4:09 p.m.
Stephen J. Turnbull wrote:
FKtPp writes:
> It seemed that, GNU's builtin #'overlays-in (beg end) supports:
>
> BEGin less than point-min, eg. 0
> END greater than point-max eg. (1+ (point-max)
>
> But XEmacs' didn't support these, and will report "Args out of
range"
> error.
Overlays are documented to be partially compatible, unsupported, and
deprecated. Policy has long been that code using them should be
converted to use extents. I personally don't consider this a bug for
that reason, at least not at the present time.
You're welcome to fix the overlay support that's in fsf-compat, as
an alternative to converting the code to use extents.
A branch of test shows that GNU's overlays- function can accept any
integer as BIG or END parameter -_-
---------------------------------->8----------------------------------
;; This buffer is for notes you don't want to save, and for Lisp
evaluation.
;; If you want to create a file, visit that file with C-x
C-f,
;; then enter the text in that file's own
buffer.
(make-overlay -100 400)
#<overlay from 1 to 255 in *scratch*>
(overlays-in -100 1000)
(#<overlay from 1 to 215 in *scratch*> #<overlay from 1 to 279 in
*scratch*> #<\
overlay from 1 to 317 in *scratch*> #<overlay from 254 to 255 in
*scratch*> #<o\
verlay from 276 to 277 in *scratch*>)
---------------------------------->8----------------------------------
But the resulted overlays are always in range [(point-min),(point-max)],
so it seemed that GNU's builtin
#'make-overlay <= checks beg and end, switch two parameter if end > beg
#'move-overlay <= checks beg and end, switch two parameter if end > beg
#'overlays-in <= checks beg and end, return nil if end > beg
attachment is a changed overlay.el that with above 3 #'functions emulate
the GNU's behavior.
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta
11:16 p.m.
New subject: overlay.el bug, incompatable overlays-in
It's me FKtPp ;) writes:
A branch of test shows that GNU's overlays- function can accept
any
integer as BIG or END parameter -_-
[[ *sigh* I really detest that style of programming. No wonder the
overlays emulation is deprecated. ]]
attachment is a changed overlay.el that with above 3 #'functions
emulate
the GNU's behavior.
Please submit it to XEmacs Patches as a patch against the Mercurial
repository. It will get faster review that way, at least from me.
You may also wish to put a copy on the tracker (this kind of thing
will eventually be done automatically but that is not implemented
yet).
I can tell you already that (1) the docstrings *need* to be amended to
describe the DWIM behavior, and (2) I'm almost certainly going to ask
that the normalization of arguments (null BUFFER -> current buffer,
truncating bounds and reordering them) be refactored as a defsubst.
Thanks!
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta
8:15 p.m.
New subject: overlay.el bug, incompatable overlays-in
Stephen J. Turnbull wrote:
It's me FKtPp ;) writes:
> A branch of test shows that GNU's overlays- function can accept any
> integer as BIG or END parameter -_-
[[ *sigh* I really detest that style of programming. No wonder the
overlays emulation is deprecated. ]]
> attachment is a changed overlay.el that with above 3 #'functions emulate
> the GNU's behavior.
Please submit it to XEmacs Patches as a patch against the Mercurial
repository. It will get faster review that way, at least from me.
You may also wish to put a copy on the tracker (this kind of thing
will eventually be done automatically but that is not implemented
yet).
I've created one with my two mail pasted in.
I can tell you already that (1) the docstrings *need* to be amended
to
describe the DWIM behavior, and (2) I'm almost certainly going to ask
that the normalization of arguments (null BUFFER -> current buffer,
truncating bounds and reordering them) be refactored as a defsubst.
Thanks!
What about this attached patch?
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta
4:15 a.m.
New subject: overlay.el bug, incompatable overlays-in
It's me FKtPp ;) wrote:
Stephen J. Turnbull wrote:
> It's me FKtPp ;) writes:
>
> > A branch of test shows that GNU's overlays- function can accept any
> > integer as BIG or END parameter -_-
>
> [[ *sigh* I really detest that style of programming. No wonder the
> overlays emulation is deprecated. ]]
>
> > attachment is a changed overlay.el that with above 3 #'functions emulate
> > the GNU's behavior.
>
> Please submit it to XEmacs Patches as a patch against the Mercurial
> repository. It will get faster review that way, at least from me.
> You may also wish to put a copy on the tracker (this kind of thing
> will eventually be done automatically but that is not implemented
> yet).
>
>
I've created one with my two mail pasted in.
issue301
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta
6116
days inactive
6118
days old
3 comments
2 participants
participants (2)
-
FKtPp -
Stephen J. Turnbull