Re: [Bug: 21.4.6] XEmacs tilde interpretation
Adrian Aichner wrote:
>>>>>"David" == David A Cobb
<Superbiskit(a)cox.net> writes:
>>>>>
>>>>>
<37 lines deleted by Adrian Aichner>
David> [Sorry for hopping in late!!]
David> I must be missing something. The only place special
David> handling of ~ or / seems appropriate in
David> minibuffer-read-file-name (or whatever) is when it is the
David> FIRST character typed by the user. Nicht wahr?
David> Whether it's electrical or post-<CR> scanning, it doesn't
David> make sense
David> to look elsewhere.
The idea, as I understand it, is that you do not need to erase
previous entry if you change your mind in the middle of a long path
entry.
e.g. in native XEmacs:
C-x d \ a \ b \ c \ d \ e \ \
electrically deletes to
C-x d \
This is annoying when done by accident, until you're tipped off (like
I was after a long time of use) that
f1 (advertised-undo)
will actually undo the electric delete.
Some for use of ~.
No, not the same for ~. If the user types / / (or \ \), that is
something that cannot appear legally in a path. She should expect some
special behavior ( or a error ). In fact, she probably knows what she's
doing and expects exactly what happens.
BUT, '~' in a filename is special only at the head of the path. To get
the self-correcting behavior for the tilde, the user should be typing /
/ ~ (or some such).
Not to stray too far afield, similar self-correcting behavior for the
colon (:) in Win NT-Native would be a nice feature -- capture the
preceding character as the drive-letter and toss out anything before that.
Would someone point me in the direction of the appropriate lisp routine?
--
David A. Cobb, Software Engineer, Public Access Advocate
"By God's Grace, I am a Christian man; by my actions a great sinner." -- The
Way of a Pilgrim: R.French, Tr.
Life is too short to tolerate crappy software!