bpo-41818: Add os.login_tty() for *nix.

[8vasu]

This follows #23536. Also, see #23546, #23686.

os.login_tty() calls the native login_tty() when it is available; otherwise, if setsid() is available and if TIOCSCTTY or ttyname() are available, it runs generic code to emulate login_tty().

Post #23686, #23546, #23740 goals:

1. add test_winsize() to "Lib/test/test_pty.py";
2. major revision of "Lib/pty.py", which has not happened since the beginning of the millennium.

Signed-off-by: Soumendra Ganguly soumendraganguly@gmail.com

https://bugs.python.org/issue41818

[8vasu]

@asvetlov Here is the implementation of os.login_tty() that you had suggested. Thank you for being patient.

Edit: This will help us make the code in "Lib/pty.py" and "Lib/test/test_pty.py" much simpler. cc: @ethanfurman @aeros @gpshead

[8vasu]

@zoulasc Sir, can you please take a look at the changes that I made to posixmodule.c? os_login_tty_impl() is the function that I mentioned earlier. Thank you for your kind help.

[zoulasc]

LGTM

[8vasu]

@zoulasc Thank you, sir.

[8vasu]

There were 2 commits that only changed the wording in the news file; rebased to have a single commit.

[github-actions]

This PR is stale because it has been open for 30 days with no activity.

[8vasu]

Still waiting for core review.

[github-actions]

This PR is stale because it has been open for 30 days with no activity.

[8vasu]

@gpshead This follows #23686; I need to update Modules/clinic/posixmodule.c.h but that can be done quickly using argument clinic when necessary. I request you to please review this if/when you have time.

[gpshead]

It looks like you're directly editing configure? this is generated code. edit the input configure.ac instead and regenerate configure. (i suspect might have just forgotten to commit your configure.ac change to the PR branch?)

[8vasu]

Sir, thank you for the prompt review. I will be honest; I actually did manually edit configure. I will learn how to do it and make the requested changes to configure.ac by Saturday morning.

[8vasu]

I have edited configure.ac and generated configure using autoconf now. Are there any steps that I might have missed/done improperly? When I generated the configure script, it added the runstatedir option to it which was not there before; please let me know if that was not supposed to happen/how to fix it. Also, it says that this branch has "conflicts that must be resolved"; Modules/clinic/posixmodule.c.h is the conflicting file; what is the best way to fix this problem; I want to avoid the XY problem this time!

[8vasu]

@gpshead Any updates on this?

[asvetlov]

I see a merge conflict now. Please fix.

[8vasu]

@asvetlov @gpshead The merge conflict in posixmodule.c.h was easy to fix but the ones in configure were a little harder; I realized that I had unknowingly added some unnecessary changes to configure since I did not have the "autoconf-archive" package installed; I decided to start fresh instead of rebasing; please check #29658. All other code remains unchanged.