cleanup of exit/quit commands for qt console

[eteq]

This implements Fernando's suggestion from the list of removing the exit/quit Quitter objects that were previously used in ipython (and moving the quitter to deathrow, as it's used nowhere else), in favor of only an updated %exit magic (still also aliased to quit, Exit, and Quit).

The re-worked exit magic tor the qtconsole also now bypasses the prompt about closing the kernel, and supports a "-k" option that leaves the kernel running but closes the qt window. This also required making some changes for the pure python version of qtconsole, but exit() wasn't working at all previously - it was catching the SystemExit exception the builtin python quit/exit uses.

I also tried merging this branch into another branch tracking the upstream trunk, and got a conflict due to some changes with the qtconsole exit prompt, although it's straightforward to merge. I'm not clear on the correct procedure for the pull request, though - I gather from the mailing list that I'm not supposed to merge in the upstream trunk here, but my git-fu is not very strong, so I'm not sure how I am supposed to include this merged fix... Or do I leave that to whomever accepts the pull request?

[minrk]

Thanks for this, sorry about the conflict, that's my code in your way. It should indeed be quite a straightforward merge.

You can give trunk a clean merge if you rebase your branch onto master, and handle the conflict there.

Does this specifically link %exit to the requesting frontend's closeEvent, is that how it works? Or are there other things happening as well?

Thanks for this!

[eteq]

Ok, I rebased the qt-exiting branch to the current head, so it should now pull without any problem. However, it doesn't appear to be reflected in this pull request... do I have to re-submit the pull request, or do something to directly update the pull request?

[fperez]

Did you push? I was literally right now reviewing your code and was about to merge it :)

[fperez]

We can meet briefly on #ipython (freenode IRC) if you want.

[eteq]

Well, when I look back at my eteq/ipython/qt-exiting branch, it says the last update was a few minutes ago... but it doesn't seem to be reflected in this pull request. Is it supposed to be automatically?

And might I say, top-notch service for this project - literally withing minutes! :)

[minrk]

Note that the times of commits don't have to change when you rebase, so you can have a commit from a week ago occur after one from today, if you rebased on top of it.

I just did steps 1-2 of the merge, and it applied cleanly, which it didn't before, so it looks like what's on GH is your rebased version.

[eteq]

Fernando pointed out some test suit failures that were due to the added support for hiding builtins that was necessary for the 'exit' and 'quit' python buildins. These are fixed by the commits I just sent up - the test suit now passes, and the branch merges cleanly.

@minrk Ah, I see now - I partly misunderstood how rebase was supposed to work.

[fperez]

Great job, many thanks! Merging and closing in a second. Welcome to the team :)