urllib.parse should not discard delimiters when associated component is empty

[gdatagmail]

BPO	24332
Nosy	@orsenthil, @vadmium
Superseder	bpo-22852: urllib.parse wrongly strips empty #fragment, ?query, //netloc

^{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 2015-05-31.04:34:30.570>
created_at = <Date 2015-05-30.18:05:17.114>
labels = []
title = 'urllib.parse should not discard delimiters when associated component is empty'
updated_at = <Date 2015-05-31.04:34:30.568>
user = 'https://bugs.python.org/gdatagmail'

bugs.python.org fields:

activity = <Date 2015-05-31.04:34:30.568>
actor = 'martin.panter'
assignee = 'none'
closed = True
closed_date = <Date 2015-05-31.04:34:30.570>
closer = 'martin.panter'
components = []
creation = <Date 2015-05-30.18:05:17.114>
creator = 'gdata gmail'
dependencies = []
files = []
hgrepos = []
issue_num = 24332
keywords = []
message_count = 2.0
messages = ['244477', '244515']
nosy_count = 3.0
nosy_names = ['orsenthil', 'martin.panter', 'gdata gmail']
pr_nums = []
priority = 'normal'
resolution = 'duplicate'
stage = 'resolved'
status = 'closed'
superseder = '22852'
type = None
url = 'https://bugs.python.org/issue24332'
versions = []

[gdatagmail]

The documenatation for urllib.parse (https://docs.python.org/3.0/library/urllib.parse.html) states several times:

"This may result in a slightly different, but equivalent URL, if the URL that was parsed originally had unnecessary delimiters (for example, a ? with an empty query; the RFC states that these are equivalent)."

This is false -- RFC 3986 explicitly states that ? with an empty query is _not_ equivalent to a URL without it. For example, the following two URL's should be considered different:

http://example.com/?
http://example.com/

https://tools.ietf.org/html/rfc3986#section-6.2.3

[vadmium]

This is essentially the same as bpo-22852. The title just refers to stripping an empty #fragment, but the netloc and query components are also affected. I have a patch there which needs reviewing, if you are interested. Or if you have any alternative ideas on how to solve this they would be welcome too.