gh-154757: Remove the deprecated Py_UNICODE type by fedonman · Pull Request #154759 · python/cpython

fedonman · 2026-07-27T08:24:27Z

Py_UNICODE is a typedef to wchar_t, deprecated by :pep:393 since 3.3 and at runtime since 3.13, listed for removal in 3.16. This removes it, and PY_UNICODE_TYPE, the same typedef deprecated alongside it in 3.13.

Nothing in CPython used either typedef. The Argument Clinic Py_UNICODE converter is unrelated: it generates const wchar_t *. Neither is in the stable ABI.

Documentation references are kept with the link suppressed (:c:type:!Py_UNICODE``), as GH-133657 did for PyWeakref_GetObject, and the anchor is recorded in `Doc/tools/removed-ids.txt`.

This covers one of the three groups in gh-154757 and one box in gh-133644. The other two groups are not touched here: GH-136277 and GH-133654 are already open for them, and Py_SetProgramName and PySys_SetArgv were reverted once before in gh-117987.

Verified with a full build (make -j), test_capi and test_clinic (1986 tests, SUCCESS). A file using Py_UNICODE or PY_UNICODE_TYPE now fails to compile with "unknown type name", while wchar_t still works. Sphinx is not installed here, so the docs build is unverified locally.

Py_UNICODE was a typedef to wchar_t, deprecated by PEP 393 since Python 3.3 and at runtime since Python 3.13, and scheduled for removal in 3.16. Nothing in CPython used it: the Argument Clinic converter of the same name generates const wchar_t *, and the typedef is not part of the stable ABI. The documentation references to it are kept, with the link suppressed, as was done for PyWeakref_GetObject in pythonGH-133657.

read-the-docs-community · 2026-07-27T08:28:20Z

Documentation build overview

📚 cpython-previews | 🛠️ Build #34769474 | 📁 Comparing 9e4a944 against main (993a0c6)

🔍 Preview build

403 files changed · + 9 added · ± 394 modified

+ Added

± Modified

vstinner

IMO PY_UNICODE_TYPE should also be removed in the same PR.

encukou · 2026-09-24T14:37:57Z

Why not keep both? Simple typedefs have virtually no maintenance cost.

vstinner · 2026-09-24T14:46:00Z

Py_UNICODE is a ghost from the past. It was only relevant when a large part of the PyUnicode C API used this type (as Py_UNICODE* strings). In Python 3.16, using Py_UNICODE is no longer relevant. Using wchar_t makes the code less Pythonic (easier to understand for people not used to Python C API).

Removing Py_UNICODE can help C extensions maintainers to discover bugs in their old code (especially if the code was written with Py_UNICODE APIs and was updated later to newer APIs)

fedonman · 2026-09-26T01:27:18Z

I left PY_UNICODE_TYPE out because only Py_UNICODE is on the 3.16 removal list. Unlike Py_UNICODE_WIDE, both typedefs have given a compiler deprecation warning since 3.13. I'll follow whatever you two settle on: remove one, both, or neither.

encukou · 2026-09-29T12:51:14Z

Removing Py_UNICODE can help C extensions maintainers to discover bugs in their old code (especially if the code was written with Py_UNICODE APIs and was updated later to newer APIs)

That seems unlikely to me. What kinds of bugs is this likely to help discover?

Removing Py_UNICODE means that any project that uses it is forced to change code. If they find the porting instructions in the removal notice, they'll replace Py_UNICODE by wchar_t. I don't think that has much potential for finding bugs. (But, of course, each such forced trivial change is a chance for the project to re-evaluate whether Python support is worth the churn.)

vstinner · 2026-09-29T14:18:01Z

@encukou: If we don't remove Py_UNICODE in Python 3.16, we should at least postpone its removal in the docs. Does it mean that we should keep forever?

I searched for (Py_UNICODE|PY_UNICODE_TYPE)\b regex in PyPI top 15,000 projects and sadly I found 29 projects:

Distance (0.1.3)
PyICU-binary (2.7.4)
PyQt5 (5.15.11)
cffi (2.1.1)
cu2qu-1.6.7.post2
cython (3.3.0)
genshi (0.7.11)
igwn_ligolw (2.1.1)
line_profiler (5.0.2)
lxml (6.1.3)
minorminer (0.2.22)
netifaces (0.11.0)
netifaces_plus (0.12.5)
nuitka (4.2.1)
orjson (3.12.0)
pyScss (1.4.0)
pyahocorasick (2.3.1)
pygments (2.21.0)
pyicu (2.16.2)
pyobjc_core (12.2.2)
pypyodbc (1.3.6)
pystack (1.7.1)
pywin32-ctypes (0.2.3)
pywinpty (3.0.5)
rcssmin (1.2.2)
rjsmin (1.2.5)
simplejson (4.1.2)
tiled (0.2.18)
twisted_iocpsupport (25.10.1)

Notes:

Cython 3.0.9 (2024-03-05) changelog says "The deprecated Py_UNICODE type is no longer used, unless required by user code."
cffi: The matching lines comes from Python 2 code (not used on Python 3.3 and newer).
lxml: Py_UNICODE is only used on Python 3.11 and older.
A single line mention PY_UNICODE_TYPE and it comes from a copy of CPython unicodeobject.h header file.

I didn't check if Py_UNICODE is actually used in other projects:

Details

PYPI-2026-09-17/cffi-2.1.1.tar.gz: cffi-2.1.1/src/c/wchar_helper.h: register Py_UNICODE *u;
PYPI-2026-09-17/cffi-2.1.1.tar.gz: cffi-2.1.1/src/c/wchar_helper.h: return PyUnicode_FromUnicode((const Py_UNICODE *)w, size);
PYPI-2026-09-17/cffi-2.1.1.tar.gz: cffi-2.1.1/src/c/wchar_helper.h: return PyUnicode_FromUnicode((const Py_UNICODE *)w, size);
PYPI-2026-09-17/cffi-2.1.1.tar.gz: cffi-2.1.1/src/c/wchar_helper.h: Py_UNICODE *u_base = PyUnicode_AS_UNICODE(result);
PYPI-2026-09-17/cffi-2.1.1.tar.gz: cffi-2.1.1/src/c/wchar_helper.h: Py_UNICODE *u = u_base;
PYPI-2026-09-17/cffi-2.1.1.tar.gz: cffi-2.1.1/src/c/wchar_helper.h: Py_UNICODE *u = PyUnicode_AS_UNICODE(unicode);
PYPI-2026-09-17/cffi-2.1.1.tar.gz: cffi-2.1.1/src/c/wchar_helper.h: Py_UNICODE *u = PyUnicode_AS_UNICODE(unicode);
PYPI-2026-09-17/cffi-2.1.1.tar.gz: cffi-2.1.1/src/c/wchar_helper.h: Py_UNICODE *u = PyUnicode_AS_UNICODE(unicode);
PYPI-2026-09-17/cffi-2.1.1.tar.gz: cffi-2.1.1/src/c/wchar_helper.h: Py_UNICODE *u = PyUnicode_AS_UNICODE(unicode);
PYPI-2026-09-17/cffi-2.1.1.tar.gz: cffi-2.1.1/src/c/wchar_helper.h: Py_UNICODE *u = PyUnicode_AS_UNICODE(unicode);
PYPI-2026-09-17/cffi-2.1.1.tar.gz: cffi-2.1.1/src/c/wchar_helper.h: Py_UNICODE *u = PyUnicode_AS_UNICODE(unicode);
PYPI-2026-09-17/cu2qu-1.6.7.post2.zip: cu2qu-1.6.7.post2/Lib/cu2qu/cython.py: int_types = ['char', 'short', 'Py_UNICODE', 'int', 'Py_UCS4', 'long', 'longlong', 'Py_ssize_t', 'size_t']
PYPI-2026-09-17/cu2qu-1.6.7.post2.zip: cu2qu-1.6.7.post2/Lib/cu2qu/cython.py: if name not in ('Py_UNICODE', 'Py_UCS4') and not name.endswith('size_t'):
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/Code.py: # return a Py_UNICODE[] constant, creating a new one if necessary
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/Code.py: # Generate legacy Py_UNICODE[] constants.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/Code.py: decls_writer.putln("static Py_UNICODE %s[] = { %s };" % (cname, utf32_array))
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/Code.py: decls_writer.putln("static Py_UNICODE %s[] = { %s };" % (cname, utf16_array))
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/ExprNodes.py: (unicode_type, PyrexTypes.c_char_ptr_type): "Unicode objects only support coercion to Py_UNICODE*.",
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/ExprNodes.py: (unicode_type, PyrexTypes.c_const_char_ptr_type): "Unicode objects only support coercion to Py_UNICODE*.",
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/ExprNodes.py: (unicode_type, PyrexTypes.c_uchar_ptr_type): "Unicode objects only support coercion to Py_UNICODE*.",
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/ExprNodes.py: (unicode_type, PyrexTypes.c_const_uchar_ptr_type): "Unicode objects only support coercion to Py_UNICODE*.",
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/ExprNodes.py: (bytes_type, PyrexTypes.c_py_unicode_ptr_type): "Cannot convert 'bytes' object to Py_UNICODE*, use 'str'.",
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/ExprNodes.py: "Cannot convert 'bytes' object to Py_UNICODE*, use 'str'."),
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/ExprNodes.py: error(self.pos, "Bytes literals cannot coerce to Py_UNICODE/Py_UCS4, use a unicode literal instead.")
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/ExprNodes.py: "surrogate pairs can be coerced into Py_UCS4/Py_UNICODE.")
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/ExprNodes.py: "than Py_UCS4/Py_UNICODE (for characters), Py_UNICODE* "
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/ExprNodes.py: # sliced Py_UNICODE* strings must coerce to Python
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/ExprNodes.py: # we infer Py_UNICODE/Py_UCS4 for unicode strings in some
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/ExprNodes.py: # Py_UNICODE/Py_UCS4 will automatically coerce to a unicode string
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/ExprNodes.py: "Py_UNICODE* has been removed in Python 3.12. This conversion to a "
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/ExprNodes.py: "Py_UNICODE* will no longer compile in the latest Python versions. "
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/Optimize.py: len(Py_UNICODE) by the equivalent Py_UNICODE_strlen() and
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/Optimize.py: """Unpack ord(Py_UNICODE) and ord('X').
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/Parsing.py: "Py_UNICODE" : (0, 0),
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/PyrexTypes.py: #  is_unicode_char       boolean     Is either Py_UCS4 or Py_UNICODE
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/PyrexTypes.py: # Py_UNICODE
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/PyrexTypes.py: # Py_UNICODE coerces from and to single character unicode strings,
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/PyrexTypes.py: # Py_UNICODE is 0..1114111, which is checked when converting from
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/PyrexTypes.py: from_py_function = "__Pyx_PyObject_AsPy_UNICODE"
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/PyrexTypes.py: return "Py_UNICODE"
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/PyrexTypes.py: #  is_pyunicode_ptr  bool           Pointer is a Py_UNICODE*.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/PyrexTypes.py: "Py_UNICODE": (0, c_py_unicode_type),
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/PyrexTypes.py: # Py_UNICODE behaves more like a string than an int
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Compiler/StringEncoding.py: """Create Py_UNICODE[] representation of a given unicode string.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Debugger/libpython.py: _type_Py_UNICODE = gdb.lookup_type('Py_UNICODE')
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Debugger/libpython.py: return _type_Py_UNICODE.sizeof
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Debugger/libpython.py: # Gather a list of ints from the Py_UNICODE array; these are either
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Debugger/libpython.py: # A more elaborate routine if sizeof(Py_UNICODE) is 2 in the
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Debugger/libpython.py: # This splits surrogate pairs if sizeof(Py_UNICODE) is 2 here (in gdb).
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Debugger/libpython.py: # If sizeof(Py_UNICODE) is 2 here (in gdb), join
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/array.pxd: Py_UNICODE *as_pyunicodes
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/long.pxd: object PyLong_FromUnicode(Py_UNICODE *u, Py_ssize_t length, int base)
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: # Return a pointer to the internal Py_UNICODE buffer of the
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: Py_UNICODE* PyUnicode_AS_UNICODE(object o)
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: # Used to return a Py_UNICODE value before Py3.3.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: # Used to return a Py_UNICODE value before Py3.3.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: # Used to return a Py_UNICODE value before Py3.3.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: # Create a Unicode Object from the Py_UNICODE buffer u of the
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: unicode PyUnicode_FromUnicode(Py_UNICODE *u, Py_ssize_t size)
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: # Py_UNICODE buffer, NULL if unicode is not a Unicode object.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: Py_UNICODE* PyUnicode_AsUnicode(object o) except NULL
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: # Py_UNICODE type is identical to the system's wchar_t.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: # Encode the Py_UNICODE buffer of the given size and return a
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: object PyUnicode_Encode(Py_UNICODE *s, Py_ssize_t size,
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: # Encode the Py_UNICODE buffer of the given size using UTF-8 and
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: bytes PyUnicode_EncodeUTF8(Py_UNICODE *s, Py_ssize_t size, char *errors)
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: # If Py_UNICODE_WIDE is defined, a single Py_UNICODE value may get
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: # Py_UNICODE values is interpreted as an UCS-2 character.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: bytes PyUnicode_EncodeUTF16(Py_UNICODE *s, Py_ssize_t size, char *errors, int byteorder)
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: # Encode the Py_UNICODE buffer of the given size using
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: object PyUnicode_EncodeUnicodeEscape(Py_UNICODE *s, Py_ssize_t size)
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: # Encode the Py_UNICODE buffer of the given size using
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: object PyUnicode_EncodeRawUnicodeEscape(Py_UNICODE *s, Py_ssize_t size, char *errors)
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: # Encode the Py_UNICODE buffer of the given size using Latin-1 and
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: bytes PyUnicode_EncodeLatin1(Py_UNICODE *s, Py_ssize_t size, char *errors)
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: # Encode the Py_UNICODE buffer of the given size using ASCII and
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: bytes PyUnicode_EncodeASCII(Py_UNICODE *s, Py_ssize_t size, char *errors)
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: # Encode the Py_UNICODE buffer of the given size using the given
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: object PyUnicode_EncodeCharmap(Py_UNICODE *s, Py_ssize_t size, object mapping, char *errors)
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: # Translate a Py_UNICODE buffer of the given length by applying a
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: object PyUnicode_TranslateCharmap(Py_UNICODE *s, Py_ssize_t size,
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: # Encode the Py_UNICODE buffer of the given size using MBCS and
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Includes/cpython/unicode.pxd: bytes PyUnicode_EncodeMBCS(Py_UNICODE *s, Py_ssize_t size, char *errors)
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Shadow.py: 'Py_UNICODE',
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Shadow.py: if name not in ('Py_UNICODE', 'Py_UCS4', 'Py_hash_t', 'ptrdiff_t') and not name.endswith('size_t'):
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Shadow.py: Py_UNICODE = py_int | str
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Shadow.py: const_Py_UNICODE : TypeAlias = const[py_int]
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Shadow.py: p_const_Py_UNICODE = pointer[const[py_int]]
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Shadow.py: pp_const_Py_UNICODE = pointer[pointer[const[py_int]]]
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Shadow.py: ppp_const_Py_UNICODE = pointer[pointer[pointer[const[py_int]]]]
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Shadow.py: p_Py_UNICODE = pointer[py_int]
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Shadow.py: pp_Py_UNICODE = pointer[pointer[py_int]]
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Shadow.py: ppp_Py_UNICODE = pointer[pointer[pointer[py_int]]]
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Tests/TestShadow.py: int_name in ['Py_UNICODE', 'Py_UCS4', 'Py_ssize_t',
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Utility/StringTools.c: static CYTHON_INLINE Py_ssize_t __Pyx_Py_UNICODE_ssize_strlen(const Py_UNICODE *u);/*proto*/
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Utility/StringTools.c: static CYTHON_INLINE Py_ssize_t __Pyx_Py_UNICODE_ssize_strlen(const Py_UNICODE *u) {
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Utility/StringTools.c: PyErr_SetString(PyExc_OverflowError, "Py_UNICODE string is too long");
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Utility/StringTools.c: static CYTHON_INLINE size_t __Pyx_Py_UNICODE_strlen(const Py_UNICODE *u); /* proto */
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Utility/StringTools.c: // Note: will not work in the limited API since Py_UNICODE is not available there.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Utility/StringTools.c: static CYTHON_INLINE size_t __Pyx_Py_UNICODE_strlen(const Py_UNICODE *u)
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Utility/StringTools.c: const Py_UNICODE *u_end = u;
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Utility/TypeConversion.c: static CYTHON_INLINE Py_UNICODE __Pyx_PyObject_AsPy_UNICODE(PyObject*);
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Utility/TypeConversion.c: static CYTHON_INLINE Py_UNICODE __Pyx_PyObject_AsPy_UNICODE(PyObject* x) {
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Utility/TypeConversion.c: "only single character unicode strings can be converted to Py_UNICODE, "
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Utility/TypeConversion.c: return (Py_UNICODE)-1;
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Utility/TypeConversion.c: "cannot convert negative value to Py_UNICODE");
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Utility/TypeConversion.c: return (Py_UNICODE)-1;
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Utility/TypeConversion.c: "value too large to convert to Py_UNICODE");
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Utility/TypeConversion.c: return (Py_UNICODE)-1;
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Utility/TypeConversion.c: return (Py_UNICODE)ival;
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/Cython/Utility/arrayarray.h: // Don't use Py_UNICODE ourselves in the union. This avoids deprecation warnings
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/runtests.py: 'run.py_unicode_strings',  # Py_UNICODE was removed
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/charptr_from_temp.pyx: cdef Py_UNICODE* cuptr
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/charptr_from_temp.pyx: 38:8: Py_UNICODE* has been removed in Python 3.12. This conversion to a Py_UNICODE* will no longer compile in the latest Python versions. Use Python C API functions like PyUnicode_AsWideCharString if you need to obtain a wchar_t* on Windows (and free the string manually after use).
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/charptr_from_temp.pyx: 41:8: Obtaining 'Py_UNICODE *' from externally modifiable global Python value
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/charptr_from_temp.pyx: 41:8: Py_UNICODE* has been removed in Python 3.12. This conversion to a Py_UNICODE* will no longer compile in the latest Python versions. Use Python C API functions like PyUnicode_AsWideCharString if you need to obtain a wchar_t* on Windows (and free the string manually after use).
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/charptr_from_temp.pyx: 44:10: Py_UNICODE* has been removed in Python 3.12. This conversion to a Py_UNICODE* will no longer compile in the latest Python versions. Use Python C API functions like PyUnicode_AsWideCharString if you need to obtain a wchar_t* on Windows (and free the string manually after use).
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/e_strcoerce.pyx: cdef Py_UNICODE u1 = u"\xFF"   # works
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/e_strcoerce.pyx: 11:14: Unicode literals do not support coercion to C types other than Py_UCS4/Py_UNICODE (for characters), Py_UNICODE* (for strings) or char*/void* (for auto-encoded strings).
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/py_ucs4_type_errors.pyx: 8:21: Only single-character Unicode string literals or surrogate pairs can be coerced into Py_UCS4/Py_UNICODE.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/py_ucs4_type_errors.pyx: 11:21: Only single-character Unicode string literals or surrogate pairs can be coerced into Py_UCS4/Py_UNICODE.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/py_ucs4_type_errors.pyx: 17:37: Bytes literals cannot coerce to Py_UNICODE/Py_UCS4, use a unicode literal instead.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/py_unicode_type_errors.pyx: cdef Py_UNICODE char_ASCII = u'A'
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/py_unicode_type_errors.pyx: cdef Py_UNICODE char_KLINGON = u'\uF8D2'
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/py_unicode_type_errors.pyx: cdef Py_UNICODE c = u'AB'
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/py_unicode_type_errors.pyx: cdef Py_UNICODE c = u'A\uF8D2'
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/py_unicode_type_errors.pyx: cdef Py_UNICODE c = b'AB'
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/py_unicode_type_errors.pyx: cdef Py_UNICODE char_bytes_latin1 = b'\xf6'
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/py_unicode_type_errors.pyx: 8:24: Only single-character Unicode string literals or surrogate pairs can be coerced into Py_UCS4/Py_UNICODE.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/py_unicode_type_errors.pyx: 11:24: Only single-character Unicode string literals or surrogate pairs can be coerced into Py_UCS4/Py_UNICODE.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/py_unicode_type_errors.pyx: 17:40: Bytes literals cannot coerce to Py_UNICODE/Py_UCS4, use a unicode literal instead.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/string_assignments.pyx: cdef Py_UNICODE* cu1 = u1
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/string_assignments.pyx: cdef Py_UNICODE* cu_f1 = c1
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/string_assignments.pyx: cdef Py_UNICODE* cu_f2 = b1
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/string_assignments.pyx: cdef Py_UNICODE* cu_f3 = s1
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/string_assignments.pyx: cdef Py_UNICODE* cu_f4 = b"abc"
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/string_assignments.pyx: 36:20: Unicode literals do not support coercion to C types other than Py_UCS4/Py_UNICODE (for characters), Py_UNICODE* (for strings) or char*/void* (for auto-encoded strings).
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/string_assignments.pyx: 37:20: Unicode objects only support coercion to Py_UNICODE*.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/string_assignments.pyx: 38:20: Unicode objects only support coercion to Py_UNICODE*.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/string_assignments.pyx: 40:25: Cannot assign type 'char *' to 'Py_UNICODE *'
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/string_assignments.pyx: 41:25: Cannot convert 'bytes' object to Py_UNICODE*, use 'str'.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/string_assignments.pyx: 43:25: Cannot convert 'bytes' object to Py_UNICODE*, use 'str'.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/errors/string_assignments.pyx: 77:20: Unicode literals do not support coercion to C types other than Py_UCS4/Py_UNICODE (for characters), Py_UNICODE* (for strings) or char*/void* (for auto-encoded strings).
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/limited_api_bugs.txt: # Py_UNICODE
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/builtin_ord.pyx: def ord_Py_UNICODE(unicode s):
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/builtin_ord.pyx: >>> ord_Py_UNICODE(uspace)
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/builtin_ord.pyx: cdef Py_UNICODE u
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/for_in_string.pyx: cdef Py_UNICODE c
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/for_in_string.pyx: cdef Py_UNICODE c
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/fstring.pyx: def format_c_values(Py_UCS4 uchar, Py_UNICODE pyunicode):
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/inop.pyx: def m_unicode_literal(Py_UNICODE a):
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/inop.pyx: def conditional_unicode(Py_UNICODE a):
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/notinop.pyx: def conditional_unicode(Py_UNICODE a):
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_strings.pyx: cdef assert_Py_UNICODE_equal(const Py_UNICODE* u1, const Py_UNICODE* u2):
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_strings.pyx: ctypedef Py_UNICODE* LPWSTR
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_strings.pyx: cdef Py_UNICODE* c_pu_str = u"unicode\u1234"
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_strings.pyx: cdef Py_UNICODE[42] c_pu_arr
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_strings.pyx: cdef Py_UNICODE* c_pu_empty = u""
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_strings.pyx: cdef Py_UNICODE* c_pu_wide_literal = u'\U00020000\U00020001'
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_strings.pyx: memcpy(c_pu_arr, c_pu_str, sizeof(Py_UNICODE) * (len(uobj) + 1))
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_strings.pyx: cdef Py_UNICODE ch = uobj[1]  # Py_UCS4 is unsigned, Py_UNICODE is usually signed.
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_strings.pyx: assert sizeof(c_pu_arr) == sizeof(Py_UNICODE) * 42
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_strings.pyx: if sizeof(Py_UNICODE) >= 4:
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_strings.pyx: assert_Py_UNICODE_equal(<Py_UNICODE*>u"u", u)
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_strings.pyx: assert_Py_UNICODE_equal(<Py_UNICODE*>u"nicode", u)
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_strings.pyx: assert_Py_UNICODE_equal(<Py_UNICODE*>u"n", u)
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_strings.pyx: assert_Py_UNICODE_equal(uwide_literal, <Py_UNICODE*>c_pu_wide_literal)
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_strings.pyx: assert len(<Py_UNICODE*>u"abc\0") == 3
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_type.pyx: cdef Py_UNICODE char_ASCII = u'A'
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_type.pyx: cdef Py_UNICODE char_KLINGON = u'\uF8D2'
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_type.pyx: def unicode_ordinal(Py_UNICODE i):
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_type.pyx: ValueError: only single character unicode strings can be converted to Py_UNICODE, got length 0
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_type.pyx: ValueError: only single character unicode strings can be converted to Py_UNICODE, got length 2
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_type.pyx: def ord_pyunicode(Py_UNICODE x):
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_type.pyx: def unicode_type_methods(Py_UNICODE uchar):
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_type.pyx: def unicode_methods(Py_UNICODE uchar):
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_type.pyx: def len_uchar(Py_UNICODE uchar):
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_type.pyx: def index_uchar(Py_UNICODE uchar, Py_ssize_t i):
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/py_unicode_type.pyx: def uchar_lookup_in_dict(obj, Py_UNICODE uchar):
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/unicode_indexing.pyx: return <Py_UNICODE>(u"12345"[i])
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/unicode_indexing.pyx: cdef Py_UNICODE result = u"12345"[i]
PYPI-2026-09-17/cython-3.3.0.tar.gz: cython-3.3.0/tests/run/unicode_indexing.pyx: cdef Py_UNICODE result = u"12345"[i]
PYPI-2026-09-17/Distance-0.1.3.tar.gz: distance/cdistance/distance.h: typedef Py_UNICODE unicode;
PYPI-2026-09-17/genshi-0.7.11.tar.gz: genshi-0.7.11/genshi/_speedups.c: Py_UNICODE *inp, *outp;
PYPI-2026-09-17/igwn_ligolw-2.1.1.tar.gz: igwn_ligolw-2.1.1/debian/changelog: * tokenizer drop use in Py_UNICODE in favour of wchar_t
PYPI-2026-09-17/line_profiler-5.0.2.tar.gz: line_profiler-5.0.2/line_profiler/python25.pxd: ctypedef unsigned short Py_UNICODE
PYPI-2026-09-17/line_profiler-5.0.2.tar.gz: line_profiler-5.0.2/line_profiler/python25.pxd: object              PyInt_FromUnicode               (Py_UNICODE*, Py_ssize_t, int)
PYPI-2026-09-17/line_profiler-5.0.2.tar.gz: line_profiler-5.0.2/line_profiler/python25.pxd: object              PyLong_FromUnicode              (Py_UNICODE*, Py_ssize_t, int)
PYPI-2026-09-17/lxml-6.1.3.tar.gz: lxml-6.1.3/CHANGES.txt: to the native UTF-8 encoding directly, instead of going through ``Py_UNICODE`` /
PYPI-2026-09-17/lxml-6.1.3.tar.gz: lxml-6.1.3/src/lxml/parser.pxi: # name of Python Py_UNICODE encoding as known to libxml2
PYPI-2026-09-17/lxml-6.1.3.tar.gz: lxml-6.1.3/src/lxml/parser.pxi: cdef Py_UNICODE *uchars = [c'<', c't', c'e', c's', c't', c'/', c'>']
PYPI-2026-09-17/lxml-6.1.3.tar.gz: lxml-6.1.3/src/lxml/parser.pxi: enc = _findEncodingName(buffer, sizeof(Py_UNICODE) * 7)
PYPI-2026-09-17/lxml-6.1.3.tar.gz: lxml-6.1.3/src/lxml/parser.pxi: # Python 3.12 removed support for "Py_UNICODE".
PYPI-2026-09-17/lxml-6.1.3.tar.gz: lxml-6.1.3/src/lxml/parser.pxi: # old Py_UNICODE string
PYPI-2026-09-17/lxml-6.1.3.tar.gz: lxml-6.1.3/src/lxml/parser.pxi: # old Py_UNICODE string
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/array.pxd: Py_UNICODE *as_pyunicodes
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/long.pxd: object PyLong_FromUnicode(Py_UNICODE *u, Py_ssize_t length, int base)
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: # Return a pointer to the internal Py_UNICODE buffer of the
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: Py_UNICODE* PyUnicode_AS_UNICODE(object o)
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: # Used to return a Py_UNICODE value before Py3.3.
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: # Used to return a Py_UNICODE value before Py3.3.
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: # Used to return a Py_UNICODE value before Py3.3.
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: # Create a Unicode Object from the Py_UNICODE buffer u of the
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: unicode PyUnicode_FromUnicode(Py_UNICODE *u, Py_ssize_t size)
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: # Py_UNICODE buffer, NULL if unicode is not a Unicode object.
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: Py_UNICODE* PyUnicode_AsUnicode(object o) except NULL
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: # Py_UNICODE type is identical to the system's wchar_t.
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: # Encode the Py_UNICODE buffer of the given size and return a
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: object PyUnicode_Encode(Py_UNICODE *s, Py_ssize_t size,
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: # Encode the Py_UNICODE buffer of the given size using UTF-8 and
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: bytes PyUnicode_EncodeUTF8(Py_UNICODE *s, Py_ssize_t size, char *errors)
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: # If Py_UNICODE_WIDE is defined, a single Py_UNICODE value may get
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: # Py_UNICODE values is interpreted as an UCS-2 character.
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: bytes PyUnicode_EncodeUTF16(Py_UNICODE *s, Py_ssize_t size, char *errors, int byteorder)
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: # Encode the Py_UNICODE buffer of the given size using
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: object PyUnicode_EncodeUnicodeEscape(Py_UNICODE *s, Py_ssize_t size)
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: # Encode the Py_UNICODE buffer of the given size using
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: object PyUnicode_EncodeRawUnicodeEscape(Py_UNICODE *s, Py_ssize_t size, char *errors)
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: # Encode the Py_UNICODE buffer of the given size using Latin-1 and
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: bytes PyUnicode_EncodeLatin1(Py_UNICODE *s, Py_ssize_t size, char *errors)
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: # Encode the Py_UNICODE buffer of the given size using ASCII and
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: bytes PyUnicode_EncodeASCII(Py_UNICODE *s, Py_ssize_t size, char *errors)
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: # Encode the Py_UNICODE buffer of the given size using the given
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: object PyUnicode_EncodeCharmap(Py_UNICODE *s, Py_ssize_t size, object mapping, char *errors)
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: # Translate a Py_UNICODE buffer of the given length by applying a
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: object PyUnicode_TranslateCharmap(Py_UNICODE *s, Py_ssize_t size,
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: # Encode the Py_UNICODE buffer of the given size using MBCS and
PYPI-2026-09-17/minorminer-0.2.22.tar.gz: minorminer-0.2.22/env/lib/python3.12/site-packages/Cython/Includes/cpython/unicode.pxd: bytes PyUnicode_EncodeMBCS(Py_UNICODE *s, Py_ssize_t size, char *errors)
PYPI-2026-09-17/netifaces-0.11.0.tar.gz: netifaces-0.11.0/netifaces.c: Py_UNICODE uc = 0;
PYPI-2026-09-17/netifaces_plus-0.12.5.tar.gz: netifaces_plus-0.12.5/netifaces/_netifaces.c: Py_UNICODE uc = 0;
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/build/include/nuitka/string_functions.h: NUITKA_MAY_BE_UNUSED static Py_UNICODE *Nuitka_UnicodeAsWideString(PyObject *str, Py_ssize_t *size) {
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/build/static_src/CompiledCodeHelpers.c: Py_UNICODE *buffer = PyUnicode_AS_UNICODE(object);
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/build/static_src/HelpersComparisonEq.c: const Py_UNICODE *data1 = a->str;
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/build/static_src/HelpersComparisonEq.c: const Py_UNICODE *data2 = b->str;
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/build/static_src/HelpersComparisonEq.c: int cmp = memcmp(data1, data2, len * sizeof(Py_UNICODE));
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/build/static_src/HelpersComparisonEq.c: const Py_UNICODE *data1 = a->str;
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/build/static_src/HelpersComparisonEq.c: const Py_UNICODE *data2 = b->str;
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/build/static_src/HelpersComparisonEq.c: int cmp = memcmp(data1, data2, len * sizeof(Py_UNICODE));
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/build/static_src/HelpersComparisonNe.c: const Py_UNICODE *data1 = a->str;
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/build/static_src/HelpersComparisonNe.c: const Py_UNICODE *data2 = b->str;
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/build/static_src/HelpersComparisonNe.c: int cmp = memcmp(data1, data2, len * sizeof(Py_UNICODE));
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/build/static_src/HelpersComparisonNe.c: const Py_UNICODE *data1 = a->str;
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/build/static_src/HelpersComparisonNe.c: const Py_UNICODE *data2 = b->str;
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/build/static_src/HelpersComparisonNe.c: int cmp = memcmp(data1, data2, len * sizeof(Py_UNICODE));
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/build/static_src/HelpersOperationInplaceAddUtils.c: operand2_size * sizeof(Py_UNICODE));
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/build/static_src/HelpersStrings.c: if (length > ((PY_SSIZE_T_MAX / (Py_ssize_t)sizeof(Py_UNICODE)) - 1)) {
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/build/static_src/HelpersStrings.c: Py_ssize_t new_size = sizeof(Py_UNICODE) * ((size_t)length + 1);
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/build/static_src/HelpersStrings.c: _PyUnicode_WSTR(unicode) = (Py_UNICODE *)NuitkaObject_Malloc(new_size);
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/code_generation/templates_c/HelperOperationComparisonUnicode.c.j2: const Py_UNICODE *data1 = a->str;
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/code_generation/templates_c/HelperOperationComparisonUnicode.c.j2: const Py_UNICODE *data2 = b->str;
PYPI-2026-09-17/nuitka-4.2.1.tar.gz: nuitka-4.2.1/nuitka/code_generation/templates_c/HelperOperationComparisonUnicode.c.j2: int cmp = memcmp(data1, data2, len * sizeof(Py_UNICODE));
PYPI-2026-09-17/orjson-3.12.0.tar.gz: orjson-3.12.0/include/cargo/pyo3-ffi-0.28.3/src/unicodeobject.rs: pub type Py_UNICODE = wchar_t;
PYPI-2026-09-17/pyahocorasick-2.3.1.tar.gz: pyahocorasick-2.3.1/src/AutomatonItemsIter.c: return PyUnicode_FromUnicode((Py_UNICODE*)(iter->buffer + 1), depth);
PYPI-2026-09-17/pygments-2.21.0.tar.gz: pygments-2.21.0/tests/examplefiles/c/ceval.c: Py_UNICODE *s = PyUnicode_AS_UNICODE(v);
PYPI-2026-09-17/pygments-2.21.0.tar.gz: pygments-2.21.0/tests/examplefiles/c/ceval.c.output: 'Py_UNICODE'  Name
PYPI-2026-09-17/pyicu-2.16.2.tar.gz: pyicu-2.16.2/bases.cpp: if (sizeof(Py_UNICODE) == sizeof(UChar))
PYPI-2026-09-17/pyicu-2.16.2.tar.gz: pyicu-2.16.2/bases.cpp: if (sizeof(Py_UNICODE) == sizeof(UChar))
PYPI-2026-09-17/pyicu-2.16.2.tar.gz: pyicu-2.16.2/bases.cpp: if (sizeof(Py_UNICODE) == sizeof(UChar))
PYPI-2026-09-17/pyicu-2.16.2.tar.gz: pyicu-2.16.2/bases.cpp: if (sizeof(Py_UNICODE) == sizeof(UChar))
PYPI-2026-09-17/pyicu-2.16.2.tar.gz: pyicu-2.16.2/bases.cpp: if (sizeof(Py_UNICODE) == sizeof(UChar))
PYPI-2026-09-17/pyicu-2.16.2.tar.gz: pyicu-2.16.2/bases.cpp: if (sizeof(Py_UNICODE) == sizeof(UChar))
PYPI-2026-09-17/pyicu-2.16.2.tar.gz: pyicu-2.16.2/bases.cpp: if (sizeof(Py_UNICODE) == sizeof(UChar))
PYPI-2026-09-17/pyicu-2.16.2.tar.gz: pyicu-2.16.2/bases.cpp: Py_UNICODE c = (Py_UNICODE) u->charAt(n);
PYPI-2026-09-17/pyicu-2.16.2.tar.gz: pyicu-2.16.2/common.cpp: else if (sizeof(Py_UNICODE) == sizeof(UChar))
PYPI-2026-09-17/pyicu-2.16.2.tar.gz: pyicu-2.16.2/common.cpp: return PyUnicode_FromUnicode((const Py_UNICODE *) utf16, len16);
PYPI-2026-09-17/pyicu-2.16.2.tar.gz: pyicu-2.16.2/common.cpp: Py_UNICODE *pchars = PyUnicode_AS_UNICODE(u);
PYPI-2026-09-17/pyicu-2.16.2.tar.gz: pyicu-2.16.2/common.cpp: if (sizeof(Py_UNICODE) == sizeof(UChar))
PYPI-2026-09-17/pyicu-2.16.2.tar.gz: pyicu-2.16.2/common.cpp: else if (sizeof(Py_UNICODE) == sizeof(UChar32))
PYPI-2026-09-17/pyicu-2.16.2.tar.gz: pyicu-2.16.2/common.cpp: Py_UNICODE *pchars = PyUnicode_AS_UNICODE(object);
PYPI-2026-09-17/PyICU-binary-2.7.4.tar.gz: PyICU-binary-2.7.4/bases.cpp: if (sizeof(Py_UNICODE) == sizeof(UChar))
PYPI-2026-09-17/PyICU-binary-2.7.4.tar.gz: PyICU-binary-2.7.4/bases.cpp: if (sizeof(Py_UNICODE) == sizeof(UChar))
PYPI-2026-09-17/PyICU-binary-2.7.4.tar.gz: PyICU-binary-2.7.4/bases.cpp: if (sizeof(Py_UNICODE) == sizeof(UChar))
PYPI-2026-09-17/PyICU-binary-2.7.4.tar.gz: PyICU-binary-2.7.4/bases.cpp: if (sizeof(Py_UNICODE) == sizeof(UChar))
PYPI-2026-09-17/PyICU-binary-2.7.4.tar.gz: PyICU-binary-2.7.4/bases.cpp: if (sizeof(Py_UNICODE) == sizeof(UChar))
PYPI-2026-09-17/PyICU-binary-2.7.4.tar.gz: PyICU-binary-2.7.4/bases.cpp: if (sizeof(Py_UNICODE) == sizeof(UChar))
PYPI-2026-09-17/PyICU-binary-2.7.4.tar.gz: PyICU-binary-2.7.4/bases.cpp: if (sizeof(Py_UNICODE) == sizeof(UChar))
PYPI-2026-09-17/PyICU-binary-2.7.4.tar.gz: PyICU-binary-2.7.4/bases.cpp: Py_UNICODE c = (Py_UNICODE) u->charAt(n);
PYPI-2026-09-17/PyICU-binary-2.7.4.tar.gz: PyICU-binary-2.7.4/common.cpp: else if (sizeof(Py_UNICODE) == sizeof(UChar))
PYPI-2026-09-17/PyICU-binary-2.7.4.tar.gz: PyICU-binary-2.7.4/common.cpp: return PyUnicode_FromUnicode((const Py_UNICODE *) utf16, len16);
PYPI-2026-09-17/PyICU-binary-2.7.4.tar.gz: PyICU-binary-2.7.4/common.cpp: Py_UNICODE *pchars = PyUnicode_AS_UNICODE(u);
PYPI-2026-09-17/PyICU-binary-2.7.4.tar.gz: PyICU-binary-2.7.4/common.cpp: if (sizeof(Py_UNICODE) == sizeof(UChar))
PYPI-2026-09-17/PyICU-binary-2.7.4.tar.gz: PyICU-binary-2.7.4/common.cpp: else if (sizeof(Py_UNICODE) == sizeof(UChar32))
PYPI-2026-09-17/PyICU-binary-2.7.4.tar.gz: PyICU-binary-2.7.4/common.cpp: Py_UNICODE *pchars = PyUnicode_AS_UNICODE(object);
PYPI-2026-09-17/pyobjc_core-12.2.2.tar.gz: pyobjc_core-12.2.2/Modules/objc/OC_PythonUnicode.m: * This can be slightly optimized on systems where sizeof(Py_UNICODE) ==
PYPI-2026-09-17/pyobjc_core-12.2.2.tar.gz: pyobjc_core-12.2.2/Modules/objc/OC_PythonUnicode.m: /* NOTE: the malloc() call can be avoided when sizeof(unichar) == sizeof(Py_UNICODE)
PYPI-2026-09-17/pypyodbc-1.3.6.tar.gz: pypyodbc-1.3.6/pypyodbc.py: #determin the size of Py_UNICODE
PYPI-2026-09-17/PyQt5-5.15.11.tar.gz: PyQt5-5.15.11/designer/pluginloader.cpp: ucs4.size() * sizeof (Py_UNICODE));
PYPI-2026-09-17/PyQt5-5.15.11.tar.gz: PyQt5-5.15.11/designer/pluginloader.cpp: native_dir.length() * sizeof (Py_UNICODE));
PYPI-2026-09-17/PyQt5-5.15.11.tar.gz: PyQt5-5.15.11/qmlscene/pluginloader.cpp: ucs4.size() * sizeof (Py_UNICODE));
PYPI-2026-09-17/PyQt5-5.15.11.tar.gz: PyQt5-5.15.11/qmlscene/pluginloader.cpp: py_plugin_dir.length() * sizeof (Py_UNICODE));
PYPI-2026-09-17/PyQt5-5.15.11.tar.gz: PyQt5-5.15.11/qpy/QtCore/qpycore_qstring.cpp: ucs4.size() * sizeof (Py_UNICODE));
PYPI-2026-09-17/PyQt5-5.15.11.tar.gz: PyQt5-5.15.11/qpy/QtCore/qpycore_qstring.cpp: qstr.length() * sizeof (Py_UNICODE));
PYPI-2026-09-17/pyScss-1.4.0.tar.gz: pyScss-1.4.0/scss/src/block_locator.c: int _strip(Py_UNICODE *begin, Py_UNICODE *end, int *lineno, Py_UNICODE **out) {
PYPI-2026-09-17/pyScss-1.4.0.tar.gz: pyScss-1.4.0/scss/src/block_locator.c: Py_UNICODE *block_start = NULL;
PYPI-2026-09-17/pyScss-1.4.0.tar.gz: pyScss-1.4.0/scss/src/block_locator.c: Py_UNICODE *block_start = NULL;
PYPI-2026-09-17/pyScss-1.4.0.tar.gz: pyScss-1.4.0/scss/src/block_locator.c: Py_UNICODE *block_start = NULL;
PYPI-2026-09-17/pyScss-1.4.0.tar.gz: pyScss-1.4.0/scss/src/block_locator.c: Py_UNICODE *codestr_end = self->codestr + self->codestr_sz;
PYPI-2026-09-17/pyScss-1.4.0.tar.gz: pyScss-1.4.0/scss/src/block_locator.h: Py_UNICODE *selprop;
PYPI-2026-09-17/pyScss-1.4.0.tar.gz: pyScss-1.4.0/scss/src/block_locator.h: Py_UNICODE *codestr;
PYPI-2026-09-17/pyScss-1.4.0.tar.gz: pyScss-1.4.0/scss/src/block_locator.h: Py_UNICODE *codestr;
PYPI-2026-09-17/pyScss-1.4.0.tar.gz: pyScss-1.4.0/scss/src/block_locator.h: Py_UNICODE *codestr_ptr;
PYPI-2026-09-17/pyScss-1.4.0.tar.gz: pyScss-1.4.0/scss/src/block_locator.h: Py_UNICODE instr;
PYPI-2026-09-17/pyScss-1.4.0.tar.gz: pyScss-1.4.0/scss/src/block_locator.h: Py_UNICODE *init;
PYPI-2026-09-17/pyScss-1.4.0.tar.gz: pyScss-1.4.0/scss/src/block_locator.h: Py_UNICODE *lose;
PYPI-2026-09-17/pyScss-1.4.0.tar.gz: pyScss-1.4.0/scss/src/block_locator.h: Py_UNICODE *start;
PYPI-2026-09-17/pyScss-1.4.0.tar.gz: pyScss-1.4.0/scss/src/block_locator.h: Py_UNICODE *end;
PYPI-2026-09-17/pystack-1.7.1.tar.gz: pystack-1.7.1/src/pystack/_pystack/cpython/string.h: typedef Py_UCS4 Py_UNICODE;
PYPI-2026-09-17/pystack-1.7.1.tar.gz: pystack-1.7.1/src/pystack/_pystack/cpython/string.h: Py_UNICODE* str;
PYPI-2026-09-17/pywin32-ctypes-0.2.3.tar.gz: pywin32-ctypes-0.2.3/win32ctypes/core/ctypes/_common.py: PPy_UNICODE = c_void_p
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: /* Py_UNICODE was the native Unicode storage format (code unit) used by
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: With PEP 393, Py_UNICODE is deprecated and replaced with a
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: #define PY_UNICODE_TYPE wchar_t
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: /* Py_DEPRECATED(3.3) */ typedef wchar_t Py_UNICODE;
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: /* --- Plain Py_UNICODE --------------------------------------------------- */
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: wstr/Py_UNICODE representation. This function is used to convert Unicode
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: /* Create a Unicode Object from the Py_UNICODE buffer u of the given
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: const Py_UNICODE *u,        /* Unicode buffer */
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: Py_UNICODE buffer.
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: If the wchar_t/Py_UNICODE representation is not yet available, this
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: Py_DEPRECATED(3.3) PyAPI_FUNC(Py_UNICODE *) PyUnicode_AsUnicode(
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: PyAPI_FUNC(const Py_UNICODE *) _PyUnicode_AsUnicode(
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: Py_UNICODE buffer and save the length at size.
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: If the wchar_t/Py_UNICODE representation is not yet available, this
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: Py_DEPRECATED(3.3) PyAPI_FUNC(Py_UNICODE *) PyUnicode_AsUnicodeAndSize(
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: /* Returns the deprecated Py_UNICODE representation's size in code units
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: If the Py_UNICODE representation is not available, it will be computed
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: /* Alias for PyUnicode_AsUnicode().  This will create a wchar_t/Py_UNICODE
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: static inline Py_UNICODE* PyUnicode_AS_UNICODE(PyObject *op)
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/cpython/unicodeobject.h: Py_UNICODE *data = PyUnicode_AS_UNICODE(op);
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/unicodeobject.h: /* Get the number of Py_UNICODE units in the
PYPI-2026-09-17/pywinpty-3.0.5.tar.gz: pywinpty-3.0.5/pythonarm64/tools/include/unicodeobject.h: is the number of Py_UNICODE characters.
PYPI-2026-09-17/rcssmin-1.2.2.tar.gz: rcssmin-1.2.2/cext.h: #define EXT_UNI_CP Py_UNICODE
PYPI-2026-09-17/rcssmin-1.2.2.tar.gz: rcssmin-1.2.2/cext.h: #error "Cannot recognize sizeof(Py_UNICODE)"
PYPI-2026-09-17/rcssmin-1.2.2.tar.gz: rcssmin-1.2.2/cext.h: ((Py_UNICODE) (((const Py_UNICODE *)(data))[(index)]))
PYPI-2026-09-17/rcssmin-1.2.2.tar.gz: rcssmin-1.2.2/cext.h: ((Py_UNICODE *)(data))[(index)] = (Py_UNICODE)(value); \
PYPI-2026-09-17/rcssmin-1.2.2.tar.gz: rcssmin-1.2.2/cext.h: (PyUnicode_FromUnicode((Py_UNICODE *)data, length))
PYPI-2026-09-17/rjsmin-1.2.5.tar.gz: rjsmin-1.2.5/cext.h: #define EXT_UNI_CP Py_UNICODE
PYPI-2026-09-17/rjsmin-1.2.5.tar.gz: rjsmin-1.2.5/cext.h: #error "Cannot recognize sizeof(Py_UNICODE)"
PYPI-2026-09-17/rjsmin-1.2.5.tar.gz: rjsmin-1.2.5/cext.h: ((Py_UNICODE) (((const Py_UNICODE *)(data))[(index)]))
PYPI-2026-09-17/rjsmin-1.2.5.tar.gz: rjsmin-1.2.5/cext.h: ((Py_UNICODE *)(data))[(index)] = (Py_UNICODE)(value); \
PYPI-2026-09-17/rjsmin-1.2.5.tar.gz: rjsmin-1.2.5/cext.h: (PyUnicode_FromUnicode((Py_UNICODE *)data, length))
PYPI-2026-09-17/simplejson-4.1.2.tar.gz: simplejson-4.1.2/simplejson/_speedups.c: #define PyUnicode_KIND(obj) (sizeof(Py_UNICODE))
PYPI-2026-09-17/simplejson-4.1.2.tar.gz: simplejson-4.1.2/simplejson/_speedups.c: #define PyUnicode_READ(kind, data, index) ((JSON_UNICHR)((const Py_UNICODE *)(data))[(index)])
PYPI-2026-09-17/simplejson-4.1.2.tar.gz: simplejson-4.1.2/simplejson/_speedups.c: #define JSON_UNICHR Py_UNICODE
PYPI-2026-09-17/simplejson-4.1.2.tar.gz: simplejson-4.1.2/simplejson/_speedups.c: Py_UNICODE c = 0;
PYPI-2026-09-17/simplejson-4.1.2.tar.gz: simplejson-4.1.2/simplejson/_speedups.c: chunk = PyUnicode_FromUnicode(&((const Py_UNICODE *)buf)[end], next - end);
PYPI-2026-09-17/simplejson-4.1.2.tar.gz: simplejson-4.1.2/simplejson/_speedups.c: Py_UNICODE *output = PyUnicode_AS_UNICODE(rval);
PYPI-2026-09-17/simplejson-4.1.2.tar.gz: simplejson-4.1.2/simplejson/_speedups.c: PyUnicode_FromUnicode(&((Py_UNICODE *)str)[(sidx)], (eidx) - (sidx))
PYPI-2026-09-17/tiled-0.2.18.tar.gz: tiled-0.2.18/tiled/structures/array.py: unicode = "U"  # fixed-length sequence of Py_UNICODE
PYPI-2026-09-17/twisted_iocpsupport-25.10.1.tar.gz: twisted_iocpsupport-25.10.1/twisted_iocpsupport/iocpsupport.pyx: ctypedef Py_UNICODE WCHAR

encukou · 2026-09-29T14:49:43Z

Does it mean that we should keep forever?

Yes, that's the current guidance in PEP-387:

If the expected maintenance overhead and security risk of the deprecated behavior is small (e.g. an old function is reimplemented in terms of a new, more general one), it can stay indefinitely (or until the situation changes).

hugovk · 2026-09-29T15:14:31Z

An LLM check on the 29 projects:

Summary: one actively maintained project breaks (lxml), and the root cause is Cython. Everything else is either dead code behind Python 2 / pre-3.3 guards, already broken on 3.12+, or not C code at all.

Details

I went through Victor's list of 29 projects by downloading each sdist and checking the preprocessor guards around every Py_UNICODE match, and by compiling lxml's shipped etree.c against headers with the typedef removed.

Would break on 3.16 with this PR (compiles fine on 3.13 today)

lxml 6.1.3 (and lxml master). _setupPythonUnicode() in src/lxml/parser.pxi declares cdef Py_UNICODE *uchars. The call is guarded by PY_VERSION_HEX < 0x030C0000 at runtime, but Cython compiles the function unconditionally. The shipped etree.c compiles cleanly against current main (with two Py_UNICODE is deprecated warnings) and fails with 14 errors against headers without the typedef, all from etree.c:129077 and :129080. lxml master still has this code; its only added guard is #define Py_UNICODE wchar_t under Py_LIMITED_API, so normal builds are still affected.

Cython 3.3.0 itself. Any cdef Py_UNICODE in user code is emitted verbatim as Py_UNICODE in the generated C: CPyUnicodeIntType.sign_and_name() returns "Py_UNICODE" in 3.3.0 and on master. So this removal breaks every Cython project using the type until Cython emits wchar_t instead. lxml is one such project.

PyICU-binary 2.7.4. Seven unguarded sizeof(Py_UNICODE) == sizeof(UChar) comparisons in bases.cpp. Abandoned fork, last released September 2021. The real PyICU 2.16.2 guards every use with PY_VERSION_HEX < 0x030d0000 and is fine.

Already broken on 3.12+ regardless (they use PyUnicode_AS_UNICODE, removed in 3.12)

pyScss 1.4.0 (2022): unconditional Py_UNICODE in block_locator.c. The C speedups are optional and fall back to pure Python.

Distance 0.1.3 (2013): unconditional typedef. The C extension only builds with an explicit --with-c flag.

Not affected

Guarded to Python 2 or pre-3.3: genshi, netifaces, netifaces_plus, simplejson, rcssmin, rjsmin, cffi (wchar_helper.h is only included when PyUnicode_KIND is undefined), pyahocorasick, pyicu 2.16.2, PyQt5, nuitka (guards at 3.0 and 3.12).

Cython declarations that never reach C: twisted_iocpsupport (the ctypedef Py_UNICODE WCHAR is inside an extern from "windows.h" block; the cythonized output contains no Py_UNICODE), line_profiler (python25.pxd is not cimported by anything).

Own definitions or non-C: pystack (its own typedef in a pystack namespace, no Python.h), orjson (Rust pyo3-ffi alias), pyobjc_core (comments), cu2qu, igwn_ligolw, minorminer (vendored virtualenv), pygments (test fixture), pypyodbc, pywin32-ctypes, pywinpty (vendored CPython header), tiled.

PY_UNICODE_TYPE matched nothing outside pywinpty's vendored copy of unicodeobject.h, so removing it as well has no additional cost on this list.

So the practical blocker is Cython: a Cython release that emits wchar_t for Py_UNICODE, plus a regenerated lxml sdist, would resolve the only real breakage.

vstinner · 2026-10-24T01:29:53Z

@fedonman: Can you update the PR to remove also PY_UNICODE_TYPE?

fedonman · 2026-10-26T09:42:21Z

Done, PY_UNICODE_TYPE is removed too.

fedonman requested review from AA-Turner, StanFromIreland, ZeroIntensity and hugovk as code owners July 27, 2026 08:24

bedevere-app Bot added the awaiting review label Jul 27, 2026

bedevere-app Bot mentioned this pull request Jul 27, 2026

26 C APIs scheduled for removal in 3.16 are still present #154757

Open

3 tasks

Merge branch 'main' into remove-py-unicode-typedef

eda7732

hugovk reviewed Aug 20, 2026

View reviewed changes

Comment thread Doc/whatsnew/3.3.rst Outdated

Address review: Py_UNICODE was removed in 3.16, not Python 4

bd1d78d

fedonman requested a review from hugovk August 28, 2026 21:57

vstinner reviewed Sep 23, 2026

View reviewed changes

Merge branch 'main' into remove-py-unicode-typedef

9e4a944

vstinner mentioned this pull request Oct 24, 2026

gh-154757: Unschedule removal of Py_UNICODE (and Py_UNICODE_TYPE) #160538

Open

Address review: remove PY_UNICODE_TYPE as well

0451d6e

bedevere-app Bot added the type-feature A feature request or enhancement label Oct 26, 2026

Conversation

fedonman commented Jul 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

read-the-docs-community Bot commented Jul 27, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Documentation build overview

Uh oh!

Uh oh!

vstinner left a comment

Choose a reason for hiding this comment

Uh oh!

encukou commented Sep 24, 2026

Uh oh!

vstinner commented Sep 24, 2026

Uh oh!

fedonman commented Sep 26, 2026

Uh oh!

encukou commented Sep 29, 2026

Uh oh!

vstinner commented Sep 29, 2026

Uh oh!

encukou commented Sep 29, 2026

Uh oh!

hugovk commented Sep 29, 2026

Would break on 3.16 with this PR (compiles fine on 3.13 today)

Already broken on 3.12+ regardless (they use PyUnicode_AS_UNICODE, removed in 3.12)

Not affected

Uh oh!

vstinner commented Oct 24, 2026

Uh oh!

fedonman commented Oct 26, 2026

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

fedonman commented Jul 27, 2026 •

edited

Loading

read-the-docs-community Bot commented Jul 27, 2026 •

edited

Loading

Already broken on 3.12+ regardless (they use `PyUnicode_AS_UNICODE`, removed in 3.12)