Implement PEP 553 - built-in breakpoint()

[warsaw]

BPO	31353
Nosy	@warsaw, @bitdancer, @serhiy-storchaka, @ethanhs, @alkuzad
PRs	PEP 553 built-in debug() function (bpo-31353) #3355

^{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/warsaw'
closed_at = <Date 2017-10-05.16:11:54.413>
created_at = <Date 2017-09-05.20:44:20.033>
labels = ['interpreter-core', 'type-feature', '3.7']
title = 'Implement PEP 553 - built-in breakpoint()'
updated_at = <Date 2018-07-09.17:47:30.913>
user = 'https://github.com/warsaw'

bugs.python.org fields:

activity = <Date 2018-07-09.17:47:30.913>
actor = 'barry'
assignee = 'barry'
closed = True
closed_date = <Date 2017-10-05.16:11:54.413>
closer = 'barry'
components = ['Interpreter Core']
creation = <Date 2017-09-05.20:44:20.033>
creator = 'barry'
dependencies = []
files = []
hgrepos = []
issue_num = 31353
keywords = []
message_count = 6.0
messages = ['301372', '302109', '302110', '303573', '321306', '321337']
nosy_count = 5.0
nosy_names = ['barry', 'r.david.murray', 'serhiy.storchaka', 'ethan smith', 'alkuzad']
pr_nums = ['3355']
priority = 'normal'
resolution = 'fixed'
stage = 'resolved'
status = 'closed'
superseder = None
type = 'enhancement'
url = 'https://bugs.python.org/issue31353'
versions = ['Python 3.7']

[warsaw]

Placeholder issue for discussion of the design of the implementation of PEP-553.

[alkuzad]

Will that be backported also to Python2.7 ? I personally do not see a reason to teach old dog new tricks but it seems appropriate in such core things. New users could transition easier, otherwise this would be next py2<->py3 difference to explain every time someone uses new code.

[bitdancer]

It is a new feature, so no it would not be backported.

[warsaw]

PEP is accepted. PR needs review.

[serhiy-storchaka]

Is it needed to specify the fromlist argument for PyImport_ImportModuleLevelObject()? The only effect of this is that Python will try to import the hook as a submodule if it is not an attribute of the parent module. But I think that callable modules is very uncommon case and we should not support is specially. Getting rid of the fromlist argument will simplify the code.

[warsaw]

It's a convenient API. I think originally I may have just don't effectively a getattr on the imported module, but I don't remember (and don't have original implementation handy - thanks, rebase!)

I don't have particularly strong feelings on the matter. If you want to submit a PR, I'll take a look. I suppose technically it *is* a functional change that might not be acceptable now that 3.7.0 is released though.