the iterable for itertools.groupby must be sorted

[matrixise]

BPO	32709
Nosy	@matrixise, @csabella
PRs	bpo-32709: the iterable for itertools.groupby must be sorted #5424

^{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 2018-01-29.12:54:45.046>
created_at = <Date 2018-01-29.12:40:46.279>
labels = ['3.7', 'docs']
title = 'the iterable for itertools.groupby must be sorted'
updated_at = <Date 2018-01-29.12:54:45.046>
user = 'https://github.com/matrixise'

bugs.python.org fields:

activity = <Date 2018-01-29.12:54:45.046>
actor = 'matrixise'
assignee = 'docs@python'
closed = True
closed_date = <Date 2018-01-29.12:54:45.046>
closer = 'matrixise'
components = ['Documentation']
creation = <Date 2018-01-29.12:40:46.279>
creator = 'matrixise'
dependencies = []
files = []
hgrepos = []
issue_num = 32709
keywords = ['patch']
message_count = 4.0
messages = ['311115', '311119', '311120', '311121']
nosy_count = 3.0
nosy_names = ['docs@python', 'matrixise', 'cheryl.sabella']
pr_nums = ['5424']
priority = 'normal'
resolution = None
stage = 'resolved'
status = 'closed'
superseder = None
type = None
url = 'https://bugs.python.org/issue32709'
versions = ['Python 3.6', 'Python 3.7']

[matrixise]

The documentation of itertools.groupby indicates that generally, the iterable needs to be sorted. In fact, this iterable MUST be sorted.

[csabella]

It depends upon your use case.

In general, if you want it to work like SQL group by, then yes, it needs to be sorted, and I believe the documentation already covers that. However, there are scenarios where you want to condense a string, but keep the ordering, such as:

'aaaaabbbbbbccccaaaaddd' into '5a6b4c4a3d'

and itertools.groupby works very nicely for that.

[matrixise]

good catch, I only use itertools.groupby when I work for a collection.

[matrixise]

I am going to close my issue because with your use case, this issue makes no sense.