Mailman 3 January 2025 - Python-checkins

GH-92897: schedule the check_home deprecation to 3.15 (#129102)
by FFY00 Jan. 20, 2025

Jan. 20, 2025

https://github.com/python/cpython/commit/e52ab564da84540a7544cb829f262b0154… commit: e52ab564da84540a7544cb829f262b0154efccae branch: main author: Filipe Laíns 🇵🇸 <lains(a)riseup.net> committer: FFY00 <filipe.lains(a)gmail.com> date: 2025-01-20T21:25:14Z summary: GH-92897: schedule the check_home deprecation to 3.15 (#129102) files: A Misc/NEWS.d/next/Library/2025-01-20-20-59-26.gh-issue-92897.G0xH8o.rst M Doc/deprecations/pending-removal-in-3.15.rst M Lib/sysconfig/__init__.py M Lib/test/test_sysconfig.py diff --git a/Doc/deprecations/pending-removal-in-3.15.rst b/Doc/deprecations/pending-removal-in-3.15.rst index 3b03e1f49e6754..390bbff2835cf8 100644 --- a/Doc/deprecations/pending-removal-in-3.15.rst +++ b/Doc/deprecations/pending-removal-in-3.15.rst @@ -51,6 +51,11 @@ Pending removal in Python 3.15 This function is only useful for Jython support, has a confusing API, and is largely untested. +* :mod:`sysconfig`: + + * The ``check_home`` argument of :func:`sysconfig.is_python_build` has been + deprecated since Python 3.12. + * :mod:`threading`: * :func:`~threading.RLock` will take no arguments in Python 3.15. diff --git a/Lib/sysconfig/__init__.py b/Lib/sysconfig/__init__.py index 86dd391aa173b5..3c3c9796ec3307 100644 --- a/Lib/sysconfig/__init__.py +++ b/Lib/sysconfig/__init__.py @@ -220,8 +220,15 @@ def _safe_realpath(path): def is_python_build(check_home=None): if check_home is not None: import warnings - warnings.warn("check_home argument is deprecated and ignored.", - DeprecationWarning, stacklevel=2) + warnings.warn( + ( + 'The check_home argument of sysconfig.is_python_build is ' + 'deprecated and its value is ignored. ' + 'It will be removed in Python 3.15.' + ), + DeprecationWarning, + stacklevel=2, + ) for fn in ("Setup", "Setup.local"): if os.path.isfile(os.path.join(_PROJECT_BASE, "Modules", fn)): return True diff --git a/Lib/test/test_sysconfig.py b/Lib/test/test_sysconfig.py index a5676027eb7438..1002d90074599a 100644 --- a/Lib/test/test_sysconfig.py +++ b/Lib/test/test_sysconfig.py @@ -712,9 +712,9 @@ def test_parse_makefile(self): class DeprecationTests(unittest.TestCase): - def deprecated(self, removal_version, deprecation_msg=None, attribute_msg=None): + def deprecated(self, removal_version, deprecation_msg=None, error=Exception, error_msg=None): if sys.version_info >= removal_version: - return self.assertRaises(AttributeError, msg=attribute_msg) + return self.assertRaises(error, msg=error_msg) else: return self.assertWarns(DeprecationWarning, msg=deprecation_msg) @@ -725,10 +725,24 @@ def test_expand_makefile_vars(self): 'sysconfig.expand_makefile_vars is deprecated and will be removed in ' 'Python 3.16. Use sysconfig.get_paths(vars=...) instead.', ), - attribute_msg="module 'sysconfig' has no attribute 'expand_makefile_vars'", + error=AttributeError, + error_msg="module 'sysconfig' has no attribute 'expand_makefile_vars'", ): sysconfig.expand_makefile_vars('', {}) + def test_is_python_build_check_home(self): + with self.deprecated( + removal_version=(3, 15), + deprecation_msg=( + 'The check_home argument of sysconfig.is_python_build is ' + 'deprecated and its value is ignored. ' + 'It will be removed in Python 3.15.' + ), + error=TypeError, + error_msg="is_python_build() takes 0 positional arguments but 1 were given", + ): + sysconfig.is_python_build('foo') + if __name__ == "__main__": unittest.main() diff --git a/Misc/NEWS.d/next/Library/2025-01-20-20-59-26.gh-issue-92897.G0xH8o.rst b/Misc/NEWS.d/next/Library/2025-01-20-20-59-26.gh-issue-92897.G0xH8o.rst new file mode 100644 index 00000000000000..632ca03bbf8dd2 --- /dev/null +++ b/Misc/NEWS.d/next/Library/2025-01-20-20-59-26.gh-issue-92897.G0xH8o.rst @@ -0,0 +1,2 @@ +Scheduled the deprecation of the ``check_home`` argument of +:func:`sysconfig.is_python_build` to Python 3.15.

1 0

[3.13] gh-111178: fix UBSan failures in `Modules/_multiprocessing/semaphore.c` (GH-129084) (#129100)
by gpshead Jan. 20, 2025

Jan. 20, 2025

https://github.com/python/cpython/commit/c5e1131cfba5c6f083b149ea58a550db0f… commit: c5e1131cfba5c6f083b149ea58a550db0f9e3a7a branch: 3.13 author: Miss Islington (bot) <31488909+miss-islington(a)users.noreply.github.com> committer: gpshead <greg(a)krypto.org> date: 2025-01-20T21:05:52Z summary: [3.13] gh-111178: fix UBSan failures in `Modules/_multiprocessing/semaphore.c` (GH-129084) (#129100) gh-111178: fix UBSan failures in `Modules/_multiprocessing/semaphore.c` (GH-129084) fix UBSan failures for `SemLockObject` (cherry picked from commit 5ed5572cac7ef204767ddf8e8888e15672ba558e) Co-authored-by: Bénédikt Tran <10796600+picnixz(a)users.noreply.github.com> files: M Modules/_multiprocessing/semaphore.c diff --git a/Modules/_multiprocessing/semaphore.c b/Modules/_multiprocessing/semaphore.c index 4de4ee6c78fbd1..8c6a4a279205d3 100644 --- a/Modules/_multiprocessing/semaphore.c +++ b/Modules/_multiprocessing/semaphore.c @@ -27,6 +27,8 @@ typedef struct { char *name; } SemLockObject; +#define _SemLockObject_CAST(op) ((SemLockObject *)(op)) + /*[python input] class SEM_HANDLE_converter(CConverter): type = "SEM_HANDLE" @@ -575,8 +577,9 @@ _multiprocessing_SemLock__rebuild_impl(PyTypeObject *type, SEM_HANDLE handle, } static void -semlock_dealloc(SemLockObject* self) +semlock_dealloc(PyObject *op) { + SemLockObject *self = _SemLockObject_CAST(op); PyTypeObject *tp = Py_TYPE(self); PyObject_GC_UnTrack(self); if (self->handle != SEM_FAILED) @@ -717,7 +720,7 @@ _multiprocessing_SemLock___exit___impl(SemLockObject *self, } static int -semlock_traverse(SemLockObject *s, visitproc visit, void *arg) +semlock_traverse(PyObject *s, visitproc visit, void *arg) { Py_VISIT(Py_TYPE(s)); return 0;

1 0

[3.12] gh-111178: fix UBSan failures in `Modules/_multiprocessing/semaphore.c` (GH-129084) (#129101)
by gpshead Jan. 20, 2025

Jan. 20, 2025

https://github.com/python/cpython/commit/9b335cc8104dd83a5a1343dc649d1f3606… commit: 9b335cc8104dd83a5a1343dc649d1f3606682098 branch: 3.12 author: Miss Islington (bot) <31488909+miss-islington(a)users.noreply.github.com> committer: gpshead <greg(a)krypto.org> date: 2025-01-20T21:00:09Z summary: [3.12] gh-111178: fix UBSan failures in `Modules/_multiprocessing/semaphore.c` (GH-129084) (#129101) gh-111178: fix UBSan failures in `Modules/_multiprocessing/semaphore.c` (GH-129084) fix UBSan failures for `SemLockObject` (cherry picked from commit 5ed5572cac7ef204767ddf8e8888e15672ba558e) Co-authored-by: Bénédikt Tran <10796600+picnixz(a)users.noreply.github.com> files: M Modules/_multiprocessing/semaphore.c diff --git a/Modules/_multiprocessing/semaphore.c b/Modules/_multiprocessing/semaphore.c index c7df82dfe2d0cc..bc718e273a8721 100644 --- a/Modules/_multiprocessing/semaphore.c +++ b/Modules/_multiprocessing/semaphore.c @@ -23,6 +23,8 @@ typedef struct { char *name; } SemLockObject; +#define _SemLockObject_CAST(op) ((SemLockObject *)(op)) + /*[python input] class SEM_HANDLE_converter(CConverter): type = "SEM_HANDLE" @@ -567,8 +569,9 @@ _multiprocessing_SemLock__rebuild_impl(PyTypeObject *type, SEM_HANDLE handle, } static void -semlock_dealloc(SemLockObject* self) +semlock_dealloc(PyObject *op) { + SemLockObject *self = _SemLockObject_CAST(op); PyTypeObject *tp = Py_TYPE(self); PyObject_GC_UnTrack(self); if (self->handle != SEM_FAILED) @@ -706,7 +709,7 @@ _multiprocessing_SemLock___exit___impl(SemLockObject *self, } static int -semlock_traverse(SemLockObject *s, visitproc visit, void *arg) +semlock_traverse(PyObject *s, visitproc visit, void *arg) { Py_VISIT(Py_TYPE(s)); return 0;

1 0

gh-111178: fix UBSan failures in `Modules/_multiprocessing/semaphore.c` (#129084)
by gpshead Jan. 20, 2025

Jan. 20, 2025

https://github.com/python/cpython/commit/5ed5572cac7ef204767ddf8e8888e15672… commit: 5ed5572cac7ef204767ddf8e8888e15672ba558e branch: main author: Bénédikt Tran <10796600+picnixz(a)users.noreply.github.com> committer: gpshead <greg(a)krypto.org> date: 2025-01-20T20:43:55Z summary: gh-111178: fix UBSan failures in `Modules/_multiprocessing/semaphore.c` (#129084) fix UBSan failures for `SemLockObject` files: M Modules/_multiprocessing/semaphore.c diff --git a/Modules/_multiprocessing/semaphore.c b/Modules/_multiprocessing/semaphore.c index 9eef7c25636899..036db2cd4c6c85 100644 --- a/Modules/_multiprocessing/semaphore.c +++ b/Modules/_multiprocessing/semaphore.c @@ -28,6 +28,8 @@ typedef struct { char *name; } SemLockObject; +#define _SemLockObject_CAST(op) ((SemLockObject *)(op)) + /*[python input] class SEM_HANDLE_converter(CConverter): type = "SEM_HANDLE" @@ -576,8 +578,9 @@ _multiprocessing_SemLock__rebuild_impl(PyTypeObject *type, SEM_HANDLE handle, } static void -semlock_dealloc(SemLockObject* self) +semlock_dealloc(PyObject *op) { + SemLockObject *self = _SemLockObject_CAST(op); PyTypeObject *tp = Py_TYPE(self); PyObject_GC_UnTrack(self); if (self->handle != SEM_FAILED) @@ -718,7 +721,7 @@ _multiprocessing_SemLock___exit___impl(SemLockObject *self, } static int -semlock_traverse(SemLockObject *s, visitproc visit, void *arg) +semlock_traverse(PyObject *s, visitproc visit, void *arg) { Py_VISIT(Py_TYPE(s)); return 0;

1 0

doc: add social cards generation support in sphinxext-opengraph (#129085)
by FFY00 Jan. 20, 2025

Jan. 20, 2025

https://github.com/python/cpython/commit/382340d5594a05d8a5362a06fe55119eec… commit: 382340d5594a05d8a5362a06fe55119eec1d8e04 branch: main author: Filipe Laíns 🇵🇸 <lains(a)riseup.net> committer: FFY00 <filipe.lains(a)gmail.com> date: 2025-01-20T20:41:27Z summary: doc: add social cards generation support in sphinxext-opengraph (#129085) files: A Doc/requirements-no-build.txt M Doc/Makefile diff --git a/Doc/Makefile b/Doc/Makefile index b8896da4a91869..53af6aca659cef 100644 --- a/Doc/Makefile +++ b/Doc/Makefile @@ -13,7 +13,8 @@ JOBS = auto PAPER = SOURCES = DISTVERSION = $(shell $(PYTHON) tools/extensions/patchlevel.py) -REQUIREMENTS = requirements.txt +IS_PYTHON_BUILD = $(shell $(PYTHON) -c 'import sysconfig; print("true") if sysconfig.is_python_build() else None') +REQUIREMENTS = $(if $(IS_PYTHON_BUILD),requirements.txt,requirements-no-build.txt) SPHINXERRORHANDLING = --fail-on-warning # Internal variables. diff --git a/Doc/requirements-no-build.txt b/Doc/requirements-no-build.txt new file mode 100644 index 00000000000000..8b3c4d410a75a3 --- /dev/null +++ b/Doc/requirements-no-build.txt @@ -0,0 +1,7 @@ +-r requirements.txt + +# Requirements when not running the build Python, +# such as optional native extensions. + +# Needed to support generating social cards in sphinxext-opengraph. +--only-binary matplotlib

1 0

[3.12] GH-128131: Completely support random read access of uncompressed unencrypted files in ZipFile (GH-128143) (#129092)
by jaraco Jan. 20, 2025

Jan. 20, 2025

https://github.com/python/cpython/commit/03bce18992fd49f5f87697a21c0390fc60… commit: 03bce18992fd49f5f87697a21c0390fc6082bfc7 branch: 3.12 author: Miss Islington (bot) <31488909+miss-islington(a)users.noreply.github.com> committer: jaraco <jaraco(a)jaraco.com> date: 2025-01-20T20:40:28Z summary: [3.12] GH-128131: Completely support random read access of uncompressed unencrypted files in ZipFile (GH-128143) (#129092) GH-128131: Completely support random read access of uncompressed unencrypted files in ZipFile (GH-128143) (cherry picked from commit dda02eb7be62bf0af850a7521c77c90ea997df6c) Co-authored-by: 5ec1cff <56485584+5ec1cff(a)users.noreply.github.com> Co-authored-by: blurb-it[bot] <43283697+blurb-it[bot](a)users.noreply.github.com> Co-authored-by: Bénédikt Tran <10796600+picnixz(a)users.noreply.github.com> files: A Misc/NEWS.d/next/Library/2024-12-21-03-20-12.gh-issue-128131.QpPmNt.rst M Lib/test/test_zipfile/test_core.py M Lib/zipfile/__init__.py diff --git a/Lib/test/test_zipfile/test_core.py b/Lib/test/test_zipfile/test_core.py index adb74345486b65..03520e57303c9d 100644 --- a/Lib/test/test_zipfile/test_core.py +++ b/Lib/test/test_zipfile/test_core.py @@ -1,3 +1,4 @@ +import _pyio import array import contextlib import importlib.util @@ -3440,5 +3441,87 @@ def test_too_short(self): b"zzz", zipfile._strip_extra(b"zzz", (self.ZIP64_EXTRA,))) +class StatIO(_pyio.BytesIO): + """Buffer which remembers the number of bytes that were read.""" + + def __init__(self): + super().__init__() + self.bytes_read = 0 + + def read(self, size=-1): + bs = super().read(size) + self.bytes_read += len(bs) + return bs + + +class StoredZipExtFileRandomReadTest(unittest.TestCase): + """Tests whether an uncompressed, unencrypted zip entry can be randomly + seek and read without reading redundant bytes.""" + def test_stored_seek_and_read(self): + + sio = StatIO() + # 20000 bytes + txt = b'0123456789' * 2000 + + # The seek length must be greater than ZipExtFile.MIN_READ_SIZE + # as `ZipExtFile._read2()` reads in blocks of this size and we + # need to seek out of the buffered data + read_buffer_size = zipfile.ZipExtFile.MIN_READ_SIZE + self.assertGreaterEqual(10002, read_buffer_size) # for forward seek test + self.assertGreaterEqual(5003, read_buffer_size) # for backward seek test + # The read length must be less than MIN_READ_SIZE, since we assume that + # only 1 block is read in the test. + read_length = 100 + self.assertGreaterEqual(read_buffer_size, read_length) # for read() calls + + with zipfile.ZipFile(sio, "w", compression=zipfile.ZIP_STORED) as zipf: + zipf.writestr("foo.txt", txt) + + # check random seek and read on a file + with zipfile.ZipFile(sio, "r") as zipf: + with zipf.open("foo.txt", "r") as fp: + # Test this optimized read hasn't rewound and read from the + # start of the file (as in the case of the unoptimized path) + + # forward seek + old_count = sio.bytes_read + forward_seek_len = 10002 + current_pos = 0 + fp.seek(forward_seek_len, os.SEEK_CUR) + current_pos += forward_seek_len + self.assertEqual(fp.tell(), current_pos) + self.assertEqual(fp._left, fp._compress_left) + arr = fp.read(read_length) + current_pos += read_length + self.assertEqual(fp.tell(), current_pos) + self.assertEqual(arr, txt[current_pos - read_length:current_pos]) + self.assertEqual(fp._left, fp._compress_left) + read_count = sio.bytes_read - old_count + self.assertLessEqual(read_count, read_buffer_size) + + # backward seek + old_count = sio.bytes_read + backward_seek_len = 5003 + fp.seek(-backward_seek_len, os.SEEK_CUR) + current_pos -= backward_seek_len + self.assertEqual(fp.tell(), current_pos) + self.assertEqual(fp._left, fp._compress_left) + arr = fp.read(read_length) + current_pos += read_length + self.assertEqual(fp.tell(), current_pos) + self.assertEqual(arr, txt[current_pos - read_length:current_pos]) + self.assertEqual(fp._left, fp._compress_left) + read_count = sio.bytes_read - old_count + self.assertLessEqual(read_count, read_buffer_size) + + # eof flags test + fp.seek(0, os.SEEK_END) + fp.seek(12345, os.SEEK_SET) + current_pos = 12345 + arr = fp.read(read_length) + current_pos += read_length + self.assertEqual(arr, txt[current_pos - read_length:current_pos]) + + if __name__ == "__main__": unittest.main() diff --git a/Lib/zipfile/__init__.py b/Lib/zipfile/__init__.py index 3e1cca472865b5..91b2e032e5f2d9 100644 --- a/Lib/zipfile/__init__.py +++ b/Lib/zipfile/__init__.py @@ -1140,13 +1140,15 @@ def seek(self, offset, whence=os.SEEK_SET): self._offset = buff_offset read_offset = 0 # Fast seek uncompressed unencrypted file - elif self._compress_type == ZIP_STORED and self._decrypter is None and read_offset > 0: + elif self._compress_type == ZIP_STORED and self._decrypter is None and read_offset != 0: # disable CRC checking after first seeking - it would be invalid self._expected_crc = None # seek actual file taking already buffered data into account read_offset -= len(self._readbuffer) - self._offset self._fileobj.seek(read_offset, os.SEEK_CUR) self._left -= read_offset + self._compress_left -= read_offset + self._eof = self._left <= 0 read_offset = 0 # flush read buffer self._readbuffer = b'' diff --git a/Misc/NEWS.d/next/Library/2024-12-21-03-20-12.gh-issue-128131.QpPmNt.rst b/Misc/NEWS.d/next/Library/2024-12-21-03-20-12.gh-issue-128131.QpPmNt.rst new file mode 100644 index 00000000000000..f4c4ebce10729c --- /dev/null +++ b/Misc/NEWS.d/next/Library/2024-12-21-03-20-12.gh-issue-128131.QpPmNt.rst @@ -0,0 +1,2 @@ +Completely support random access of uncompressed unencrypted read-only +zip files obtained by :meth:`ZipFile.open <zipfile.ZipFile.open>`.

1 0

[3.12] doc: fix venv creating for the local Python using uv (GH-129094) (#129097)
by hugovk Jan. 20, 2025

Jan. 20, 2025

https://github.com/python/cpython/commit/728c6b89a66b7cb431a99fef835674ddbd… commit: 728c6b89a66b7cb431a99fef835674ddbd765d37 branch: 3.12 author: Miss Islington (bot) <31488909+miss-islington(a)users.noreply.github.com> committer: hugovk <1324225+hugovk(a)users.noreply.github.com> date: 2025-01-20T21:13:23+02:00 summary: [3.12] doc: fix venv creating for the local Python using uv (GH-129094) (#129097) Co-authored-by: Filipe Laíns 🇵🇸 <lains(a)riseup.net> files: M Doc/Makefile diff --git a/Doc/Makefile b/Doc/Makefile index 1a66642a4a03ed..b8896da4a91869 100644 --- a/Doc/Makefile +++ b/Doc/Makefile @@ -172,7 +172,7 @@ venv: else \ echo "Creating venv in $(VENVDIR)"; \ if $(UV) --version >/dev/null 2>&1; then \ - $(UV) venv $(VENVDIR); \ + $(UV) venv --python=$(PYTHON) $(VENVDIR); \ VIRTUAL_ENV=$(VENVDIR) $(UV) pip install -r $(REQUIREMENTS); \ else \ $(PYTHON) -m venv $(VENVDIR); \

1 0

[3.13] doc: fix venv creating for the local Python using uv (GH-129094) (#129096)
by hugovk Jan. 20, 2025

Jan. 20, 2025

https://github.com/python/cpython/commit/207708246327c8b179529020007cedd94b… commit: 207708246327c8b179529020007cedd94b19cd86 branch: 3.13 author: Miss Islington (bot) <31488909+miss-islington(a)users.noreply.github.com> committer: hugovk <1324225+hugovk(a)users.noreply.github.com> date: 2025-01-20T21:13:12+02:00 summary: [3.13] doc: fix venv creating for the local Python using uv (GH-129094) (#129096) Co-authored-by: Filipe Laíns 🇵🇸 <lains(a)riseup.net> files: M Doc/Makefile diff --git a/Doc/Makefile b/Doc/Makefile index 1a66642a4a03ed..b8896da4a91869 100644 --- a/Doc/Makefile +++ b/Doc/Makefile @@ -172,7 +172,7 @@ venv: else \ echo "Creating venv in $(VENVDIR)"; \ if $(UV) --version >/dev/null 2>&1; then \ - $(UV) venv $(VENVDIR); \ + $(UV) venv --python=$(PYTHON) $(VENVDIR); \ VIRTUAL_ENV=$(VENVDIR) $(UV) pip install -r $(REQUIREMENTS); \ else \ $(PYTHON) -m venv $(VENVDIR); \

1 0

doc: fix venv creating for the local Python using uv (#129094)
by hugovk Jan. 20, 2025

Jan. 20, 2025

https://github.com/python/cpython/commit/da36bccdac6d45bda6f83563410567c04f… commit: da36bccdac6d45bda6f83563410567c04f4b288c branch: main author: Filipe Laíns 🇵🇸 <lains(a)riseup.net> committer: hugovk <1324225+hugovk(a)users.noreply.github.com> date: 2025-01-20T21:04:05+02:00 summary: doc: fix venv creating for the local Python using uv (#129094) files: M Doc/Makefile diff --git a/Doc/Makefile b/Doc/Makefile index 1a66642a4a03ed..b8896da4a91869 100644 --- a/Doc/Makefile +++ b/Doc/Makefile @@ -172,7 +172,7 @@ venv: else \ echo "Creating venv in $(VENVDIR)"; \ if $(UV) --version >/dev/null 2>&1; then \ - $(UV) venv $(VENVDIR); \ + $(UV) venv --python=$(PYTHON) $(VENVDIR); \ VIRTUAL_ENV=$(VENVDIR) $(UV) pip install -r $(REQUIREMENTS); \ else \ $(PYTHON) -m venv $(VENVDIR); \

1 0

[3.13] GH-128131: Completely support random read access of uncompressed unencrypted files in ZipFile (GH-128143) (#129091)
by jaraco Jan. 20, 2025

Jan. 20, 2025

https://github.com/python/cpython/commit/03b2ecf41c6f4cc57aac3dd147b8b5dedc… commit: 03b2ecf41c6f4cc57aac3dd147b8b5dedcafcd3d branch: 3.13 author: Miss Islington (bot) <31488909+miss-islington(a)users.noreply.github.com> committer: jaraco <jaraco(a)jaraco.com> date: 2025-01-20T18:28:52Z summary: [3.13] GH-128131: Completely support random read access of uncompressed unencrypted files in ZipFile (GH-128143) (#129091) GH-128131: Completely support random read access of uncompressed unencrypted files in ZipFile (GH-128143) (cherry picked from commit dda02eb7be62bf0af850a7521c77c90ea997df6c) Co-authored-by: 5ec1cff <56485584+5ec1cff(a)users.noreply.github.com> Co-authored-by: blurb-it[bot] <43283697+blurb-it[bot](a)users.noreply.github.com> Co-authored-by: Bénédikt Tran <10796600+picnixz(a)users.noreply.github.com> files: A Misc/NEWS.d/next/Library/2024-12-21-03-20-12.gh-issue-128131.QpPmNt.rst M Lib/test/test_zipfile/test_core.py M Lib/zipfile/__init__.py diff --git a/Lib/test/test_zipfile/test_core.py b/Lib/test/test_zipfile/test_core.py index 4ff9f9c34237c6..4b56f6a380f219 100644 --- a/Lib/test/test_zipfile/test_core.py +++ b/Lib/test/test_zipfile/test_core.py @@ -1,3 +1,4 @@ +import _pyio import array import contextlib import importlib.util @@ -3454,5 +3455,87 @@ def test_too_short(self): b"zzz", zipfile._Extra.strip(b"zzz", (self.ZIP64_EXTRA,))) +class StatIO(_pyio.BytesIO): + """Buffer which remembers the number of bytes that were read.""" + + def __init__(self): + super().__init__() + self.bytes_read = 0 + + def read(self, size=-1): + bs = super().read(size) + self.bytes_read += len(bs) + return bs + + +class StoredZipExtFileRandomReadTest(unittest.TestCase): + """Tests whether an uncompressed, unencrypted zip entry can be randomly + seek and read without reading redundant bytes.""" + def test_stored_seek_and_read(self): + + sio = StatIO() + # 20000 bytes + txt = b'0123456789' * 2000 + + # The seek length must be greater than ZipExtFile.MIN_READ_SIZE + # as `ZipExtFile._read2()` reads in blocks of this size and we + # need to seek out of the buffered data + read_buffer_size = zipfile.ZipExtFile.MIN_READ_SIZE + self.assertGreaterEqual(10002, read_buffer_size) # for forward seek test + self.assertGreaterEqual(5003, read_buffer_size) # for backward seek test + # The read length must be less than MIN_READ_SIZE, since we assume that + # only 1 block is read in the test. + read_length = 100 + self.assertGreaterEqual(read_buffer_size, read_length) # for read() calls + + with zipfile.ZipFile(sio, "w", compression=zipfile.ZIP_STORED) as zipf: + zipf.writestr("foo.txt", txt) + + # check random seek and read on a file + with zipfile.ZipFile(sio, "r") as zipf: + with zipf.open("foo.txt", "r") as fp: + # Test this optimized read hasn't rewound and read from the + # start of the file (as in the case of the unoptimized path) + + # forward seek + old_count = sio.bytes_read + forward_seek_len = 10002 + current_pos = 0 + fp.seek(forward_seek_len, os.SEEK_CUR) + current_pos += forward_seek_len + self.assertEqual(fp.tell(), current_pos) + self.assertEqual(fp._left, fp._compress_left) + arr = fp.read(read_length) + current_pos += read_length + self.assertEqual(fp.tell(), current_pos) + self.assertEqual(arr, txt[current_pos - read_length:current_pos]) + self.assertEqual(fp._left, fp._compress_left) + read_count = sio.bytes_read - old_count + self.assertLessEqual(read_count, read_buffer_size) + + # backward seek + old_count = sio.bytes_read + backward_seek_len = 5003 + fp.seek(-backward_seek_len, os.SEEK_CUR) + current_pos -= backward_seek_len + self.assertEqual(fp.tell(), current_pos) + self.assertEqual(fp._left, fp._compress_left) + arr = fp.read(read_length) + current_pos += read_length + self.assertEqual(fp.tell(), current_pos) + self.assertEqual(arr, txt[current_pos - read_length:current_pos]) + self.assertEqual(fp._left, fp._compress_left) + read_count = sio.bytes_read - old_count + self.assertLessEqual(read_count, read_buffer_size) + + # eof flags test + fp.seek(0, os.SEEK_END) + fp.seek(12345, os.SEEK_SET) + current_pos = 12345 + arr = fp.read(read_length) + current_pos += read_length + self.assertEqual(arr, txt[current_pos - read_length:current_pos]) + + if __name__ == "__main__": unittest.main() diff --git a/Lib/zipfile/__init__.py b/Lib/zipfile/__init__.py index 8b636094fad3ca..82e307f78e8e3d 100644 --- a/Lib/zipfile/__init__.py +++ b/Lib/zipfile/__init__.py @@ -1163,13 +1163,15 @@ def seek(self, offset, whence=os.SEEK_SET): self._offset = buff_offset read_offset = 0 # Fast seek uncompressed unencrypted file - elif self._compress_type == ZIP_STORED and self._decrypter is None and read_offset > 0: + elif self._compress_type == ZIP_STORED and self._decrypter is None and read_offset != 0: # disable CRC checking after first seeking - it would be invalid self._expected_crc = None # seek actual file taking already buffered data into account read_offset -= len(self._readbuffer) - self._offset self._fileobj.seek(read_offset, os.SEEK_CUR) self._left -= read_offset + self._compress_left -= read_offset + self._eof = self._left <= 0 read_offset = 0 # flush read buffer self._readbuffer = b'' diff --git a/Misc/NEWS.d/next/Library/2024-12-21-03-20-12.gh-issue-128131.QpPmNt.rst b/Misc/NEWS.d/next/Library/2024-12-21-03-20-12.gh-issue-128131.QpPmNt.rst new file mode 100644 index 00000000000000..f4c4ebce10729c --- /dev/null +++ b/Misc/NEWS.d/next/Library/2024-12-21-03-20-12.gh-issue-128131.QpPmNt.rst @@ -0,0 +1,2 @@ +Completely support random access of uncompressed unencrypted read-only +zip files obtained by :meth:`ZipFile.open <zipfile.ZipFile.open>`.

1 0