Would be nice if there were more docs for the typing module (PEP 484). Looking for volunteers. (There's stuff in the PEP that can serve as a starting point.)

Note: support for isinstance() and issubclass() will be withdrawn in beta 2.

A first iteration of the documentation based on the PEP and the module source.

This still needs a lot of work.

Addressing comments from the review. Here is an updated patch.

This should probably be committed before RC1 -- it's better than nothing!

Is anyone working on this right now? If not, I could implement the comments by Guido to the latest patch by Daniel.

I dropped the ball here. The last message from the review system was misclassified by Gmail as spam... Sorry about that.

I can take a look at this tonight but if anyone can address the comments earlier, go for it.

New version of the patch by Daniel, latest comments by Guido are taken into account.

Thank you Ivan! I am going to commit this, it's much better than the two lines we had before. :-) There's still a lot of work left to do -- towards the end there are many classes without description. Maybe you and Daniel can divide up the work for those? We can always iterate (even during the release candidates and even after the 3.5.0 release).

I could add the docs for the first half of the remaining classes by the end of this week.

New changeset a8dcbd2711d6 by Guido van Rossum in branch '3.5':
Issue #24272: Initial docs for typing.py (PEP 484).
https://hg.python.org/cpython/rev/a8dcbd2711d6

New changeset 0c74fd4219aa by Guido van Rossum in branch 'default':
Issue #24272: Initial docs for typing.py (PEP 484). (Merge from 3.5.)
https://hg.python.org/cpython/rev/0c74fd4219aa

I've landed this now (for real). Next time please strip all trailing whitespace from typing.rst, else the hg push fails!

@Ivan, I'll work on the remaining half.

Here is a new patch for the first part of classes at the end of docs.
I also did some minor changes (in particular added more info on generics and removed the section about ellipsis for default arguments, we could return it if we decide to write something about stub files here).

I assume what we've got is good enough for rc1?

Can we mark this closed / reduce the priority from "release blocker"?

Note that there are few (if any) restrictions on fixing docs during the release candidates.  (Though after today we'll be switching to the Bitbucket "pull request" workflow, so that will complicate matters slightly.)

Yeah, there's another diff but I didn't get to it.

Here's a patch for the second half of remaining classes (this file is a complement of Ivan Levkivskyi's patch).

Missing Generator.

New changeset 192c654a7c93 by Zachary Ware in branch '3.5':
Issue #24272: Remove usage of default reST role from typing docs
https://hg.python.org/cpython/rev/192c654a7c93

New changeset 10a63ded324c by Zachary Ware in branch 'default':
Issue #24272: Merge with 3.5
https://hg.python.org/cpython/rev/10a63ded324c

Guido, are you going to commit our patches before 3.5.final? Is there a merge conflict now? Should I rebase my diff?

If necessary I could make a clean patch combining two patches (my and Daniel's) on top of what Zachary did.

I apologize, I've been distracted. I will try to get to this before 3.5.0 final goes out! I don't know if the fixes will end up in the tar ball but they will certainly be on docs.python.org.

Doc fixes are always welcome.  Barring mishaps I should tag 3.5.0 final on Saturday; it'd be sweet if you could get me your diff no later than Friday night.

Pull Request: https://bitbucket.org/larry/cpython350/pull-requests/24/docs-update-for-typing-module/diff

Thank you, Guido!

I have noticed single quotes in line 38: 

type hinted using `Callable[[Arg1Type, Arg2Type], ReturnType]`

I think it would be logical to replace it with double quotes: 

``Callable[[Arg1Type, Arg2Type], ReturnType]`` 

as Zachary did in the rest of the text.

By the way, you have marked my patch as Daniels and vice versa in the commit descriptions, but that probably does not matter :)

I apologize for the mix-up! The code review tool didn't help.

Regarding the single quotes, there seem to be a bunch more of these, and as cutting a PR is a fair bit of labor I'd rather not fix that. We can fix this once 3.5.0 it out of the door (hopefully this weekend).

Pull request accepted!  Please forward-merge, thanks!

Merged and pushed. Let's please use a new issue for any further patches.

By the way, for some reason the changes in revision e361130782d4 are not in Doc/library/typing.rst.

For example,

io documentation on current default branch: https://hg.python.org/cpython/file/default/Doc/library/typing.rst#l376
io documentation on current 3.5 branch: https://hg.python.org/cpython/file/3.5/Doc/library/typing.rst#l376
Updates to the io documntation in revision e361130782d4: https://hg.python.org/cpython/rev/e361130782d4#l1.67

Also, my clone is up-to-date:

changeset:   97855:d433d74e0377
tag:         tip
user:        Victor Stinner <victor.stinner@gmail.com>
date:        Thu Sep 10 16:00:06 2015 +0200
files:       Python/pytime.c
description:
pytime: oops, fix typos on Windows

I suck at merge. :-( No idea what happened. To fix it, I have now just
copied the file again from the larry branch and committed that in 3.5, then
merged up to 3.6 (default).

changeset:   97856:8a7814a3613b
branch:      3.5
parent:      97847:6c8e2c6e3a99
user:        Guido van Rossum <guido@python.org>
date:        Thu Sep 10 10:52:11 2015 -0700
summary:     Restore doc updates to typing.rst by Ivan Levkivskyi and
Daniel Andrade Groppe.

changeset:   97857:7fc4306d537b
tag:         tip
parent:      97855:d433d74e0377
parent:      97856:8a7814a3613b
user:        Guido van Rossum <guido@dropbox.com>
date:        Thu Sep 10 10:54:10 2015 -0700
summary:     Restore doc updates to typing.rst by Ivan Levkivskyi and
Daniel Andrade Groppe.