8:33 a.m.
I don't know what happens in XEmacs--I only work on Emacs.
The code you sent creates overlays. They would continue to exist and
be displayed until something removes them with delete-overlay. There
seems to be code in flyspell.el that tries to delete them, so the
question is why it does not delete these overlays.
Manuel, could you debug it?
I still think the problem is in Emacs or in the
documentation about
overlay. In the section "Overlays properties" it is said:
-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----
`before-string'
This property's value is a string to add to the display at the
beginning of the overlay. The string does not appear in the
buffer in any sense--only on the screen.
`after-string'
This property's value is a string to add to the display at the end
of the overlay. The string does not appear in the buffer in any
sense--only on the screen.
-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----
So, I would have guessed that when a region using an overlay with
a before-string and an after-string properties is deleted, the before-string
and after-string disappear of the screen. That is, I have tried:
-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----
(require 'flyspell)
(defvar ov nil)
(defun add-ov (beg end)
(setq ov (make-overlay beg end nil t nil))
(overlay-put ov 'face 'flyspell-incorrect-face)
(overlay-put ov 'mouse-face 'highlight)
(overlay-put ov 'before-string "->")
(overlay-put ov 'after-string "<-"))
(defun del-ov ()
(delete-overlay ov))
-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----
Then, on a random buffer:
-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----
(add-ov 1 4)
(delete-region 1 4)
-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----
The strings "->" and "<-" are still on the screen. Is this
really normal?
For your information, XEmacs 21.4 removes the strings.
Anyway, I will fix flyspell.el so that it runs correctly with Emacs 21.1.
--
Manuel
ps: I'm using Emacs 21.1.1 on a Linux box.
11:28 a.m.
New subject: [Manuel Serrano <Manuel.Serrano@sophia.inria.fr>] Re: flyspell before and after-string
>>>> "Manuel" == Manuel Serrano
<Manuel.Serrano(a)sophia.inria.fr> writes:
Manuel> The strings "->" and "<-" are still on the
screen. Is this
Manuel> really normal? For your information, XEmacs 21.4 removes
Manuel> the strings.
In XEmacs the overlay API is implemented using the extent primitives,
and by default new extents are "detachable":
------------------------------------------------------------------------
Detached Extents
================
A detached extent is an extent that is not attached to a buffer or
string but can be re-inserted. Detached extents have a start position
and end position of `nil'. Extents can be explicitly detached using
`detach-extent'. An extent is also detached when all of its characters
are all killed by a deletion, if its `detachable' property is set; if
this property is not set, the extent becomes a zero-length extent.
------------------------------------------------------------------------
See also C-h f set-extent-property RET.
So if you set the detachable property to nil, you should get the same
behavior as in Emacs. Depending on how you manipulate those extents
you may prefer to do
(when (featurep 'xemacs)
(set-extent-property ov 'detachable nil))
at creation rather than checking for xemacs at deletion time. This
seems to expose a bug in XEmacs's overlay handling (for zero-length
overlays), but it's irrelevant if you're just going to go ahead and
delete it anyway.
HTH
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Don't ask how you can "do" free software business;
ask what your business can "do for" free software.
2:15 p.m.
New subject: [Manuel Serrano <Manuel.Serrano@sophia.inria.fr>] Re: flyspell before and after-string
>>>>> "Manuel" == Manuel Serrano
<Manuel.Serrano(a)sophia.inria.fr> writes:
Manuel> The strings "->" and "<-" are still on the
screen. Is this
Manuel> really normal? For your information, XEmacs 21.4 removes
Manuel> the strings.
In XEmacs the overlay API is implemented using the extent primitives,
and by default new extents are "detachable":
------------------------------------------------------------------------
Detached Extents
================
A detached extent is an extent that is not attached to a buffer or
string but can be re-inserted. Detached extents have a start position
and end position of `nil'. Extents can be explicitly detached using
`detach-extent'. An extent is also detached when all of its characters
are all killed by a deletion, if its `detachable' property is set; if
this property is not set, the extent becomes a zero-length extent.
------------------------------------------------------------------------
See also C-h f set-extent-property RET.
So if you set the detachable property to nil, you should get the same
behavior as in Emacs. Depending on how you manipulate those extents
you may prefer to do
(when (featurep 'xemacs)
(set-extent-property ov 'detachable nil))
at creation rather than checking for xemacs at deletion time. This
seems to expose a bug in XEmacs's overlay handling (for zero-length
overlays), but it's irrelevant if you're just going to go ahead and
delete it anyway.
HTH
I certainly don't want the Emacs behavior!
There is something I don't understand with Emacs overlay. I put an
overlay at some location on a buffer. I put a before-string for that
overlay. Now I edit this buffer and I kill the line containing the
overlay. With Emacs the string attached to the before-string property
will stay still on the buffer and I have no way to get rid of it using
user commands. I have to use Elisp code to get rid of the before string.
To my opinion this is cumbersome (most likely I miss understand
the correct way to use overlay...).
--
Manuel
2:55 p.m.
New subject: [Manuel Serrano <Manuel.Serrano@sophia.inria.fr>] Re: flyspell before and after-string
> I don't know what happens in XEmacs--I only work on Emacs.
>
> The code you sent creates overlays. They would continue to exist and
> be displayed until something removes them with delete-overlay. There
> seems to be code in flyspell.el that tries to delete them, so the
> question is why it does not delete these overlays.
>
> Manuel, could you debug it?
I still think the problem is in Emacs or in the documentation about
overlay. In the section "Overlays properties" it is said:
-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----
`before-string'
This property's value is a string to add to the display at the
beginning of the overlay. The string does not appear in the
buffer in any sense--only on the screen.
`after-string'
This property's value is a string to add to the display at the end
of the overlay. The string does not appear in the buffer in any
sense--only on the screen.
-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----
So, I would have guessed that when a region using an overlay with
a before-string and an after-string properties is deleted, the before-string
and after-string disappear of the screen. That is, I have tried:
-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----
(require 'flyspell)
(defvar ov nil)
(defun add-ov (beg end)
(setq ov (make-overlay beg end nil t nil))
(overlay-put ov 'face 'flyspell-incorrect-face)
(overlay-put ov 'mouse-face 'highlight)
(overlay-put ov 'before-string "->")
(overlay-put ov 'after-string "<-"))
(defun del-ov ()
(delete-overlay ov))
-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----
Then, on a random buffer:
-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----
(add-ov 1 4)
(delete-region 1 4)
-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----
The strings "->" and "<-" are still on the screen. Is this
really normal?
Check out the `evaporate' property of overlays.
You could set it to t so that your overlays get automatically
deleted when their size gets down to 0.
Stefan
3:08 p.m.
New subject: [Manuel Serrano <Manuel.Serrano@sophia.inria.fr>] Re: flyspell before and after-string
>>>> "Manuel" == Manuel Serrano
<Manuel.Serrano(a)sophia.inria.fr> writes:
Manuel> To my opinion this is cumbersome (most likely I miss
Manuel> understand the correct way to use overlay...).
Emacs 21's Lispref describes this property for overlays:
------------------------------------------------------------------------
＠item evaporate
＠kindex evaporate ＠r{(overlay property)}
If this property is non-＠code{nil}, the overlay is deleted automatically
if it ever becomes empty (i.e., if it spans no characters).
------------------------------------------------------------------------
According to etc/ONEWS this was introduced in 19.29, so you can
probably rely on it even for very old Emacsen.
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Don't ask how you can "do" free software business;
ask what your business can "do for" free software.
3:15 p.m.
New subject: [Manuel Serrano <Manuel.Serrano@sophia.inria.fr>] Re: flyspell before and after-string
>>>> "Stefan" == Stefan Monnier
<monnier+gnu/emacs(a)rum.cs.yale.edu> writes:
Stefan> Check out the `evaporate' property of overlays. You could
Stefan> set it to t
Is there a reason why this doesn't default to t? (I've never heard
anybody complain that the "detachable" property should default to nil
for extents, so I'm curious.)
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Don't ask how you can "do" free software business;
ask what your business can "do for" free software.
3:40 p.m.
New subject: [Manuel Serrano <Manuel.Serrano@sophia.inria.fr>] Re: flyspell before and after-string
Check out the `evaporate' property of overlays.
You could set it to t so that your overlays get automatically
deleted when their size gets down to 0.
Great. Many thanks for the information. I
will be able to produce a better
support for Emacs. The new version is available at:
http://www-sop.inria.fr/mimosa/personnel/Manuel.Serrano/flyspell/flyspell...
--
Manuel
8122
days inactive
8122
days old
6 comments
3 participants
participants (3)
-
Manuel Serrano -
Stefan Monnier -
Stephen J. Turnbull