I've done the implementation for both non-chunked and chunked reads. I haven't benchmarked chunked reads because I don't currently have a convenient way to generate a high-bandwidth chunked stream, but I don't see any reason that it shouldn't enjoy the same benefits that the non-chunked case does. I've used the benchmark attached to the bpo bug to verify that performance now matches the unsized read case.
Automerge-Triggered-By: @methane
In absence of explicit declaration, resource compiler uses system
codepage. When this codepage is DBCS or UTF-8, Python's copyright
string is corrupted, because it contains copyright sign encoded
as \xA9.
The fix is to explicitly declare codepage 1252.
Py_InitializeFromConfig() now always creates the empty tuple
singleton as soon as possible.
Optimize PyTuple_New(0): it no longer has to check if the empty tuple
was created or not, it is always creatd.
* Add tuple_create_empty_tuple_singleton() function.
* Add tuple_get_empty() function.
* Remove state parameter of tuple_alloc().
Each interpreter now has its own Unicode latin1 singletons.
Remove "ifdef EXPERIMENTAL_ISOLATED_SUBINTERPRETERS"
and "ifdef LATIN1_SINGLETONS": always enable latin1 singletons.
Optimize unicode_result_ready(): only attempt to get a latin1
singleton for PyUnicode_1BYTE_KIND.
Replace MIDL-generated file with manual GUID definition.
Use the same .def file for release and debug builds.
Update setup build to support latest toolset
Functions of unicodeobject.c, like PyUnicode_New(), no longer check
if the empty Unicode singleton has been initialized or not. Consider
that it is always initialized. The Unicode API must not be used
before _PyUnicode_Init() or after _PyUnicode_Fini().
Each interpreter now has its own MemoryError free list: it is not
longer shared by all interpreters.
Add _Py_exc_state structure and PyInterpreterState.exc_state member.
Move also errnomap into _Py_exc_state.
Use linker comment #pragma and preprocessor for re-exporting stable
API functions and variables.
Module definition file, custom build targets and entry point code
become unnecessary and can be removed.
This change also fixes missing _PyErr_BadInternalCall export on x86.
Remove auto-generated resource header. Pass definitions required
by resource files (ORIGINAL_FILENAME and FIELD3) directly to resource
compiler.
Remove unused MS_DLL_ID resource string and related dead code.
* Add missing header files to pythoncore.
* Add missing file filters ("Resource Files" in particular) to
all projects.
* Add new sub-filters for private headers in pythoncore and
for 3rd party source files.
* Add missing _zoneinfo configurations in pcbuild.sln.
* Update bdist_wininst with the new zlib location.
* Revert "bpo-40521: Make the empty frozenset per interpreter (GH-21068)"
This reverts commit 261cfedf76.
* bpo-40521: Empty frozensets are no longer singletons
* Complete the removal of the frozenset singleton
Each interpreter now has its own empty bytes string and single byte
character singletons.
Replace STRINGLIB_EMPTY macro with STRINGLIB_GET_EMPTY() macro.
Each interpreter now has its own dict free list:
* Move dict free lists into PyInterpreterState.
* Move PyDict_MAXFREELIST define to pycore_interp.h
* Add _Py_dict_state structure.
* Add tstate parameter to _PyDict_ClearFreeList() and _PyDict_Fini().
* In debug mode, ensure that the dict free lists are not used after
_PyDict_Fini() is called.
* Remove "#ifdef EXPERIMENTAL_ISOLATED_SUBINTERPRETERS".
* Add a glossary entry for the term "callback"
* Link to it in loop.call_soon() and in the "Concurrency and Multithreading" section
Co-authored-by: Kyle Stanley <aeros167@gmail.com>
Unexpected errors in calling the __iter__ method are no longer
masked by TypeError in the "in" operator and functions
operator.contains(), operator.indexOf() and operator.countOf().
`GET_INVALID_TARGET` might unexpectedly return `NULL`, which if not
caught will cause a SEGFAULT. Therefore, this commit introduces a new
inline function `RAISE_SYNTAX_ERROR_INVALID_TARGET` that always
checks for `GET_INVALID_TARGET` returning NULL and can be used in
the grammar, replacing the long C ternary operation used till now.
There are a bunch of other fd: int uses in this file, I expect many if not
all of them would be better off using the fildes converter. This particular
one was flagged by Coverity as it presumably flags fpathconf as not accepting
negative fds. I'd expect the other fd's to have been flagged as well
otherwise.
I'm marking this one as skip news as it really is a no-op.
Serhiy Storchaka
Bruce Merry
Arisaka97
Nikita Nemkin
Victor Stinner
Jürgen Gmach
Gareth Rees
Anthony Sottile
Raymond Hettinger
Ram Rachum
WildCard65
Roger Iyengar
Dong-hee Na
Krishna Chivukula
Christian Heimes
Gregory P. Smith
Joannah Nanjekye
Lysandros Nikolaou