Remove namespace copy from dataclasses.make_dataclass()

[ericvsmith]

BPO	43820
Nosy	@ericvsmith
PRs	bpo-43820: Remove an unnecessary copy of the 'namespace' parameter to make_dataclass() #25372

^{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 = 'https://github.com/ericvsmith'
closed_at = <Date 2021-04-14.23:31:18.151>
created_at = <Date 2021-04-12.21:12:10.351>
labels = ['type-bug', 'library', '3.10']
title = 'Remove namespace copy from dataclasses.make_dataclass()'
updated_at = <Date 2021-04-14.23:31:18.151>
user = 'https://github.com/ericvsmith'

bugs.python.org fields:

activity = <Date 2021-04-14.23:31:18.151>
actor = 'eric.smith'
assignee = 'eric.smith'
closed = True
closed_date = <Date 2021-04-14.23:31:18.151>
closer = 'eric.smith'
components = ['Library (Lib)']
creation = <Date 2021-04-12.21:12:10.351>
creator = 'eric.smith'
dependencies = []
files = []
hgrepos = []
issue_num = 43820
keywords = ['patch']
message_count = 1.0
messages = ['390907']
nosy_count = 1.0
nosy_names = ['eric.smith']
pr_nums = ['25372']
priority = 'normal'
resolution = 'fixed'
stage = 'resolved'
status = 'closed'
superseder = None
type = 'behavior'
url = 'https://bugs.python.org/issue43820'
versions = ['Python 3.10']

[ericvsmith]

In make_dataclasses(), the "namespace" argument is copied because it is mutated. This isn't really necessary: the callback passed to types.new_class() could just update its "ns" parameter with "namespace" and with the new attributes (anotations and default field values).

This would eliminate the need to copy "namespace".