Remove reference to undefined dictionary ordering in Tutorial

[DimitrisJim]

BPO	29453
Nosy	@rhettinger, @bitdancer, @methane, @serhiy-storchaka, @zhangyangyu, @DimitrisJim, @Mariatta
PRs	bpo-29453: Remove reference to undefined dictionary ordering in Tutorial #140 [3.6] bpo-29453: Remove reference to undefined dictionary ordering in Tutor… #208
Files	controlflowdiff.patch controlflowdiff2.patch

^{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/Mariatta'
closed_at = <Date 2017-02-21.18:32:35.967>
created_at = <Date 2017-02-05.15:17:27.092>
labels = ['3.7', 'docs']
title = 'Remove reference to undefined dictionary ordering in Tutorial'
updated_at = <Date 2017-02-21.18:32:35.966>
user = 'https://github.com/DimitrisJim'

bugs.python.org fields:

activity = <Date 2017-02-21.18:32:35.966>
actor = 'Mariatta'
assignee = 'Mariatta'
closed = True
closed_date = <Date 2017-02-21.18:32:35.967>
closer = 'Mariatta'
components = ['Documentation']
creation = <Date 2017-02-05.15:17:27.092>
creator = 'Jim Fasarakis-Hilliard'
dependencies = []
files = ['46524', '46525']
hgrepos = []
issue_num = 29453
keywords = ['patch']
message_count = 20.0
messages = ['287048', '287050', '287051', '287053', '287054', '287056', '287057', '287058', '287059', '287060', '287061', '287092', '287138', '287219', '288047', '288081', '288098', '288272', '288313', '288314']
nosy_count = 8.0
nosy_names = ['rhettinger', 'r.david.murray', 'methane', 'docs@python', 'serhiy.storchaka', 'xiang.zhang', 'Jim Fasarakis-Hilliard', 'Mariatta']
pr_nums = ['140', '208']
priority = 'normal'
resolution = 'fixed'
stage = 'resolved'
status = 'closed'
superseder = None
type = None
url = 'https://bugs.python.org/issue29453'
versions = ['Python 3.6', 'Python 3.7']

[DimitrisJim]

Removes keys = sorted(keywords.keys()) from function example and removes the text that describes why this was necessary. As per PEP-468, this note is obsolete for 3.6+

Also changes the ordering of the function call to match the previous output.

[methane]

LGTM

[bitdancer]

It is not (yet) a language requirement that ordinary dictionaries be ordered. This patch may become appropriate in 3.7, but that has not yet been determined. It is not appropriate for 3.6. In 3.6, the order of keys in an ordinary dictionary is still undefined, even though it is in practice consistent in CPython.

[DimitrisJim]

Isn't it a language requirement that **kwargs be ordered in 3.6, David?

PEP-468 states that **kwargs is to be an ordered mapping and, if I'm not mistaken, that was done in order to not depend on the fact that dicts became ordered. I might have understood something wrong, though :-)

[zhangyangyu]

David, actually I have the same thoughts as Jim. Ordered ordinary dicts is not a feature but ordered **kwargs is in 3.6. They seems not the same thing.

[serhiy-storchaka]

I would not change the order of keyword arguments, but rather change the output.

[DimitrisJim]

It was a random decision on my part, Serhiy, since I didn't see any difference. Why would you go the other way around?

[serhiy-storchaka]

Because it shows preserving the order of keyword arguments (rather than sorting by keyword name).

[DimitrisJim]

Indeed, good point. Changed it to the suggested way.

[serhiy-storchaka]

LGTM. Thanks Jim.

But maybe it is worth to mention that the output corresponds to the order of passed keyword arguments.

[bitdancer]

You are correct, I didn't read the full context of the diff. My apologies.

[rhettinger]

Patch 2 looks fine to me.

[DimitrisJim]

But maybe it is worth to mention that the output corresponds to the order of passed keyword arguments

Should I add this note? It looks fine to me as is but I'm not the experienced one here :-)

[serhiy-storchaka]

I don't know. David, Raymond, what are your thoughts?

[rhettinger]

> But maybe it is worth to mention that the output
> corresponds to the order of passed keyword arguments

Should I add this note?

Yes, please. This is a section on keyword arguments, making it the preferred place to mention the new guaranteed output order.

[DimitrisJim]

Added the following short sentence to the PR, which I believe makes the point clear:

Note that the order in which the keyword arguments are printed is guaranteed to match the order in which they were provided in the function call.

[rhettinger]

Thanks for the revision. Assigning to Mariatta so that she can apply the PR.

[Mariatta]

New changeset 32e8f9b by Mariatta in branch 'master':
bpo-29453: Remove reference to undefined dictionary ordering in Tutorial (GH-140)
32e8f9b

[Mariatta]

New changeset 9b49133 by GitHub in branch '3.6':
bpo-29453: Remove reference to undefined dictionary ordering in Tutorial (GH-140) (#208)
9b49133

[Mariatta]

Thanks everyone. PR has been merged and backported to 3.6 :)
Closing this issue.