01:36:31 +0000 (03:36 +0200)] 
gh-152132: Test Py_CompileString() in test_capi.test_run (#153470)

PyRun functions now raises ValueError if the start argument is
invalid.

* Add Modules/_testlimitedcapi/run.c.
* Rename _Py_CompileStringObjectWithModule() to
  _Py_CompileStringObject().

12 days agogh-152132: Enhance Py_RunMain() and fix bugs (#153461)
Victor Stinner [Thu, 9 Jul 2026 23:29:53 +0000 (01:29 +0200)] 
gh-152132: Enhance Py_RunMain() and fix bugs (#153461)

* Check for signals more often. Previously, a pending exception could
  be removed by PyErr_Clear().
* Only call _PyInterpreterState_SetNotRunningMain() if
  _PyInterpreterState_SetRunningMain() has been called.
* Add pymain_error() and pyrun_error(): write the error message to
  sys.stderr instead of the C stream stderr.
* Convert _PyPathConfig_UpdateGlobal() PyStatus error to an
  exception.
* Simplify pymain_run_startup(): avoid strerror() which is decoded
  from the wrong encoding. Instead display the exception, Py_fopen()
  raises an OSError with the path object (and can raise other exceptions).
* Add pymain_set_path0() function.
* Add _PySys_FormatV() function.

12 days agoUse enums for pycore_uop_ids.h. (gh-153351)
Neil Schemenauer [Thu, 9 Jul 2026 22:08:36 +0000 (15:08 -0700)] 
Use enums for pycore_uop_ids.h. (gh-153351)

This generally reduces the sizes of diffs when uops are added or
removed.  It also helps avoid merge conflicts.

12 days agogh-98894: Quote test_dtrace tracer subcommands (#152901)
stratakis [Thu, 9 Jul 2026 20:43:54 +0000 (22:43 +0200)] 
gh-98894: Quote test_dtrace tracer subcommands (#152901)

Use shlex.join() when building tracer -c command strings
for avoiding issues with spaces.

12 days agogh-121608: Fix test_samply_profiler: only check samply once (#153424)
Victor Stinner [Thu, 9 Jul 2026 19:27:40 +0000 (21:27 +0200)] 
gh-121608: Fix test_samply_profiler: only check samply once (#153424)

Only run samply_command_works() once. Remove also dead code.

12 days agogh-152769: Enable perf trampoline on musl (Alpine Linux) (#152774)
Xiaowei Lu [Thu, 9 Jul 2026 19:17:40 +0000 (03:17 +0800)] 
gh-152769: Enable perf trampoline on musl (Alpine Linux) (#152774)

12 days agogh-152132: Fix Py_RunMain() to return an exit code (#153446)
Victor Stinner [Thu, 9 Jul 2026 18:28:15 +0000 (20:28 +0200)] 
gh-152132: Fix Py_RunMain() to return an exit code (#153446)

* Fix Py_RunMain() to return an exit code, rather than calling
  Py_Exit(), when running a script, a command, or the REPL.
* _PyRun_SimpleFile() now logs errors to stderr, for example if setting
  __main__.__file__ fails.
* Add tests on Py_RunMain() exitcode.
* Rename functions:

  * _PyRun_SimpleStringFlagsWithName() => _PyRun_SimpleString()
  * _PyRun_SimpleFileObject() => _PyRun_SimpleFile()
  * _PyRun_AnyFileObject() => _PyRun_AnyFile()
  * _PyRun_InteractiveLoopObject() => _PyRun_InteractiveLoop()

* Change _PyRun_SimpleString(), _PyRun_SimpleFile(), _PyRun_AnyFile() and
  _PyRun_InteractiveLoop() return type to PyObject*.
* pymain_repl() now displays the error if PySys_Audit() or
  import _pyrepl failed.
* PyRun_SimpleFileExFlags() and PyRun_AnyFileExFlags() now log
  PyUnicode_DecodeFSDefault() error. So these functions can no longer
  return -1 with an exception set.

12 days agogh-146531: Broaden `wm_iconbitmap` skip on macOS 26 Intel (#153348)
Hugo van Kemenade [Thu, 9 Jul 2026 18:23:31 +0000 (21:23 +0300)] 
gh-146531: Broaden `wm_iconbitmap` skip on macOS 26 Intel (#153348)

12 days agogh-152905: Decode LC_TIME items in nl_langinfo() from glibc wide data (GH-152911)
Serhiy Storchaka [Thu, 9 Jul 2026 17:36:08 +0000 (20:36 +0300)] 
gh-152905: Decode LC_TIME items in nl_langinfo() from glibc wide data (GH-152911)

On glibc, locale.nl_langinfo() now decodes the LC_TIME text items from the
wide (_NL_W*) locale data, independently of the LC_CTYPE encoding.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
12 days agogh-152997: Support system locale encodings via an iconv-based codec (GH-153001)
Serhiy Storchaka [Thu, 9 Jul 2026 17:01:32 +0000 (20:01 +0300)] 
gh-152997: Support system locale encodings via an iconv-based codec (GH-153001)

Where the C library provides iconv(), the codecs module now exposes every
encoding iconv() knows but Python has no built-in codec for -- the POSIX
counterpart of the Windows code-page support.  Use it by name (e.g.
"cp1133"), or with an "iconv:" prefix to force it over a built-in codec.

The codec is a last-resort search function and never shadows a built-in.
Both directions pivot through native-endian UTF-32, keeping one input unit
per code point so error handlers get the exact string position.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
12 days agogh-153444: Fix compile warning in `Modules/_interpchannelsmodule.c` (#153445)
sobolevn [Thu, 9 Jul 2026 16:08:29 +0000 (19:08 +0300)] 
gh-153444: Fix compile warning in `Modules/_interpchannelsmodule.c` (#153445)

12 days agogh-49555: Support international mailbox names in imaplib (GH-153391)
Serhiy Storchaka [Thu, 9 Jul 2026 16:02:56 +0000 (19:02 +0300)] 
gh-49555: Support international mailbox names in imaplib (GH-153391)

Non-ASCII mailbox names are now encoded as modified UTF-7 (RFC 3501,
section 5.1.3), so they can be passed as an ordinary str.

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
12 days agogh-90092: Fix test_curses when stdin is write-only (e.g. under nohup) (GH-153420)
Serhiy Storchaka [Thu, 9 Jul 2026 15:43:26 +0000 (18:43 +0300)] 
gh-90092: Fix test_curses when stdin is write-only (e.g. under nohup) (GH-153420)

newterm() needs a readable input fd, but nohup can leave stdin write-only.
Fall back to the output fd, as already done when stdin has no fileno.

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
12 days agogh-153422: Return bool from some tkinter query methods (GH-153429)
Serhiy Storchaka [Thu, 9 Jul 2026 14:13:42 +0000 (17:13 +0300)] 
gh-153422: Return bool from some tkinter query methods (GH-153429)

The winfo_exists(), winfo_ismapped() and winfo_viewable() methods of
tkinter widgets and Text.edit_modified() now return a bool instead of an
integer or, depending on wantobjects, a string.

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
12 days agogh-151292: Change `total_samples:u32` in the binary format to `total_samples:u64...
Maurycy Pawłowski-Wieroński [Thu, 9 Jul 2026 13:59:28 +0000 (15:59 +0200)] 
gh-151292: Change `total_samples:u32` in the binary format to `total_samples:u64` (#153425)

12 days agogh-153406: Raise ValueError, not OverflowError, for out-of-range dates in email.utils...
tonghuaroot (童话) [Thu, 9 Jul 2026 13:58:56 +0000 (21:58 +0800)] 
gh-153406: Raise ValueError, not OverflowError, for out-of-range dates in email.utils.parsedate_to_datetime (#153407)

email.utils.parsedate_to_datetime documented that it raises ValueError for an invalid date, but it leaked OverflowError when the parsed year or timezone offset was too large for the datetime and timedelta constructors, and that OverflowError also escaped the modern header parsing path since DateHeader.parse only caught ValueError. Wrap the datetime and timezone construction so an OverflowError is re-raised as a ValueError with the original chained as the cause, which restores the documented contract and lets the existing header handler record an InvalidDateDefect instead of raising.

12 days agogh-153417: Fix BytesWarning in imaplib error messages for bytes arguments (GH-153423)
Serhiy Storchaka [Thu, 9 Jul 2026 13:51:45 +0000 (16:51 +0300)] 
gh-153417: Fix BytesWarning in imaplib error messages for bytes arguments (GH-153423)

IMAP4.select() and IMAP4.uid() formatted the mailbox and command argument
with %s in their error messages, which raised BytesWarning under -bb when the
argument was bytes and masked the real error.  Use %r instead, which is safe
for bytes and also quotes the value.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
12 days agogh-152132: Test all PyRun functions in test_capi.test_run (#153352)
Victor Stinner [Thu, 9 Jul 2026 10:50:40 +0000 (12:50 +0200)] 
gh-152132: Test all PyRun functions in test_capi.test_run (#153352)

PyRun_InteractiveOneObjectEx() now raises TypeError if the second
parameter type is not str.

12 days agogh-69134: Harden NotebookTest.test_traversal (GH-153403)
Serhiy Storchaka [Thu, 9 Jul 2026 09:54:13 +0000 (12:54 +0300)] 
gh-69134: Harden NotebookTest.test_traversal (GH-153403)

identify(5, 5) could run before the notebook reached its requested size,
so the pixel fell outside the first tab and returned ''.  Guard it with a
new opt-in wait_until_mapped(full_size=True).

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
13 days agogh-151757: Skip curses variation-selector test on older macOS (GH-153344)
Serhiy Storchaka [Thu, 9 Jul 2026 06:28:32 +0000 (09:28 +0300)] 
gh-151757: Skip curses variation-selector test on older macOS (GH-153344)

Older macOS reports a variation selector as a spacing character (wcwidth()
returns 1) instead of a zero-width combining mark, so curses cannot put it in
the same cell as its base.

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
13 days agogh-153380: Adapt ttk tests to Tk 9.1 changes (GH-153381)
Serhiy Storchaka [Thu, 9 Jul 2026 06:07:34 +0000 (09:07 +0300)] 
gh-153380: Adapt ttk tests to Tk 9.1 changes (GH-153381)

- selectmode: new 'single' and 'multiple' modes
- set(): includes the tree column '#0'
- insert()/selection(): item ids keep the passed value's type
- heading_callback: avoid a double click on the second click
- theme_create_image: don't assert the exact (now scaled) width

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
13 days agogh-130578: clarify urllib.parse.quote parameter defaults behavior (#130598)
Kanishk Pachauri [Thu, 9 Jul 2026 05:11:36 +0000 (10:41 +0530)] 
gh-130578: clarify urllib.parse.quote parameter defaults behavior (#130598)

Co-authored-by: Serhiy Storchaka <storchaka@gmail.com>
13 days agogh-47005: fix do_open() to let regular headers override unredirected … (#146506)
CHINMAY [Thu, 9 Jul 2026 03:40:18 +0000 (20:40 -0700)] 
gh-47005: fix do_open() to let regular headers override unredirected … (#146506)

* gh-47005: fix do_open() to let regular headers override unredirected headers

AbstractHTTPHandler.do_open() was building the request header dict by
starting with unredirected_hdrs and only inserting regular headers that
were not already present, giving unredirected headers priority.  This
contradicts get_header() and header_items(), both of which give regular
headers the higher priority.

Fix by unconditionally updating with req.headers so that a header set
via add_header() always overrides one set via add_unredirected_header().

---------

Co-authored-by: Senthil Kumaran <senthil@python.org>
13 days agogh-151213: Document asyncio debugging tools (#151392)
Bartosz Sławecki [Wed, 8 Jul 2026 23:17:05 +0000 (01:17 +0200)] 
gh-151213: Document asyncio debugging tools (#151392)

13 days agogh-119592: gh-152967: Fix ProcessPoolExecutor stranding submitted work when a max_tas...
Gregory P. Smith [Wed, 8 Jul 2026 20:06:20 +0000 (13:06 -0700)] 
gh-119592: gh-152967: Fix ProcessPoolExecutor stranding submitted work when a max_tasks_per_child worker exits (GH-152978)

gh-119592: Fix ProcessPoolExecutor stranding submitted work when a max_tasks_per_child worker exits

Worker replacement went through the executor object: the manager thread
read executor attributes that shutdown(wait=False) clears concurrently,
and could not replace workers at all once the executor was garbage
collected. A worker exiting at its max_tasks_per_child limit in those
states left the remaining submitted work permanently unexecuted and hung
interpreter exit; the racing case could crash the manager thread.

Replace workers from the executor manager thread using its own state
plus configuration read through the live executor weakref, which
shutdown() never clears:

- After shutdown(wait=False) with the executor still referenced, a
  replacement is spawned and the remaining work is executed as
  documented.
- Once the executor has been garbage collected (gh-152967), or a
  replacement worker cannot be started and no workers remain, the
  remaining futures now fail with BrokenProcessPool instead of never
  resolving.
- A new _force_shutting_down flag stops both spawn paths from starting
  workers that would escape terminate_workers()/kill_workers().

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
Reviewed-multiple-times-by: Gregory P. Smith
13 days agogh-61310: Document package precedence over same-named modules (#153211)
YiYi [Wed, 8 Jul 2026 17:28:30 +0000 (01:28 +0800)] 
gh-61310: Document package precedence over same-named modules (#153211)

Co-authored-by: TheD0ubleC <cc@scmd.cc>
Co-authored-by: Stan Ulbrych <stan@python.org>
13 days agogh-153300: Set global configuration variables in PyConfig_Set() (#153301)
Victor Stinner [Wed, 8 Jul 2026 16:56:07 +0000 (18:56 +0200)] 
gh-153300: Set global configuration variables in PyConfig_Set() (#153301)

PyConfig_Set() now also set global configuration variables. For
example, PyConfig_Set("inspect", value) now also sets Py_InspectFlag.

Use PyConfig_Set() in main.c to set the inspect flag. Python code can
now see the updated sys.flags.inspect value.

13 days agogh-46927: Prevent readline from overriding environment variables (GH-153184)
Mathieu Dubois-Briand [Wed, 8 Jul 2026 16:45:15 +0000 (18:45 +0200)] 
gh-46927: Prevent readline from overriding environment variables (GH-153184)

Readline sets COLUMNS and LINES to the terminal size at initialization, and
ncurses prefers those stale variables over an ioctl() query, breaking resize
handling. Set rl_change_environment to 0, when available, so importing
readline no longer modifies the environment.

Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
13 days agogh-143750: Compile OpenSSL with TSan for TSan CI (#153316)
Kumar Aditya [Wed, 8 Jul 2026 15:15:31 +0000 (20:45 +0530)] 
gh-143750: Compile OpenSSL with TSan for TSan CI (#153316)

Co-authored-by: Sam Gross <colesbury@gmail.com>
13 days agogh-153298: Fix data race in `GenericAlias` parameter initialization on FT (#153318)
sobolevn [Wed, 8 Jul 2026 14:57:22 +0000 (17:57 +0300)] 
gh-153298: Fix data race in `GenericAlias` parameter initialization on FT (#153318)

13 days agogh-153339: Remove extra skip in `test_generators.py` (#153340)
sobolevn [Wed, 8 Jul 2026 14:13:24 +0000 (17:13 +0300)] 
gh-153339: Remove extra skip in `test_generators.py` (#153340)

13 days agobpo-42861: Add `next_network` function to ipaddress library (#24180)
777GE90 [Wed, 8 Jul 2026 14:06:55 +0000 (15:06 +0100)] 
bpo-42861: Add `next_network` function to ipaddress library (#24180)

Add `next_network` method that returns the  next closest network of prefix size *next_prefix*. to Lib.ipaddress

---------

Co-authored-by: Faisal Mahmood <killerklient@hotmail.co.uk>
Co-authored-by: Senthil Kumaran <senthil@python.org>
13 days agogh-153083: Defer GC tracking of an array to the end of construction. (gh-153284)
Donghee Na [Wed, 8 Jul 2026 12:27:32 +0000 (21:27 +0900)] 
gh-153083: Defer GC tracking of an array to the end of construction. (gh-153284)

13 days agogh-143990: Do not assume the requested font size in test_font (GH-153322)
Serhiy Storchaka [Wed, 8 Jul 2026 12:12:20 +0000 (15:12 +0300)] 
gh-143990: Do not assume the requested font size in test_font (GH-153322)

On platforms with only bitmap fonts Tk substitutes the nearest
available size, so query the resolved size instead of assuming it.

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
13 days agogh-86017: Document displayof for Font.actual() and clarify Font.copy() (GH-153288)
Serhiy Storchaka [Wed, 8 Jul 2026 12:00:15 +0000 (15:00 +0300)] 
gh-86017: Document displayof for Font.actual() and clarify Font.copy() (GH-153288)

Explain the displayof argument of tkinter.font.Font.actual() and describe
what Font.copy() returns, including for a wrapped font description.

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
13 days agogh-66788: Add the utf-7-imap codec (GH-153149)
Serhiy Storchaka [Wed, 8 Jul 2026 11:55:27 +0000 (14:55 +0300)] 
gh-66788: Add the utf-7-imap codec (GH-153149)

Implement the modified UTF-7 encoding used for international IMAP4
mailbox names (RFC 3501, section 5.1.3).  It differs from UTF-7:
"&" is the shift character ("&-" encodes a literal "&"), "," replaces
"/" in the Base64 alphabet, and all non-printable-ASCII characters are
Base64-encoded.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
13 days agogh-150191: temporarily skip two openssl tests which have internal data races under...
Kumar Aditya [Wed, 8 Jul 2026 11:48:38 +0000 (17:18 +0530)] 
gh-150191: temporarily skip two openssl tests which have internal data races under TSAN (#153320)

13 days agogh-88574: Do not swallow the line after a terminating literal in imaplib (GH-153317)
Serhiy Storchaka [Wed, 8 Jul 2026 11:45:58 +0000 (14:45 +0300)] 
gh-88574: Do not swallow the line after a terminating literal in imaplib (GH-153317)

GH-152751 skipped a spurious blank line after a literal unconditionally,
corrupting a response that ends with a literal (such as a mailbox name
returned by LIST): its empty trailer was mistaken for the blank and the
following line was swallowed.

The blank is now skipped only inside an unclosed parenthesis.  After a
literal that ends the response it instead arrives before the next
response and is skipped there.

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
13 days agogh-149740: Remove redundant self.empty() check in asyncio.Queue.get() (#149741)
Timofei [Wed, 8 Jul 2026 10:15:18 +0000 (13:15 +0300)] 
gh-149740: Remove redundant self.empty() check in asyncio.Queue.get() (#149741)

Co-authored-by: Kumar Aditya <kumaraditya@python.org>
13 days agogh-153292: Fix data race in `threading.RLock.__repr__` in FT builds (#153299)
sobolevn [Wed, 8 Jul 2026 09:41:38 +0000 (12:41 +0300)] 
gh-153292: Fix data race in `threading.RLock.__repr__` in FT builds (#153299)

2 weeks agogh-150358: Remove unused _complete_fut attribute from asyncio.StreamWriter (#150359)
Timofei [Wed, 8 Jul 2026 09:24:24 +0000 (12:24 +0300)] 
gh-150358: Remove unused _complete_fut attribute from asyncio.StreamWriter (#150359)

2 weeks agogh-143990: Preserve the size when creating a Font from a named font (GH-153267)
Serhiy Storchaka [Wed, 8 Jul 2026 08:30:12 +0000 (11:30 +0300)] 
gh-143990: Preserve the size when creating a Font from a named font (GH-153267)

tkinter.font.Font now copies the options of a named font (via "font
configure") instead of the options resolved by "font actual", which
would resolve a size specified in pixels (a negative size) to points.
A font description is still resolved, as it cannot be parsed otherwise.

Font.copy(), which has always been equivalent to constructing a Font
from the original font, is updated to match and now preserves the size
too.

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
2 weeks agogh-153056: Fix a data race compiling the string.Template pattern in free-threading...
tonghuaroot (童话) [Wed, 8 Jul 2026 00:40:54 +0000 (08:40 +0800)] 
gh-153056: Fix a data race compiling the string.Template pattern in free-threading builds (#153057)

* gh-153056: Fix a data race compiling the string.Template pattern in free-threading builds

Template compiles its substitution pattern lazily and caches it on the class.  On the free-threaded build two concurrent first uses could race: a thread that observed the pattern another thread had just compiled would try to recompile it, and re.compile() rejects flags on an already-compiled pattern, raising a spurious ValueError.  Return the already-compiled pattern instead.

As a side effect, a subclass that supplies an already-compiled pattern now works too; previously it raised the same ValueError at class definition.

* Trim test comments and NEWS wording

* Document that the pattern attribute accepts a string or a compiled regex

* Comment the three states of pattern and note the documented-behavior fix in NEWS

* Update Doc/library/string.rst

---------

Co-authored-by: Barry Warsaw <barry@python.org>
2 weeks agogh-152586: Make `tempfile.TemporaryFileWrapper` public (#152646)
Aniket [Tue, 7 Jul 2026 21:34:16 +0000 (03:04 +0530)] 
gh-152586: Make `tempfile.TemporaryFileWrapper` public (#152646)

Deprecate old compatibility `tempfile._TemporaryFileWrapper` name, schedule it to be removed in Python 3.21

Co-authored-by: sobolevn <mail@sobolevn.me>
Co-authored-by: blurb-it[bot] <43283697+blurb-it[bot]@users.noreply.github.com>
2 weeks agogh-153133: Fix socket leak in error path of asyncio create_connection (#153134)
Timofei [Tue, 7 Jul 2026 19:23:17 +0000 (22:23 +0300)] 
gh-153133: Fix socket leak in error path of asyncio create_connection (#153134)

2 weeks agogh-153252: Fix error handling in `_PyCompile_CodeGen` (#153253)
sobolevn [Tue, 7 Jul 2026 16:41:05 +0000 (19:41 +0300)] 
gh-153252: Fix error handling in `_PyCompile_CodeGen` (#153253)

2 weeks agoOn Unix, avoid a conversion to wchar_t in getpath.c (#153230)
Victor Stinner [Tue, 7 Jul 2026 16:28:05 +0000 (18:28 +0200)] 
On Unix, avoid a conversion to wchar_t in getpath.c (#153230)

* Replace _Py_wstat() with _Py_stat() in isfile() and isxfile().
* Replace _Py_wfopen() with Py_fopen() in readlines().

2 weeks agogh-152798: update sys.thread_info.lock documentation to match implementation (#153263)
Wojusensei [Tue, 7 Jul 2026 16:13:35 +0000 (00:13 +0800)] 
gh-152798: update sys.thread_info.lock documentation to match implementation (#153263)

2 weeks agogh-153210: Restore accidental `array.ArrayType` removal (#153261)
sobolevn [Tue, 7 Jul 2026 16:09:02 +0000 (19:09 +0300)] 
gh-153210: Restore accidental `array.ArrayType` removal (#153261)

2 weeks agogh-153205: Set `MemoryError` in stackref array conversion (#153206)
An Long [Tue, 7 Jul 2026 16:03:34 +0000 (01:03 +0900)] 
gh-153205: Set `MemoryError` in stackref array conversion (#153206)

Co-authored-by: blurb-it[bot] <43283697+blurb-it[bot]@users.noreply.github.com>
Co-authored-by: sobolevn <mail@sobolevn.me>
2 weeks agogh-150191: Avoid data race in test_sni_callback_race (#150193)
Sam Gross [Tue, 7 Jul 2026 15:52:00 +0000 (11:52 -0400)] 
gh-150191: Avoid data race in test_sni_callback_race (#150193)

2 weeks agoTemporarily allow the TSan (free-threading) job to fail (#152173)
Stan Ulbrych [Tue, 7 Jul 2026 14:59:57 +0000 (16:59 +0200)] 
Temporarily allow the TSan (free-threading) job to fail (#152173)

2 weeks agogh-143990: Allow tkinter.font.Font to wrap a font description (GH-152025)
Serhiy Storchaka [Tue, 7 Jul 2026 09:37:04 +0000 (12:37 +0300)] 
gh-143990: Allow tkinter.font.Font to wrap a font description (GH-152025)

With exists=True and no name, Font now wraps the font description as is,
without creating a new named font, so that it is used without loss of
precision by actual(), measure() and metrics().  Its name attribute is then
the description rather than a string.  Keyword options now override the
corresponding settings of the given font instead of being ignored.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2 weeks agogh-153227: Use a development version of Pygments (#153228)
Peter Bierma [Tue, 7 Jul 2026 09:08:34 +0000 (05:08 -0400)] 
gh-153227: Use a development version of Pygments (#153228)

2 weeks agogh-153210: Fix `array` module import crash under a memory pressure (#153238)
sobolevn [Tue, 7 Jul 2026 08:25:30 +0000 (11:25 +0300)] 
gh-153210: Fix `array` module import crash under a memory pressure (#153238)

2 weeks agogh-145177: Update Emscripten to 6.0.2 (#153235)
Hood Chatham [Mon, 6 Jul 2026 23:54:16 +0000 (16:54 -0700)] 
gh-145177: Update Emscripten to 6.0.2 (#153235)

Removes the patched `getentropy()` call since it has been included upstream.

2 weeks agogh-140729: Fix the cProfile module when the executed script contains calls to multipr...
Duprat [Mon, 6 Jul 2026 23:39:06 +0000 (01:39 +0200)] 
gh-140729: Fix the cProfile module when the executed script contains calls to multiprocessing.Process (#144715)

Co-authored-by: Pablo Galindo Salgado <pablogsal@gmail.com>
2 weeks agogh-153068: Preserve cprofile enable errors (#153070)
An Long [Mon, 6 Jul 2026 23:36:26 +0000 (08:36 +0900)] 
gh-153068: Preserve cprofile enable errors (#153070)

2 weeks agogh-150479: reject CR and LF in email.utils.formataddr (#150480)
metsw24-max [Mon, 6 Jul 2026 20:16:43 +0000 (01:46 +0530)] 
gh-150479: reject CR and LF in email.utils.formataddr (#150480)

2 weeks agogh-150490: Raise PyType_Modified for insertion into split dictionary (take 2) (#152914)
Dino Viehland [Mon, 6 Jul 2026 19:56:32 +0000 (12:56 -0700)] 
gh-150490: Raise PyType_Modified for insertion into split dictionary (take 2) (#152914)

Raise PyType_Modified for insertion into split dictionary - now with better locking discipline in free-threaded builds

2 weeks agoSeed pip caches for docs on `main`/`3.x` branches (#153127)
Hugo van Kemenade [Mon, 6 Jul 2026 19:46:22 +0000 (22:46 +0300)] 
Seed pip caches for docs on `main`/`3.x` branches (#153127)

2 weeks agogh-153200: Fix math.isqrt() for int subclasses with overridden comparison operators...
Serhiy Storchaka [Mon, 6 Jul 2026 18:10:23 +0000 (21:10 +0300)] 
gh-153200: Fix math.isqrt() for int subclasses with overridden comparison operators (GH-153203)

The final check-and-correct comparison in the arbitrary precision path
could call a comparison operator overridden in an int subclass.
Compare by value with int's tp_richcompare.

2 weeks agogh-150880: Clarify DirEntry.path construction semantics (GH-153218)
Zain Nadeem [Mon, 6 Jul 2026 17:14:28 +0000 (22:14 +0500)] 
gh-150880: Clarify DirEntry.path construction semantics (GH-153218)

2 weeks agogh-152433: Allow _ssl module to build for Windows UWP API set (GH-152791)
thexai [Mon, 6 Jul 2026 16:32:36 +0000 (18:32 +0200)] 
gh-152433: Allow _ssl module to build for Windows UWP API set (GH-152791)

2 weeks agogh-152785: Upgrade Ubuntu from 24.04 to 26.04 in GitHub Actions (#152717)
Victor Stinner [Mon, 6 Jul 2026 15:44:21 +0000 (17:44 +0200)] 
gh-152785: Upgrade Ubuntu from 24.04 to 26.04 in GitHub Actions (#152717)

* Replace "ubuntu-24.04" with "ubuntu-26.04".
* Replace "ubuntu-latest" with "ubuntu-26.04" for Cross build Linux.
* Replace "ubuntu-latest" with "ubuntu-slim" for small workloads.
* Update ".github/actionlint.yaml" to allow "ubuntu-26.04"
  and "ubuntu-26.04-arm" images.
* Install Ubuntu libmpdec-dev package, rather than installing libmpdec
  from source (tarball).
* No longer run https://apt.llvm.org/llvm.sh, since ubuntu-26.04 provides
  clang-21 by default.

Co-authored-by: Hugo van Kemenade <1324225+hugovk@users.noreply.github.com>
2 weeks agoDocs: Link to the `contextlib.chdir` context manager in the `os.chdir` docs (#153152)
Cody Maloney [Mon, 6 Jul 2026 15:40:45 +0000 (08:40 -0700)] 
Docs: Link to the `contextlib.chdir` context manager in the `os.chdir` docs (#153152)

Co-authored-by: Stan Ulbrych <stan@python.org>
2 weeks agogh-152849: Fix `OverflowError` message for out-of-range timestamps in the `time`...
tonghuaroot (童话) [Mon, 6 Jul 2026 14:50:15 +0000 (22:50 +0800)] 
gh-152849: Fix `OverflowError` message for out-of-range timestamps in the `time` module (#152850)

Co-authored-by: Stan Ulbrych <stan@python.org>
2 weeks agogh-148286: Remove invalid test in `test_decodeescape` (#153187)
Stan Ulbrych [Mon, 6 Jul 2026 14:47:55 +0000 (16:47 +0200)] 
gh-148286: Remove invalid test in `test_decodeescape` (#153187)

2 weeks agogh-145098: Use `macos-26[-intel]` instead of `macos-15[-intel]` in `{jit,tail-call...
Hugo van Kemenade [Mon, 6 Jul 2026 14:25:32 +0000 (17:25 +0300)] 
gh-145098: Use `macos-26[-intel]` instead of `macos-15[-intel]` in `{jit,tail-call}.yml` (#153194)

2 weeks agoAdd tests for syntax error messages that had no test coverage (GH-153192)
Serhiy Storchaka [Mon, 6 Jul 2026 13:49:17 +0000 (16:49 +0300)] 
Add tests for syntax error messages that had no test coverage (GH-153192)

"illegal target for annotation" and "cannot use dict unpacking here"
were not tested at all, and "f-string: expecting '!', or ':', or '}'"
was only tested for its t-string variant.

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2 weeks agogh-153171: Produce specialized syntax errors for 'not' after an operator in more...
Serhiy Storchaka [Mon, 6 Jul 2026 10:00:05 +0000 (13:00 +0300)] 
gh-153171: Produce specialized syntax errors for 'not' after an operator in more positions (GH-153175)

Move the invalid_factor rule from term to factor and the
invalid_arithmetic rule from shift_expr to sum, so that they are
tried at all positions where the invalid construct can occur.
Previously a generic "invalid syntax" error was reported for
constructs like "1 << 2 + not x" or "1 * + not x".

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2 weeks agogh-145855: Remove the DELETE_ATTR bytecode (GH-146216)
An Long [Mon, 6 Jul 2026 09:41:47 +0000 (18:41 +0900)] 
gh-145855: Remove the DELETE_ATTR bytecode (GH-146216)

2 weeks agogh-148765: Fix description of *include_hidden* parameter in `glob.glob()` docs (...
Eoin Shaughnessy [Mon, 6 Jul 2026 07:44:48 +0000 (08:44 +0100)] 
gh-148765: Fix description of *include_hidden* parameter in `glob.glob()` docs (#148768)

Co-authored-by: Stan Ulbrych <stan@python.org>
2 weeks agogh-152026: Track mark-saving contexts with a counter (GH-153160)
Serhiy Storchaka [Mon, 6 Jul 2026 05:22:37 +0000 (08:22 +0300)] 
gh-152026: Track mark-saving contexts with a counter (GH-153160)

Replace the state->repeat checks that guard mark saving and restoring
with an explicit save_marks counter maintained where repeat and
possessive contexts are entered and left.  Each push is now paired
with a pop decided by the same condition.

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
2 weeks agogh-82151: Update and improve urlsplit docstring (#152985)
Pranav Choudhary [Mon, 6 Jul 2026 03:19:17 +0000 (08:49 +0530)] 
gh-82151: Update and improve urlsplit docstring (#152985)

* gh-82151: Update and improve urlsplit docstring

Co-authored-by: idomic <michael.ido@gmail.com>
* Slightly improved wording for % escapes. Consistent with rest of the doc.

---------

Co-authored-by: idomic <michael.ido@gmail.com>
Co-authored-by: Senthil Kumaran <senthil@python.org>
2 weeks agogh-151640: Track sharing of `BytesIO` buffer in free-threaded builds (GH-151651)
Pedram Karimi [Mon, 6 Jul 2026 01:17:20 +0000 (04:47 +0330)] 
gh-151640: Track sharing of `BytesIO` buffer in free-threaded builds (GH-151651)

2 weeks agobpo-45706: Add imaplib.IMAP4.login_plain (GH-29398)
Przemysław Buczkowski [Sun, 5 Jul 2026 21:26:44 +0000 (06:26 +0900)] 
bpo-45706: Add imaplib.IMAP4.login_plain (GH-29398)

Adds authentication using PLAIN SASL mechanism.

This is a plain-text authentication mechanism which can be used
instead of IMAP4.login() when UTF-8 support is required.

2 weeks agogh-153141: Fix mutable default argument in _SharedMemoryTracker.__init__ (GH-153142)
Vineet Kumar [Sun, 5 Jul 2026 19:02:58 +0000 (00:32 +0530)] 
gh-153141: Fix mutable default argument in _SharedMemoryTracker.__init__ (GH-153142)

Fix default argument for segment_names in _SharedMemoryTracker constructor to not use a mutable list.

2 weeks agogh-98092: Add imaplib.IMAP4.id method (GH-153136)
Serhiy Storchaka [Sun, 5 Jul 2026 17:32:32 +0000 (20:32 +0300)] 
gh-98092: Add imaplib.IMAP4.id method (GH-153136)

Add a wrapper for the IMAP ID command (RFC 2971).  It takes a mapping
of field names to values and returns the server identification
information from the untagged ID response.

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2 weeks agogh-105708: 'V' could be case insensitive for IPvFuture hostnames (#105709)
Chandra [Sun, 5 Jul 2026 17:08:28 +0000 (12:08 -0500)] 
gh-105708: 'V' could be case insensitive for IPvFuture hostnames (#105709)

* gh-105708: 'V' could be case insensitive for IPvFuture hostnames

* gh-105708: 'V' could be case insensitive for IPvFuture hostnames & checking empty hostnames

* 📜🤖 Added by blurb_it.

* Fix the Merge changing \z to \Z.

* Fixed the News Entry.

* Fix the lint.

---------

Co-authored-by: blurb-it[bot] <43283697+blurb-it[bot]@users.noreply.github.com>
Co-authored-by: Senthil Kumaran <senthil@python.org>
2 weeks agogh-77508: Add imaplib.IMAP4.move method (GH-153121)
Serhiy Storchaka [Sun, 5 Jul 2026 17:02:45 +0000 (20:02 +0300)] 
gh-77508: Add imaplib.IMAP4.move method (GH-153121)

Add a wrapper for the IMAP MOVE command (RFC 6851), analogous to
copy().  The arguments of MOVE and UID MOVE are quoted when necessary.

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2 weeks agogh-60055: Allow passing a Request instance for RobotParser URLs (#103753)
Andrew James [Sun, 5 Jul 2026 15:57:09 +0000 (01:57 +1000)] 
gh-60055: Allow passing a Request instance for RobotParser URLs (#103753)

* Allow passing a Request instance for the url parameter

* 📜🤖 Added by blurb_it.

* gh-60055: rebase onto main and document Request support

* Fix the lint warnings and CI Errors.

* Addressing RobotFileParser review comments.

* Update Doc/library/urllib.robotparser.rst

Co-authored-by: Serhiy Storchaka <storchaka@gmail.com>
* Address additional Review Comments.

* Changed versionchanged to next.

---------

Co-authored-by: blurb-it[bot] <43283697+blurb-it[bot]@users.noreply.github.com>
Co-authored-by: Senthil Kumaran <senthil@python.org>
Co-authored-by: Serhiy Storchaka <storchaka@gmail.com>
2 weeks agogh-143921: Narrow the control character check in imaplib commands (GH-153067)
Serhiy Storchaka [Sun, 5 Jul 2026 15:25:36 +0000 (18:25 +0300)] 
gh-143921: Narrow the control character check in imaplib commands (GH-153067)

Only NUL, CR and LF are rejected now.  Other control characters are
valid in quoted strings and can occur in mailbox names returned by
the server, so they are now accepted and sent quoted.

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2 weeks agogh-68147: Fix RFC references in imaplib (GH-153126)
Serhiy Storchaka [Sun, 5 Jul 2026 14:08:58 +0000 (17:08 +0300)] 
gh-68147: Fix RFC references in imaplib (GH-153126)

Refer to RFC 3501, which obsoleted RFC 2060.  "]" is disallowed
in flags, not in tags.

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2 weeks agogh-149930: Clarify imaplib response helper types (GH-149963)
Savage Mechanic [Sun, 5 Jul 2026 13:16:54 +0000 (14:16 +0100)] 
gh-149930: Clarify imaplib response helper types (GH-149963)

Co-authored-by: savagemechanic <20458938+savagemechanic@users.noreply.github.com>
2 weeks agogh-43699: Defer the drag cursor in tkinter.dnd until the pointer moves (GH-152371)
Serhiy Storchaka [Sun, 5 Jul 2026 12:49:43 +0000 (15:49 +0300)] 
gh-43699: Defer the drag cursor in tkinter.dnd until the pointer moves (GH-152371)

The drag cursor is now changed only once the pointer starts moving past
a small threshold rather than on the initial button press, so that a
plain click no longer flashes it.

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
2 weeks agogh-59396: Use themed widgets in tkinter.filedialog (GH-152036)
Serhiy Storchaka [Sun, 5 Jul 2026 11:27:51 +0000 (14:27 +0300)] 
gh-59396: Use themed widgets in tkinter.filedialog (GH-152036)

The FileDialog, LoadFileDialog and SaveFileDialog dialogs are now built from
the themed tkinter.ttk widgets by default instead of the classic tkinter
widgets, and gained a use_ttk parameter that selects between the classic Tk
widgets and the themed ttk widgets.

They were also brought closer to the native Tk file dialog: the buttons and
field labels gained Alt key accelerators, the default ring follows the keyboard
focus, the Escape key cancels the dialog, the focus traverses the widgets in
their visual order, and the directory and file lists gained a horizontal
scrollbar and type-ahead selection.  The dialog is now transient and centered
over its parent, and the SaveFileDialog overwrite confirmation uses a themed
message box.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2 weeks agoFix minor typos in 'Compound statements' docs (#149666)
Jonathan Dung [Sun, 5 Jul 2026 11:23:49 +0000 (19:23 +0800)] 
Fix minor typos in 'Compound statements' docs (#149666)

Co-authored-by: Stan Ulbrych <stan@python.org>
2 weeks agogh-76607: Improve documentation of the tkinter cursor option (GH-152481)
Serhiy Storchaka [Sun, 5 Jul 2026 11:13:52 +0000 (14:13 +0300)] 
gh-76607: Improve documentation of the tkinter cursor option (GH-152481)

Link to the cursors(3tk) manual page for the available cursor names and
note that they are platform-dependent.

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
2 weeks agogh-79638: Restore "Treat an unreachable robots.txt as disallow all" (GH-152525)
Serhiy Storchaka [Sun, 5 Jul 2026 11:13:01 +0000 (14:13 +0300)] 
gh-79638: Restore "Treat an unreachable robots.txt as disallow all" (GH-152525)

This change was accidentally reverted by f0daba1652c (gh-106693,
GH-149514), which only intended to revert the ob_sval change.
The tests were already restored by GH-149569.

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2 weeks agogh-75952: Document negative offsets in tkinter geometry strings (GH-152531)
Serhiy Storchaka [Sun, 5 Jul 2026 11:07:49 +0000 (14:07 +0300)] 
gh-75952: Document negative offsets in tkinter geometry strings (GH-152531)

Tk geometry strings can contain a negative offset (e.g. 200x100+-9+-8)
when a window edge is positioned beyond the corresponding screen edge.
Note this in the geometry() and winfo_geometry() documentation.

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
2 weeks agogh-123193: Document tkinter Variable lifetime (GH-152625)
Serhiy Storchaka [Sun, 5 Jul 2026 11:07:12 +0000 (14:07 +0300)] 
gh-123193: Document tkinter Variable lifetime (GH-152625)

A Tk variable wrapper unsets its Tcl variable when garbage collected, so a
reference must be kept while a widget uses it.  Otherwise Tk recreates the Tcl
variable but never unsets it again, leaking it.

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
2 weeks ago Fix various typos in the `http.server` docs (#153089)
Bénédikt Tran [Sun, 5 Jul 2026 10:58:53 +0000 (12:58 +0200)] 
 Fix various typos in the `http.server` docs (#153089)

2 weeks agogh-69134: Wait until mapped in SimpleDialog keyboard tests (GH-152690)
Serhiy Storchaka [Sun, 5 Jul 2026 10:51:34 +0000 (13:51 +0300)] 
gh-69134: Wait until mapped in SimpleDialog keyboard tests (GH-152690)

The SimpleDialog keyboard tests generate key events after focus_force(),
which on Windows are dropped until the toplevel is mapped, so they could
fail intermittently (seen on the Windows10 buildbot as
test_return_no_default).  Wait until the window is mapped in each of
these tests, as GH-152599 did for the other keyboard tests.

Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
2 weeks agogh-54930: Send a status line in error responses to malformed request lines (GH-152980)
Serhiy Storchaka [Sun, 5 Jul 2026 10:48:23 +0000 (13:48 +0300)] 
gh-54930: Send a status line in error responses to malformed request lines (GH-152980)

Previously such error responses were sent in the bare HTTP/0.9 style,
without a status line and headers.

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2 weeks agogh-151945: fix Sphinx reference warnings in `http.server` docs (#153084)
Bénédikt Tran [Sun, 5 Jul 2026 09:25:30 +0000 (11:25 +0200)] 
gh-151945: fix Sphinx reference warnings in `http.server` docs (#153084)