The @test.support.requires_legacy_unicode_capi() decorators uses from _testcapi import unicode_legacy_string, but this module attribute has been remove since Python 3.12 by commit f9c9354, and so all tests using the decorator are now always skipped.
Should we now remove tests using @requires_legacy_unicode_capi() decorator? Or are these tests kept on purpose?
For example, this test is now always skipped:
$ ./python -m test test_capi -m test.test_capi.test_getargs.String_TestCase.test_Z -v
test_Z (test.test_capi.test_getargs.String_TestCase.test_Z) ... skipped 'requires legacy Unicode C API'
See also issue #92536 and PEP 623: Remove wstr from Unicode.
cc @methane @serhiy-storchaka
Linked PRs
The
@test.support.requires_legacy_unicode_capi()decorators usesfrom _testcapi import unicode_legacy_string, but this module attribute has been remove since Python 3.12 by commit f9c9354, and so all tests using the decorator are now always skipped.Should we now remove tests using
@requires_legacy_unicode_capi()decorator? Or are these tests kept on purpose?For example, this test is now always skipped:
See also issue #92536 and PEP 623: Remove wstr from Unicode.
cc @methane @serhiy-storchaka
Linked PRs