Mailman 3 August 2025 - Python-checkins

gh-138130: Fix return value of libc_ver() on Emscripten (#138132)
by freakboy3742 Aug. 31, 2025

Aug. 31, 2025

https://github.com/python/cpython/commit/11a5fc82381308f0b6060b8c06af89fd19… commit: 11a5fc82381308f0b6060b8c06af89fd197516e1 branch: main author: Hood Chatham <roberthoodchatham(a)gmail.com> committer: freakboy3742 <russell(a)keith-magee.com> date: 2025-09-01T06:50:29+08:00 summary: gh-138130: Fix return value of libc_ver() on Emscripten (#138132) Emscripten's libc is a hybrid of musl and llvm libc; but it reports that it is "glibc". This modifies the return value of `platform.libc_ver()` to return something that is Emscripten-specific. files: M Lib/platform.py M Lib/test/test_platform.py diff --git a/Lib/platform.py b/Lib/platform.py index da15bb4717bb96..762f5d06099cf0 100644 --- a/Lib/platform.py +++ b/Lib/platform.py @@ -173,6 +173,11 @@ def libc_ver(executable=None, lib='', version='', chunksize=16384): """ if not executable: + if sys.platform == "emscripten": + # Emscripten's os.confstr reports that it is glibc, so special case + # it. + ver = ".".join(str(x) for x in sys._emscripten_info.emscripten_version) + return ("emscripten", ver) try: ver = os.confstr('CS_GNU_LIBC_VERSION') # parse 'glibc 2.28' as ('glibc', '2.28') diff --git a/Lib/test/test_platform.py b/Lib/test/test_platform.py index 2e26134bae8323..73d4f3cbdb689d 100644 --- a/Lib/test/test_platform.py +++ b/Lib/test/test_platform.py @@ -532,8 +532,10 @@ def test_ios_ver(self): self.assertEqual(override.model, "Whiz") self.assertTrue(override.is_simulator) - @unittest.skipIf(support.is_emscripten, "Does not apply to Emscripten") def test_libc_ver(self): + if support.is_emscripten: + assert platform.libc_ver() == ("emscripten", "4.0.12") + return # check that libc_ver(executable) doesn't raise an exception if os.path.isdir(sys.executable) and \ os.path.exists(sys.executable+'.exe'):

1 0

gh-138286: Run ``ruff`` on ``Tools/i18n`` (#138287)
by AA-Turner Aug. 31, 2025

Aug. 31, 2025

https://github.com/python/cpython/commit/78acd8e95e78bc410b8207cd60e1323ece… commit: 78acd8e95e78bc410b8207cd60e1323ece01b3d5 branch: main author: Stan Ulbrych <89152624+StanFromIreland(a)users.noreply.github.com> committer: AA-Turner <9087854+AA-Turner(a)users.noreply.github.com> date: 2025-08-31T20:29:02Z summary: gh-138286: Run ``ruff`` on ``Tools/i18n`` (#138287) Co-authored-by: Adam Turner <9087854+AA-Turner(a)users.noreply.github.com> Co-authored-by: Tomas R. <tomas.roun8(a)gmail.com> files: A Tools/i18n/.ruff.toml M .pre-commit-config.yaml M Tools/i18n/makelocalealias.py M Tools/i18n/msgfmt.py M Tools/i18n/pygettext.py diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml index d101f5c37e60b1..61db8c456c2e64 100644 --- a/.pre-commit-config.yaml +++ b/.pre-commit-config.yaml @@ -14,6 +14,10 @@ repos: name: Run Ruff (lint) on Tools/build/ args: [--exit-non-zero-on-fix, --config=Tools/build/.ruff.toml] files: ^Tools/build/ + - id: ruff + name: Run Ruff (lint) on Tools/i18n/ + args: [--exit-non-zero-on-fix, --config=Tools/i18n/.ruff.toml] + files: ^Tools/i18n/ - id: ruff name: Run Ruff (lint) on Argument Clinic args: [--exit-non-zero-on-fix, --config=Tools/clinic/.ruff.toml] diff --git a/Tools/i18n/.ruff.toml b/Tools/i18n/.ruff.toml new file mode 100644 index 00000000000000..a8f4f2f5a96d7b --- /dev/null +++ b/Tools/i18n/.ruff.toml @@ -0,0 +1,10 @@ +extend = "../../.ruff.toml" # Inherit the project-wide settings + +target-version = "py313" + +[lint] +select = [ + "F", # pyflakes + "I", # isort + "UP", # pyupgrade +] diff --git a/Tools/i18n/makelocalealias.py b/Tools/i18n/makelocalealias.py index 7f001abc09745d..f825862ffdf350 100755 --- a/Tools/i18n/makelocalealias.py +++ b/Tools/i18n/makelocalealias.py @@ -8,6 +8,7 @@ """ import locale import sys + _locale = locale # Location of the X11 alias file. @@ -100,16 +101,15 @@ def parse_glibc_supported(filename): def pprint(data): items = sorted(data.items()) for k, v in items: - print(' %-40s%a,' % ('%a:' % k, v)) + print(f" {k!a:<40}{v!a},") def print_differences(data, olddata): items = sorted(olddata.items()) for k, v in items: if k not in data: - print('# removed %a' % k) + print(f'# removed {k!a}') elif olddata[k] != data[k]: - print('# updated %a -> %a to %a' % \ - (k, olddata[k], data[k])) + print(f'# updated {k!a} -> {olddata[k]!a} to {data[k]!a}') # Additions are not mentioned def optimize(data): @@ -132,7 +132,7 @@ def check(data): errors = 0 for k, v in data.items(): if locale.normalize(k) != v: - print('ERROR: %a -> %a != %a' % (k, locale.normalize(k), v), + print(f'ERROR: {k!a} -> {locale.normalize(k)!a} != {v!a}', file=sys.stderr) errors += 1 return errors @@ -142,10 +142,10 @@ def check(data): parser = argparse.ArgumentParser() parser.add_argument('--locale-alias', default=LOCALE_ALIAS, help='location of the X11 alias file ' - '(default: %a)' % LOCALE_ALIAS) + f'(default: {LOCALE_ALIAS})') parser.add_argument('--glibc-supported', default=SUPPORTED, help='location of the glibc SUPPORTED locales file ' - '(default: %a)' % SUPPORTED) + f'(default: {SUPPORTED})') args = parser.parse_args() data = locale.locale_alias.copy() diff --git a/Tools/i18n/msgfmt.py b/Tools/i18n/msgfmt.py index 65254a7c375456..3351511bbc8f2b 100755 --- a/Tools/i18n/msgfmt.py +++ b/Tools/i18n/msgfmt.py @@ -24,14 +24,14 @@ Display version information and exit. """ -import os -import sys +import array import ast +import codecs import getopt +import os import struct -import array +import sys from email.parser import HeaderParser -import codecs __version__ = "1.2" @@ -113,7 +113,7 @@ def make(filename, outfile): try: with open(infile, 'rb') as f: lines = f.readlines() - except IOError as msg: + except OSError as msg: print(msg, file=sys.stderr) sys.exit(1) @@ -126,6 +126,7 @@ def make(filename, outfile): sys.exit(1) section = msgctxt = None + msgid = msgstr = b'' fuzzy = 0 # Start off assuming Latin-1, so everything decodes without failure, @@ -177,7 +178,7 @@ def make(filename, outfile): # This is a message with plural forms elif l.startswith('msgid_plural'): if section != ID: - print('msgid_plural not preceded by msgid on %s:%d' % (infile, lno), + print(f'msgid_plural not preceded by msgid on {infile}:{lno}', file=sys.stderr) sys.exit(1) l = l[12:] @@ -188,7 +189,7 @@ def make(filename, outfile): section = STR if l.startswith('msgstr['): if not is_plural: - print('plural without msgid_plural on %s:%d' % (infile, lno), + print(f'plural without msgid_plural on {infile}:{lno}', file=sys.stderr) sys.exit(1) l = l.split(']', 1)[1] @@ -196,7 +197,7 @@ def make(filename, outfile): msgstr += b'\0' # Separator of the various plural forms else: if is_plural: - print('indexed msgstr required for plural on %s:%d' % (infile, lno), + print(f'indexed msgstr required for plural on {infile}:{lno}', file=sys.stderr) sys.exit(1) l = l[6:] @@ -212,8 +213,7 @@ def make(filename, outfile): elif section == STR: msgstr += l.encode(encoding) else: - print('Syntax error on %s:%d' % (infile, lno), \ - 'before:', file=sys.stderr) + print(f'Syntax error on {infile}:{lno} before:', file=sys.stderr) print(l, file=sys.stderr) sys.exit(1) # Add last entry @@ -226,7 +226,7 @@ def make(filename, outfile): try: with open(outfile,"wb") as f: f.write(output) - except IOError as msg: + except OSError as msg: print(msg, file=sys.stderr) diff --git a/Tools/i18n/pygettext.py b/Tools/i18n/pygettext.py index f46b05067d7fde..ddf4474d2bce55 100755 --- a/Tools/i18n/pygettext.py +++ b/Tools/i18n/pygettext.py @@ -193,7 +193,7 @@ def make_escapes(pass_nonascii): escape = escape_ascii else: escape = escape_nonascii - escapes = [r"\%03o" % i for i in range(256)] + escapes = [fr"\{i:03o}" for i in range(256)] for i in range(32, 127): escapes[i] = chr(i) escapes[ord('\\')] = r'\\' @@ -796,7 +796,7 @@ class Options: try: with open(options.excludefilename) as fp: options.toexclude = fp.readlines() - except IOError: + except OSError: print(f"Can't read --exclude-file: {options.excludefilename}", file=sys.stderr) sys.exit(1)

1 0

gh-138013: Make ``test.test_io`` into a package (#138153)
by AA-Turner Aug. 31, 2025

Aug. 31, 2025

https://github.com/python/cpython/commit/15e37ea6b7f7070c112aa3118ba46325c4… commit: 15e37ea6b7f7070c112aa3118ba46325c4cfde2c branch: main author: Cody Maloney <cmaloney(a)users.noreply.github.com> committer: AA-Turner <9087854+AA-Turner(a)users.noreply.github.com> date: 2025-08-31T20:14:19+01:00 summary: gh-138013: Make ``test.test_io`` into a package (#138153) files: A Lib/test/test_io/__init__.py A Lib/test/test_io/__main__.py A Lib/test/test_io/test_general.py D Lib/test/test_io.py M Makefile.pre.in diff --git a/Lib/test/test_io/__init__.py b/Lib/test/test_io/__init__.py new file mode 100644 index 00000000000000..4b16ecc31156a5 --- /dev/null +++ b/Lib/test/test_io/__init__.py @@ -0,0 +1,5 @@ +import os +from test.support import load_package_tests + +def load_tests(*args): + return load_package_tests(os.path.dirname(__file__), *args) diff --git a/Lib/test/test_io/__main__.py b/Lib/test/test_io/__main__.py new file mode 100644 index 00000000000000..40a23a297ec2b4 --- /dev/null +++ b/Lib/test/test_io/__main__.py @@ -0,0 +1,4 @@ +from . import load_tests +import unittest + +unittest.main() diff --git a/Lib/test/test_io.py b/Lib/test/test_io/test_general.py similarity index 99% rename from Lib/test/test_io.py rename to Lib/test/test_io/test_general.py index 92be2763e5ed1e..a56a2c0157f764 100644 --- a/Lib/test/test_io.py +++ b/Lib/test/test_io/test_general.py @@ -5,7 +5,7 @@ # * test_memoryio - tests BytesIO and StringIO # * test_fileio - tests FileIO # * test_file - tests the file interface -# * test_io - tests everything else in the io module +# * test_io.test_general - tests everything else in the io module # * test_univnewlines - tests universal newline support # * test_largefile - tests operations on a file greater than 2**32 bytes # (only enabled with -ulargefile) diff --git a/Makefile.pre.in b/Makefile.pre.in index 9ce6ec65f142d8..eb07f66f14ffc5 100644 --- a/Makefile.pre.in +++ b/Makefile.pre.in @@ -2670,6 +2670,7 @@ TESTSUBDIRS= idlelib/idle_test \ test/test_importlib/source \ test/test_inspect \ test/test_interpreters \ + test/test_io \ test/test_json \ test/test_module \ test/test_multiprocessing_fork \

1 0

[3.13] gh-138295: Fix a grammar issue in the descriptor HOWTO (GH-138296) (GH-138300)
by ZeroIntensity Aug. 31, 2025

Aug. 31, 2025

https://github.com/python/cpython/commit/aebf072a52b396e32f530dbaeab7c41554… commit: aebf072a52b396e32f530dbaeab7c4155425eaef branch: 3.13 author: Miss Islington (bot) <31488909+miss-islington(a)users.noreply.github.com> committer: ZeroIntensity <zintensitydev(a)gmail.com> date: 2025-08-31T15:59:22Z summary: [3.13] gh-138295: Fix a grammar issue in the descriptor HOWTO (GH-138296) (GH-138300) gh-138295: Fix a grammar issue in the descriptor HOWTO (GH-138296) "an str" -> "a str" (cherry picked from commit 552cf86b83e8aafebcfe9abbc77101bdbfda14c8) Co-authored-by: Fangyi Zhou <me(a)fangyi.io> files: M Doc/howto/descriptor.rst diff --git a/Doc/howto/descriptor.rst b/Doc/howto/descriptor.rst index 50a368b7712938..ba6dd4e0505f53 100644 --- a/Doc/howto/descriptor.rst +++ b/Doc/howto/descriptor.rst @@ -420,7 +420,7 @@ Here are three practical data validation utilities: def validate(self, value): if not isinstance(value, str): - raise TypeError(f'Expected {value!r} to be an str') + raise TypeError(f'Expected {value!r} to be a str') if self.minsize is not None and len(value) < self.minsize: raise ValueError( f'Expected {value!r} to be no smaller than {self.minsize!r}'

1 0

gh-138295: Fix a grammar issue in the descriptor HOWTO (GH-138296)
by ZeroIntensity Aug. 31, 2025

Aug. 31, 2025

https://github.com/python/cpython/commit/552cf86b83e8aafebcfe9abbc77101bdbf… commit: 552cf86b83e8aafebcfe9abbc77101bdbfda14c8 branch: main author: Fangyi Zhou <me(a)fangyi.io> committer: ZeroIntensity <zintensitydev(a)gmail.com> date: 2025-08-31T11:50:23-04:00 summary: gh-138295: Fix a grammar issue in the descriptor HOWTO (GH-138296) "an str" -> "a str" files: M Doc/howto/descriptor.rst diff --git a/Doc/howto/descriptor.rst b/Doc/howto/descriptor.rst index f6c3e473f1c36d..9d5a9ac8b718cb 100644 --- a/Doc/howto/descriptor.rst +++ b/Doc/howto/descriptor.rst @@ -420,7 +420,7 @@ Here are three practical data validation utilities: def validate(self, value): if not isinstance(value, str): - raise TypeError(f'Expected {value!r} to be an str') + raise TypeError(f'Expected {value!r} to be a str') if self.minsize is not None and len(value) < self.minsize: raise ValueError( f'Expected {value!r} to be no smaller than {self.minsize!r}'

1 0

gh-138279: Remove unused attr in RuleCollectorVisitor in parser_generator.py (#138280)
by pablogsal Aug. 31, 2025

Aug. 31, 2025

https://github.com/python/cpython/commit/7bca924fdbecb13e9c5fd72568cdd8f771… commit: 7bca924fdbecb13e9c5fd72568cdd8f77101fe07 branch: main author: chemelnucfin <3982092+chemelnucfin(a)users.noreply.github.com> committer: pablogsal <Pablogsal(a)gmail.com> date: 2025-08-31T14:23:40Z summary: gh-138279: Remove unused attr in RuleCollectorVisitor in parser_generator.py (#138280) files: M Tools/peg_generator/pegen/parser_generator.py diff --git a/Tools/peg_generator/pegen/parser_generator.py b/Tools/peg_generator/pegen/parser_generator.py index 976f5e6e57d7c6..a4360eb86ada50 100644 --- a/Tools/peg_generator/pegen/parser_generator.py +++ b/Tools/peg_generator/pegen/parser_generator.py @@ -44,8 +44,7 @@ class RuleCollectorVisitor(GrammarVisitor): """Visitor that invokes a provided callmaker visitor with just the NamedItem nodes""" - def __init__(self, rules: Dict[str, Rule], callmakervisitor: GrammarVisitor) -> None: - self.rules = rules + def __init__(self, callmakervisitor: GrammarVisitor) -> None: self.callmaker = callmakervisitor def visit_Rule(self, rule: Rule) -> None: @@ -163,7 +162,7 @@ def collect_rules(self) -> None: for rule in self.all_rules.values(): keyword_collector.visit(rule) - rule_collector = RuleCollectorVisitor(self.rules, self.callmakervisitor) + rule_collector = RuleCollectorVisitor(self.callmakervisitor) done: Set[str] = set() while True: computed_rules = list(self.all_rules)

1 0

[3.14] gh-138257: Fix internal `RuleCollectorVisitor` attribute name (GH-138208) (#138259)
by hugovk Aug. 31, 2025

Aug. 31, 2025

https://github.com/python/cpython/commit/5ec6d56405bd7fe365d428b3c1db32feff… commit: 5ec6d56405bd7fe365d428b3c1db32feff260ae8 branch: 3.14 author: Miss Islington (bot) <31488909+miss-islington(a)users.noreply.github.com> committer: hugovk <1324225+hugovk(a)users.noreply.github.com> date: 2025-08-31T09:54:59+03:00 summary: [3.14] gh-138257: Fix internal `RuleCollectorVisitor` attribute name (GH-138208) (#138259) Co-authored-by: chemelnucfin <3982092+chemelnucfin(a)users.noreply.github.com> files: M Tools/peg_generator/pegen/parser_generator.py diff --git a/Tools/peg_generator/pegen/parser_generator.py b/Tools/peg_generator/pegen/parser_generator.py index 7dd56f98a652cc..976f5e6e57d7c6 100644 --- a/Tools/peg_generator/pegen/parser_generator.py +++ b/Tools/peg_generator/pegen/parser_generator.py @@ -45,7 +45,7 @@ class RuleCollectorVisitor(GrammarVisitor): """Visitor that invokes a provided callmaker visitor with just the NamedItem nodes""" def __init__(self, rules: Dict[str, Rule], callmakervisitor: GrammarVisitor) -> None: - self.rulses = rules + self.rules = rules self.callmaker = callmakervisitor def visit_Rule(self, rule: Rule) -> None:

1 0

[3.14] gh-138158: Use the `"data"` tarfile extraction filter in `Tools/ssl/multissltests.py` (GH-138147) (#138262)
by hugovk Aug. 31, 2025

Aug. 31, 2025

https://github.com/python/cpython/commit/b79becefb6d5ca2ec7e12a8c3b9b019ef4… commit: b79becefb6d5ca2ec7e12a8c3b9b019ef410955e branch: 3.14 author: Miss Islington (bot) <31488909+miss-islington(a)users.noreply.github.com> committer: hugovk <1324225+hugovk(a)users.noreply.github.com> date: 2025-08-31T09:53:21+03:00 summary: [3.14] gh-138158: Use the `"data"` tarfile extraction filter in `Tools/ssl/multissltests.py` (GH-138147) (#138262) Co-authored-by: Tommaso Bona <piergeolo(a)gmail.com> files: M Tools/ssl/multissltests.py diff --git a/Tools/ssl/multissltests.py b/Tools/ssl/multissltests.py index 3bede231748e6c..da21e7d97965d1 100755 --- a/Tools/ssl/multissltests.py +++ b/Tools/ssl/multissltests.py @@ -294,7 +294,7 @@ def _unpack_src(self): raise ValueError(member.name, base) member.name = member.name[len(base):].lstrip('/') log.info("Unpacking files to {}".format(self.build_dir)) - tf.extractall(self.build_dir, members) + tf.extractall(self.build_dir, members, filter='data') def _build_src(self, config_args=()): """Now build openssl"""

1 0

[3.14] gh-133829: Remove some specifics from the ``zipimport`` example (GH-133835) (#138274)
by hugovk Aug. 31, 2025

Aug. 31, 2025

https://github.com/python/cpython/commit/a37a382f78f3ab52fc01fbdd6bf7c58185… commit: a37a382f78f3ab52fc01fbdd6bf7c58185459f4f branch: 3.14 author: Miss Islington (bot) <31488909+miss-islington(a)users.noreply.github.com> committer: hugovk <1324225+hugovk(a)users.noreply.github.com> date: 2025-08-31T09:44:56+03:00 summary: [3.14] gh-133829: Remove some specifics from the ``zipimport`` example (GH-133835) (#138274) Co-authored-by: R Chintan Meher <meherrihaan(a)gmail.com> Co-authored-by: Adam Turner <9087854+AA-Turner(a)users.noreply.github.com> Co-authored-by: Peter Bierma <zintensitydev(a)gmail.com> Co-authored-by: sobolevn <mail(a)sobolevn.me> Co-authored-by: Stan Ulbrych <89152624+StanFromIreland(a)users.noreply.github.com> files: M Doc/library/zipimport.rst diff --git a/Doc/library/zipimport.rst b/Doc/library/zipimport.rst index cd76f29a556939..ead87c2d6ad2e5 100644 --- a/Doc/library/zipimport.rst +++ b/Doc/library/zipimport.rst @@ -188,17 +188,20 @@ Here is an example that imports a module from a ZIP archive - note that the .. code-block:: shell-session - $ unzip -l example.zip - Archive: example.zip + $ unzip -l example_archive.zip + Archive: example_archive.zip Length Date Time Name -------- ---- ---- ---- - 8467 11-26-02 22:30 jwzthreading.py + 8467 01-01-00 12:30 example.py -------- ------- 8467 1 file - $ ./python - Python 2.3 (#1, Aug 1 2003, 19:54:32) + +.. code-block:: pycon + >>> import sys - >>> sys.path.insert(0, 'example.zip') # Add .zip file to front of path - >>> import jwzthreading - >>> jwzthreading.__file__ - 'example.zip/jwzthreading.py' + >>> # Add the archive to the front of the module search path + >>> sys.path.insert(0, 'example_archive.zip') + >>> import example + >>> example.__file__ + 'example_archive.zip/example.py' +

1 0

gh-116738: Make `_json` module thread-safe in the free-threading (#119438)
by kumaraditya303 Aug. 31, 2025

Aug. 31, 2025

https://github.com/python/cpython/commit/43573028c6ae21c66c118b8bae866c8968… commit: 43573028c6ae21c66c118b8bae866c8968b87b68 branch: main author: Pieter Eendebak <pieter.eendebak(a)gmail.com> committer: kumaraditya303 <kumaraditya(a)python.org> date: 2025-08-31T09:42:45+05:30 summary: gh-116738: Make `_json` module thread-safe in the free-threading (#119438) Co-authored-by: Kumar Aditya <kumaraditya(a)python.org> files: A Lib/test/test_free_threading/test_json.py A Misc/NEWS.d/next/Core_and_Builtins/2024-06-04-20-26-21.gh-issue-116738.q_hPYq.rst M Modules/_json.c diff --git a/Lib/test/test_free_threading/test_json.py b/Lib/test/test_free_threading/test_json.py new file mode 100644 index 00000000000000..8a541e960a63ad --- /dev/null +++ b/Lib/test/test_free_threading/test_json.py @@ -0,0 +1,79 @@ +from threading import Barrier, Thread +from test.test_json import CTest +from test.support import threading_helper + + +def encode_json_helper( + json, worker, data, number_of_threads=12, number_of_json_encodings=100 +): + worker_threads = [] + barrier = Barrier(number_of_threads) + for index in range(number_of_threads): + worker_threads.append( + Thread(target=worker, args=[barrier, data, index]) + ) + for t in worker_threads: + t.start() + for ii in range(number_of_json_encodings): + json.dumps(data) + data.clear() + for t in worker_threads: + t.join() + + +class MyMapping(dict): + def __init__(self): + self.mapping = [] + + def items(self): + return self.mapping + + +(a)threading_helper.reap_threads +(a)threading_helper.requires_working_threading() +class TestJsonEncoding(CTest): + # Test encoding json with concurrent threads modifying the data cannot + # corrupt the interpreter + + def test_json_mutating_list(self): + def worker(barrier, data, index): + barrier.wait() + while data: + for d in data: + if len(d) > 5: + d.clear() + else: + d.append(index) + + data = [[], []] + encode_json_helper(self.json, worker, data) + + def test_json_mutating_exact_dict(self): + def worker(barrier, data, index): + barrier.wait() + while data: + for d in data: + if len(d) > 5: + try: + key = list(d)[0] + d.pop() + except (KeyError, IndexError): + pass + else: + d[index] = index + + data = [{}, {}] + encode_json_helper(self.json, worker, data) + + def test_json_mutating_mapping(self): + def worker(barrier, data, index): + barrier.wait() + while data: + for d in data: + if len(d.mapping) > 3: + d.mapping.clear() + else: + d.mapping.append((index, index)) + + data = [MyMapping(), MyMapping()] + encode_json_helper(self.json, worker, data) diff --git a/Misc/NEWS.d/next/Core_and_Builtins/2024-06-04-20-26-21.gh-issue-116738.q_hPYq.rst b/Misc/NEWS.d/next/Core_and_Builtins/2024-06-04-20-26-21.gh-issue-116738.q_hPYq.rst new file mode 100644 index 00000000000000..fe6a7229223de8 --- /dev/null +++ b/Misc/NEWS.d/next/Core_and_Builtins/2024-06-04-20-26-21.gh-issue-116738.q_hPYq.rst @@ -0,0 +1 @@ +Make the module :mod:`json` safe to use under the free-threading build. diff --git a/Modules/_json.c b/Modules/_json.c index e1d6042cb78ab5..9a1fc3aba36116 100644 --- a/Modules/_json.c +++ b/Modules/_json.c @@ -10,6 +10,7 @@ #include "Python.h" #include "pycore_ceval.h" // _Py_EnterRecursiveCall() +#include "pycore_critical_section.h" // Py_BEGIN_CRITICAL_SECTION_SEQUENCE_FAST() #include "pycore_global_strings.h" // _Py_ID() #include "pycore_pyerrors.h" // _PyErr_FormatNote #include "pycore_runtime.h" // _PyRuntime @@ -1456,7 +1457,7 @@ write_newline_indent(PyUnicodeWriter *writer, static PyObject * encoder_call(PyObject *op, PyObject *args, PyObject *kwds) { - /* Python callable interface to encode_listencode_obj */ + /* Python callable interface to encoder_listencode_obj */ static char *kwlist[] = {"obj", "_current_indent_level", NULL}; PyObject *obj; Py_ssize_t indent_level; @@ -1743,15 +1744,84 @@ encoder_encode_key_value(PyEncoderObject *s, PyUnicodeWriter *writer, bool *firs return 0; } +static inline int +_encoder_iterate_mapping_lock_held(PyEncoderObject *s, PyUnicodeWriter *writer, + bool *first, PyObject *dct, PyObject *items, + Py_ssize_t indent_level, PyObject *indent_cache, + PyObject *separator) +{ + _Py_CRITICAL_SECTION_ASSERT_OBJECT_LOCKED(items); + PyObject *key, *value; + for (Py_ssize_t i = 0; i < PyList_GET_SIZE(items); i++) { + PyObject *item = PyList_GET_ITEM(items, i); +#ifdef Py_GIL_DISABLED + // gh-119438: in the free-threading build the critical section on items can get suspended + Py_INCREF(item); +#endif + if (!PyTuple_Check(item) || PyTuple_GET_SIZE(item) != 2) { + PyErr_SetString(PyExc_ValueError, "items must return 2-tuples"); +#ifdef Py_GIL_DISABLED + Py_DECREF(item); +#endif + return -1; + } + + key = PyTuple_GET_ITEM(item, 0); + value = PyTuple_GET_ITEM(item, 1); + if (encoder_encode_key_value(s, writer, first, dct, key, value, + indent_level, indent_cache, + separator) < 0) { +#ifdef Py_GIL_DISABLED + Py_DECREF(item); +#endif + return -1; + } +#ifdef Py_GIL_DISABLED + Py_DECREF(item); +#endif + } + + return 0; +} + +static inline int +_encoder_iterate_dict_lock_held(PyEncoderObject *s, PyUnicodeWriter *writer, + bool *first, PyObject *dct, Py_ssize_t indent_level, + PyObject *indent_cache, PyObject *separator) +{ + _Py_CRITICAL_SECTION_ASSERT_OBJECT_LOCKED(dct); + PyObject *key, *value; + Py_ssize_t pos = 0; + while (PyDict_Next(dct, &pos, &key, &value)) { +#ifdef Py_GIL_DISABLED + // gh-119438: in the free-threading build the critical section on dct can get suspended + Py_INCREF(key); + Py_INCREF(value); +#endif + if (encoder_encode_key_value(s, writer, first, dct, key, value, + indent_level, indent_cache, + separator) < 0) { +#ifdef Py_GIL_DISABLED + Py_DECREF(key); + Py_DECREF(value); +#endif + return -1; + } +#ifdef Py_GIL_DISABLED + Py_DECREF(key); + Py_DECREF(value); +#endif + } + return 0; +} + static int encoder_listencode_dict(PyEncoderObject *s, PyUnicodeWriter *writer, - PyObject *dct, + PyObject *dct, Py_ssize_t indent_level, PyObject *indent_cache) { /* Encode Python dict dct a JSON term */ PyObject *ident = NULL; - PyObject *items = NULL; - PyObject *key, *value; bool first = true; if (PyDict_GET_SIZE(dct) == 0) { @@ -1788,34 +1858,30 @@ encoder_listencode_dict(PyEncoderObject *s, PyUnicodeWriter *writer, } if (s->sort_keys || !PyDict_CheckExact(dct)) { - items = PyMapping_Items(dct); - if (items == NULL || (s->sort_keys && PyList_Sort(items) < 0)) + PyObject *items = PyMapping_Items(dct); + if (items == NULL || (s->sort_keys && PyList_Sort(items) < 0)) { + Py_XDECREF(items); goto bail; + } - for (Py_ssize_t i = 0; i < PyList_GET_SIZE(items); i++) { - PyObject *item = PyList_GET_ITEM(items, i); - - if (!PyTuple_Check(item) || PyTuple_GET_SIZE(item) != 2) { - PyErr_SetString(PyExc_ValueError, "items must return 2-tuples"); - goto bail; - } - - key = PyTuple_GET_ITEM(item, 0); - value = PyTuple_GET_ITEM(item, 1); - if (encoder_encode_key_value(s, writer, &first, dct, key, value, - indent_level, indent_cache, - separator) < 0) - goto bail; + int result; + Py_BEGIN_CRITICAL_SECTION_SEQUENCE_FAST(items); + result = _encoder_iterate_mapping_lock_held(s, writer, &first, dct, + items, indent_level, indent_cache, separator); + Py_END_CRITICAL_SECTION_SEQUENCE_FAST(); + Py_DECREF(items); + if (result < 0) { + goto bail; } - Py_CLEAR(items); } else { - Py_ssize_t pos = 0; - while (PyDict_Next(dct, &pos, &key, &value)) { - if (encoder_encode_key_value(s, writer, &first, dct, key, value, - indent_level, indent_cache, - separator) < 0) - goto bail; + int result; + Py_BEGIN_CRITICAL_SECTION(dct); + result = _encoder_iterate_dict_lock_held(s, writer, &first, dct, + indent_level, indent_cache, separator); + Py_END_CRITICAL_SECTION(); + if (result < 0) { + goto bail; } } @@ -1837,11 +1903,43 @@ encoder_listencode_dict(PyEncoderObject *s, PyUnicodeWriter *writer, return 0; bail: - Py_XDECREF(items); Py_XDECREF(ident); return -1; } +static inline int +_encoder_iterate_fast_seq_lock_held(PyEncoderObject *s, PyUnicodeWriter *writer, + PyObject *seq, PyObject *s_fast, + Py_ssize_t indent_level, PyObject *indent_cache, PyObject *separator) +{ + for (Py_ssize_t i = 0; i < PySequence_Fast_GET_SIZE(s_fast); i++) { + PyObject *obj = PySequence_Fast_GET_ITEM(s_fast, i); +#ifdef Py_GIL_DISABLED + // gh-119438: in the free-threading build the critical section on s_fast can get suspended + Py_INCREF(obj); +#endif + if (i) { + if (PyUnicodeWriter_WriteStr(writer, separator) < 0) { +#ifdef Py_GIL_DISABLED + Py_DECREF(obj); +#endif + return -1; + } + } + if (encoder_listencode_obj(s, writer, obj, indent_level, indent_cache)) { + _PyErr_FormatNote("when serializing %T item %zd", seq, i); +#ifdef Py_GIL_DISABLED + Py_DECREF(obj); +#endif + return -1; + } +#ifdef Py_GIL_DISABLED + Py_DECREF(obj); +#endif + } + return 0; +} + static int encoder_listencode_list(PyEncoderObject *s, PyUnicodeWriter *writer, PyObject *seq, @@ -1849,10 +1947,8 @@ encoder_listencode_list(PyEncoderObject *s, PyUnicodeWriter *writer, { PyObject *ident = NULL; PyObject *s_fast = NULL; - Py_ssize_t i; - ident = NULL; - s_fast = PySequence_Fast(seq, "_iterencode_list needs a sequence"); + s_fast = PySequence_Fast(seq, "encoder_listencode_list needs a sequence"); if (s_fast == NULL) return -1; if (PySequence_Fast_GET_SIZE(s_fast) == 0) { @@ -1890,16 +1986,13 @@ encoder_listencode_list(PyEncoderObject *s, PyUnicodeWriter *writer, goto bail; } } - for (i = 0; i < PySequence_Fast_GET_SIZE(s_fast); i++) { - PyObject *obj = PySequence_Fast_GET_ITEM(s_fast, i); - if (i) { - if (PyUnicodeWriter_WriteStr(writer, separator) < 0) - goto bail; - } - if (encoder_listencode_obj(s, writer, obj, indent_level, indent_cache)) { - _PyErr_FormatNote("when serializing %T item %zd", seq, i); - goto bail; - } + int result; + Py_BEGIN_CRITICAL_SECTION_SEQUENCE_FAST(seq); + result = _encoder_iterate_fast_seq_lock_held(s, writer, seq, s_fast, + indent_level, indent_cache, separator); + Py_END_CRITICAL_SECTION_SEQUENCE_FAST(); + if (result < 0) { + goto bail; } if (ident != NULL) { if (PyDict_DelItem(s->markers, ident))

1 0