Remove Py_UNICODE APIs except PEP 623

[methane]

BPO	41123
Nosy	@vstinner, @methane, @miss-islington
PRs	bpo-41123: Remove Py_UNICODE_str* macros #21164 bpo-41123: Doc: Document Py_UNICODE APIs will be removed in 3.10 #21181 bpo-41123: Remove PyUnicode_GetMax() #21192 bpo-41123: Remove PyLong_FromUnicode() #21204 [3.9] bpo-41123: Doc: PyLong_FromUnicode will be removed in 3.10 #21205 [3.8] bpo-41123: Doc: PyLong_FromUnicode will be removed in 3.10 (GH-21205) #21207 bpo-41123: Remove PyUnicode_AsUnicodeCopy #21209 [3.9] bpo-41123: Remove PyUnicode_AsUnicodeCopy in 3.10 #21227 [3.8] bpo-41123: Remove PyUnicode_AsUnicodeCopy in 3.10 (GH-21227) #21230 bpo-41123: Remove Py_UNICODE_COPY() and Py_UNICODE_FILL() #28887

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields:

assignee = None
closed_at = <Date 2021-03-16.04:23:19.290>
created_at = <Date 2020-06-26.03:53:21.785>
labels = ['expert-C-API', '3.10']
title = 'Remove Py_UNICODE APIs except PEP 623'
updated_at = <Date 2021-10-11.21:36:45.040>
user = 'https://github.com/methane'

bugs.python.org fields:

activity = <Date 2021-10-11.21:36:45.040>
actor = 'vstinner'
assignee = 'none'
closed = True
closed_date = <Date 2021-03-16.04:23:19.290>
closer = 'methane'
components = ['C API']
creation = <Date 2020-06-26.03:53:21.785>
creator = 'methane'
dependencies = []
files = []
hgrepos = []
issue_num = 41123
keywords = ['patch']
message_count = 14.0
messages = ['372407', '372460', '372468', '372489', '372537', '372538', '372539', '372544', '372545', '372548', '372646', '372657', '372666', '403690']
nosy_count = 3.0
nosy_names = ['vstinner', 'methane', 'miss-islington']
pr_nums = ['21164', '21181', '21192', '21204', '21205', '21207', '21209', '21227', '21230', '28887']
priority = 'normal'
resolution = 'fixed'
stage = 'resolved'
status = 'closed'
superseder = None
type = None
url = 'https://bugs.python.org/issue41123'
versions = ['Python 3.10']

[methane]

# APIs relating to wstr

Since some APIs did not have Py_DEPRECATE until 3.9 (see #65140), it can not be removed in 3.10. I wrote PEP-623 for them.

This issue doesn't about them.

# Deprecated since Python 3.3, and not documented.

In Python 3.3 what's new:

• :c:macro:`Py_UNICODE_strlen`: use :c:func:`PyUnicode_GetLength` or
  :c:macro:`PyUnicode_GET_LENGTH`
• :c:macro:`Py_UNICODE_strcat`: use :c:func:`PyUnicode_CopyCharacters` or
  :c:func:`PyUnicode_FromFormat`
• :c:macro:`Py_UNICODE_strcpy`, :c:macro:`Py_UNICODE_strncpy`,
  :c:macro:`Py_UNICODE_COPY`: use :c:func:`PyUnicode_CopyCharacters` or
  :c:func:`PyUnicode_Substring`
• :c:macro:`Py_UNICODE_strcmp`: use :c:func:`PyUnicode_Compare`
• :c:macro:`Py_UNICODE_strncmp`: use :c:func:`PyUnicode_Tailmatch`
• :c:macro:`Py_UNICODE_strchr`, :c:macro:`Py_UNICODE_strrchr`: use
  :c:func:`PyUnicode_FindChar`

These functions are not documented. But they has Py_DEPRECATED(3.3) from Python 3.6.
Let's remove them in 3.10.

# Deprecated since Python 3.3 with document

Some APIs has document with .. deprecated:: 3.3 4.0.

• PyLong_FromUnicode
• PyUnicode_TransformDecimalToASCII
• PyUnicode_AsUnicodeCopy
• PyUnicode_Encode
• PyUnicode_EncodeUTF7
• PyUnicode_EncodeUTF8
• PyUnicode_EncodeUTF16
• PyUnicode_EncodeUTF32
• PyUnicode_EncodeUnicodeEscape
• PyUnicode_EncodeRawUnicodeEscape
• PyUnicode_EncodeLatin1
• PyUnicode_EncodeASCII
• PyUnicode_EncodeCharmap
• PyUnicode_TranslateCharmap
• PyUnicode_EncodeMBCS

a) Can we replace 4.0 with 3.10 and remove them in 3.10?
b) Or should we replace 4.0 with 3.11 and wait one more year?

[methane]

New changeset 20a7902 by Inada Naoki in branch 'master':
bpo-41123: Remove Py_UNICODE_str* functions (GH-21164)
20a7902

[methane]

a) Can we replace 4.0 with 3.10 and remove them in 3.10?
b) Or should we replace 4.0 with 3.11 and wait one more year?

I chose this because they are deprecated since 3.3, and Py_DEPRECATED is used since 3.6.

[methane]

APIs have /* Py_DEPRECATED(3.3) */

• PyUnicode_EncodeDecimal
  It is used in PyLong_FromUnicode. Deprecate in 3.10 and remove in 3.12

• PyUnicode_TransformDecimalToASCII
  While PEP-393 deprecate this too, its document doesn't has deprecated. Deprecate in 3.10 and remove in 3.12

• _PyUnicode_ToLowercase, _PyUnicode_ToUppercase
  They are not deprecated by PEP-393, but bpo-12736. Deprecate in 3.10 and remove in 3.12.
  On the other hand, _PyUnicode_ToTitlecase has Py_DEPRECATED. It can be removed in 3.10.

[methane]

I confirmed PyUnicode_GetMax is not used in top4000 packages.

[methane]

New changeset d9f2a13 by Inada Naoki in branch 'master':
bpo-41123: Remove PyUnicode_GetMax() (GH-21192)
d9f2a13

[methane]

New changeset e4f1fe6 by Inada Naoki in branch 'master':
bpo-41123: Remove PyLong_FromUnicode() (GH-21204)
e4f1fe6

[methane]

New changeset 02134da by Inada Naoki in branch '3.9':
bpo-41123: Doc: PyLong_FromUnicode will be removed in 3.10 (GH-21205)
02134da

[miss-islington]

New changeset ea16430 by Miss Islington (bot) in branch '3.8':
bpo-41123: Doc: PyLong_FromUnicode will be removed in 3.10 (GH-21205)
ea16430

[methane]

PyUnicode_AsUnicodeCopy is not used in top 4000 packages. Let's remove it too.

[methane]

New changeset b333266 by Inada Naoki in branch 'master':
bpo-41123: Remove PyUnicode_AsUnicodeCopy (GH-21209)
b333266

[methane]

New changeset 2ea6a99 by Inada Naoki in branch '3.9':
bpo-41123: Remove PyUnicode_AsUnicodeCopy in 3.10 (GH-21227)
2ea6a99

[miss-islington]

New changeset 41d6e3f by Miss Islington (bot) in branch '3.8':
bpo-41123: Remove PyUnicode_AsUnicodeCopy in 3.10 (GH-21227)
41d6e3f

[vstinner]

New changeset 1f316ea by Victor Stinner in branch 'main':
bpo-41123: Remove Py_UNICODE_COPY() and Py_UNICODE_FILL() (GH-28887)
1f316ea