python / cpython

…21021)

…H-21054)

This commit changes the parsing of f-string expressions with the new parser. The parser gets pre-fed with the location of the expression itself (not the f-string, which was what we were doing before). This allows us to completely skip the shifting of the AST nodes after the parsing is completed.

…assignment rule (GH-20387)

tracemalloc_get_frame() checked filename == NULL two times in a row.

Signed-off-by: Christian Heimes <christian@python.org>

They are undocumented and deprecated since Python 3.3.

…Error (GH-21084)

Prefix the error message with `fstring: `, when parsing an f-string expression throws a `SyntaxError`.

…lon (GH-21160)

They are deprecated since Python 3.0.

Use new test.support helper submodules in tests:

* distutils tests
* test_buffer
* test_compile
* test_filecmp
* test_fileinput
* test_readline
* test_smtpnet
* test_structmembers
* test_tools

…racters. (GH-21035)

Always create the empty bytes string singleton.

Optimize PyBytes_FromStringAndSize(str, 0): it no longer has to check
if the empty string singleton was created or not, it is always
available.

Add functions:

* _PyBytes_Init()
* bytes_get_empty(), bytes_new_empty()
* bytes_create_empty_string_singleton()
* unicode_create_empty_string_singleton()

_Py_unicode_state: rename empty structure member to empty_string.

Adds a simple check for whether or not the package is being installed in the GUI or using installer on the command line. This addresses an issue where CLI-based software management tools (such as Munki) unexpectedly open Finder windows into a GUI session during installation runs.

* Fix support of non-ASCII names in functions OpenDatabase()
  and init_database().
* Fix support of non-ASCII SQL in method Database.OpenView().

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

Automerge-Triggered-By: @merwok

…H-20985)

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().

Automerge-Triggered-By: @matrixise