Fix sage keyboard trap

[davidfarmer]

Once you tab into a sage cell, there is no obvious way to get out.

I am pretty sure I can fix this the same way as we esc to get out of a knowl.
It is just a matter of keeping track of where to go when you hit esc.

[davidfarmer]

In fact, if you open a knowl containing a sage cell, get into the cell, and then hit esc,
you jump out of the sage cell and now are sitting on the knowl link. That is not optimal,
but at least you were not trapped.

Should esc in a sage cell take you to the object immediately after the cell (which presumably
is the Evaluate button?)

[rbeezer]

I'm sure this sort of thing is a big part of what Sina wants to inform us about.

…

On 05/25/2018 11:04 AM, David W. Farmer wrote: In fact, if you open a knowl containing a sage cell, get into the call, and then hit esc, you jump out of the sage cell and now are sitting on the knowl link. That is not optimal, but at least you were not trapped. Should esc in a sage cell take you to the object immediately after the cell (which presumably is the Evaluate button?) — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#888 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABy2cqP1zfS3QvzrJBGddhU-V09-XnnMks5t2EesgaJpZM4UOXn7>.

[Alex-Jordan]

From Jim Fowler in an email:

If the problem is the code editor, the sagecell code editor could be
replaced. (The mathjax code I have implementing \sage separates the
sagecell widget from the execution of the sage code.)

[kcrisman]

From Jim Fowler in an email: If the problem is the code editor, the sagecell code editor could be replaced. (The mathjax code I have implementing \sage separates the sagecell widget from the execution of the sage code.)

Fair enough, but just be sure to let upstream know about it too then; maybe they should just switch from CodeMirror or whatever Jason and now Andrey are using.