PyType_Spec basicsize/itemsize should allow -1 for "inherit"

[encukou]

BPO	45525
Nosy	@encukou

^{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 2021-10-20.10:40:01.685>
created_at = <Date 2021-10-19.15:43:58.181>
labels = []
title = 'PyType_Spec basicsize/itemsize should allow -1 for "inherit"'
updated_at = <Date 2021-10-20.10:40:01.683>
user = 'https://github.com/encukou'

bugs.python.org fields:

activity = <Date 2021-10-20.10:40:01.683>
actor = 'petr.viktorin'
assignee = 'none'
closed = True
closed_date = <Date 2021-10-20.10:40:01.685>
closer = 'petr.viktorin'
components = []
creation = <Date 2021-10-19.15:43:58.181>
creator = 'petr.viktorin'
dependencies = []
files = []
hgrepos = []
issue_num = 45525
keywords = []
message_count = 2.0
messages = ['404311', '404421']
nosy_count = 1.0
nosy_names = ['petr.viktorin']
pr_nums = []
priority = 'normal'
resolution = 'postponed'
stage = 'resolved'
status = 'closed'
superseder = None
type = None
url = 'https://bugs.python.org/issue45525'
versions = []

[encukou]

For cases where you aren't adding new C-level state, especially if don't have the superclass' struct available, it would be good to allow inheriting basicsize/itemsize from the superclass.

(SuperType->tp_size is not easily usable for the same reason PyType_FromSpecWithBases exists: on MSVC, if SuperType comes from a different DLL, it is not a constant expression.)

[encukou]

Hm, after sleeping on it, I think I filed this too soon.
If you don't have the superclass's C memory layout (and you can't add new C-level state), it's not likely that you want to define a class in C.

I can construct theoretical cases where it might be useful, but IMO it should wait for an *actual* motivating use case. (The one I had was making a specific C-API test a bit easier to write, which is very weak.)

No need to have this sit on CPython's TODO list. I'll close this and let someone reopen (or file a new issue) if they actually need it.

[encukou]

FWIW, there is a use case now, along with a more general solution proposed: https://mail.python.org/archives/list/capi-sig@python.org/