12:22 a.m.
So today I decide that it would be nifty if you could spell-check
comments in code, and I discovered the `ispell-comments-and-strings'
function. Actually the context was showing a colleague how XEmacs is
the slickest thing since icy roads, and how it could do *anything*.
Imagine my embarrassment and utter dismay when
M-x ispell-comments-and-strings gave me "Unterminated string or comment". ;-)
{needless to say, the buffer did not contain any unterminated strings
or comments}
I've observed this misbehaviour with 21.1.14 on Solaris and 21.1.12 on Linux.
Searching the archives I see that Hrovje reported the same bug back in
2000
http://list-archive.xemacs.org/xemacs-beta/200005/msg00342.html
and had a possible workaround. Was there ever any follow-up to this
bug report? (Nothing I can find in the archives)
There is also a newer version of ispell.el (3.5) available at
http://kdstevens.com/~stevens/ispell-page.html.
The ispell.el bundled with XEmacs-21.1. is version 3.3. But the newer
version 3.5 still gives me the same "Unterminated string or comment"
bug.
Is this an XEmacs bug (syntax-table breakage) or an ispell.el bug?
Any expert opinions?
------------------------------------------------------------------
12:38 a.m.
>>>> "CW" == Charles G Waldman <Charles>
writes:
CW> So today I decide that it would be nifty if you could spell-check
CW> comments in code, and I discovered the `ispell-comments-and-strings'
CW> function. Actually the context was showing a colleague how XEmacs is
CW> the slickest thing since icy roads, and how it could do *anything*.
CW> Imagine my embarrassment and utter dismay when
CW> M-x ispell-comments-and-strings gave me "Unterminated string or comment".
;-)
CW> {needless to say, the buffer did not contain any unterminated strings
CW> or comments}
Which language were you trying? I'm using this sometimes with
Java/JDE and it works for me there?
XEmacs 21.4 (patch 5) "Civil Service" [Lucid] (sparc-sun-solaris2.8)
of Tue Oct 23 2001 on jena
Thanks,
-- Marco
CW> I've observed this misbehaviour with 21.1.14 on Solaris and 21.1.12 on Linux.
CW> Searching the archives I see that Hrovje reported the same bug back in
CW> 2000
CW> http://list-archive.xemacs.org/xemacs-beta/200005/msg00342.html
CW> and had a possible workaround. Was there ever any follow-up to this
CW> bug report? (Nothing I can find in the archives)
CW> There is also a newer version of ispell.el (3.5) available at
CW> http://kdstevens.com/~stevens/ispell-page.html.
CW> The ispell.el bundled with XEmacs-21.1. is version 3.3. But the newer
CW> version 3.5 still gives me the same "Unterminated string or comment"
CW> bug.
CW> Is this an XEmacs bug (syntax-table breakage) or an ispell.el bug?
CW> Any expert opinions?
CW> ------------------------------------------------------------------
--
12:47 a.m.
Marco Walther writes:
Which language were you trying? I'm using this sometimes with
Java/JDE and it works for me there?
XEmacs 21.4 (patch 5) "Civil Service" [Lucid] (sparc-sun-solaris2.8)
of Tue Oct 23 2001 on jena
Language = C / English.
I think that language is not the issue... the difference is that you
are running xemacs-21.4.5 and I'm using xemacs-21.1.14. I don't have
a 21.4 to test with. (I've been quite hesitant to trade in my lovely
Channel Islands for a less stable version with a bunch of enhancements
which don't appeal to my tastes)
1:39 a.m.
>>>> "CW" == Charles G Waldman <Charles>
writes:
CW> Marco Walther writes:
>
> Which language were you trying? I'm using this sometimes with
> Java/JDE and it works for me there?
>
> XEmacs 21.4 (patch 5) "Civil Service" [Lucid] (sparc-sun-solaris2.8)
> of Tue Oct 23 2001 on jena
CW> Language = C / English.
CW> I think that language is not the issue... the difference is that you
CW> are running xemacs-21.4.5 and I'm using xemacs-21.1.14. I don't have
Confirmed! I'm seeing the same problem on
XEmacs 21.1 (patch 14) "Cuyahoga Valley" [Lucid] (sparc-sun-solaris2.8) of Mon
Feb 5 2001 on jena
with basically the same xemacs-packages etc.
So it looks like this is really a problem of the `core' XEmacs.
Thanks,
-- Marco
CW> a 21.4 to test with. (I've been quite hesitant to trade in my lovely
CW> Channel Islands for a less stable version with a bunch of enhancements
CW> which don't appeal to my tastes)
--
1:59 a.m.
On Thu, 28 Feb 2002, cgw(a)alum.mit.edu wrote:
Marco Walther writes:
>
> Which language were you trying? I'm using this sometimes with
> Java/JDE and it works for me there?
>
> XEmacs 21.4 (patch 5) "Civil Service" [Lucid] (sparc-sun-solaris2.8)
> of Tue Oct 23 2001 on jena
Language = C / English.
I think that language is not the issue... the difference is that you
are running xemacs-21.4.5 and I'm using xemacs-21.1.14. I don't have
a 21.4 to test with. (I've been quite hesitant to trade in my lovely
Channel Islands for a less stable version with a bunch of enhancements
which don't appeal to my tastes)
I can verify that this problem does not occur in 21.4.6 and that it
does occur is 21.1. The backtrace is not overly interesting as
ispell-comments-and-strings is calling error based on the results of
parse-partial-sexp.
(defun ispell-comments-and-strings ()
....
(setq state (parse-partial-sexp start (point-max)
nil nil state 'syntax-table))
(if (or (nth 3 state) (nth 4 state))
(error "Unterminated string or comment"))
...)
I would recommend upgrading to 21.4.6 over 21.1.14.
BTW, you can have flyspell enabled in programming modes to check
comments as you type them.
-jeff
8302
days inactive
8302
days old
4 comments
3 participants
participants (3)
-
Charles Waldman -
Jeff Mincy -
Marco Walther