error when marking messages in Chinese expirable

error when marking messages in Chinese expirable

[Max Froumentin]

Since a few releases ago (Oort 0.13 or 0.14), when I mark an article
expirable, if that summary line happens to have non-ASCII characters I
get the error:

gnus-summary-update-mark: Characters in subst-char-in-region have
different byte-lengths

From what I could figure out, it happens only on lines with Chinese characters.

Sorry if this has already been reported.

-- 
Max.

Re: error when marking messages in Chinese expirable

[Kai Großjohann]

Max Froumentin <mf@w3.org> writes:

> Since a few releases ago (Oort 0.13 or 0.14), when I mark an article
> expirable, if that summary line happens to have non-ASCII characters I
> get the error:
>
> gnus-summary-update-mark: Characters in subst-char-in-region have
> different byte-lengths

Can you provide a backtrace?  M-x toggle-debug-on-error RET, then
repeat the error.

Does it also happen when killing things?  I killed some Chinese spam
today, but got no error.
-- 
file-error; Data: (Opening input file no such file or directory ~/.signature)

Re: error when marking messages in Chinese expirable

[Max Froumentin]

kai.grossjohann@gmx.net (Kai Großjohann) wrote:

> Can you provide a backtrace?  M-x toggle-debug-on-error RET, then
> repeat the error.

Sorry, forgot to include it:

Debugger entered--Lisp error: (error "Characters in subst-char-in-region have different byte-lengths")                                                         
  subst-char-in-region(1800 1801 43041 32)
  gnus-summary-update-mark(32 score)
  gnus-summary-update-line(nil)
  gnus-summary-update-mark(69 unread)
  gnus-summary-mark-article-as-read(67)
  gnus-summary-catchup(nil nil nil fast)
  gnus-summary-catchup-and-exit(nil)
  call-interactively(gnus-summary-catchup-and-exit)

and my emacs-version: "GNU Emacs 21.2.1 (i386-debian-linux-gnu, X toolkit, Xaw3d scroll bars) of 2002-03-22 on raven, modified by Debian"

and my gnus-version: "Oort Gnus v0.18"

> Does it also happen when killing things?  I killed some Chinese spam
> today, but got no error.

I have been trying for a week but didn't get any spam with offending
Chinese characters in the subject, so I couldn't test if C-k trigered
the error. But I know that marking them expirable with 'E' does
trigger the error.

Max.

Re: error when marking messages in Chinese expirable

[Kai Großjohann]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset=iso-2022-jp-2, Size: 1080 bytes --]

Max Froumentin <mf@w3.org> writes:

> kai.grossjohann@gmx.net (Kai Gro^[.A^[N_johann) wrote:
>
>> Can you provide a backtrace?  M-x toggle-debug-on-error RET, then
>> repeat the error.
>
> Sorry, forgot to include it:
>
> Debugger entered--Lisp error: (error "Characters in subst-char-in-region have different byte-lengths")                                                         
>   subst-char-in-region(1800 1801 43041 32)
>   gnus-summary-update-mark(32 score)
>   gnus-summary-update-line(nil)
>   gnus-summary-update-mark(69 unread)
>   gnus-summary-mark-article-as-read(67)
>   gnus-summary-catchup(nil nil nil fast)
>   gnus-summary-catchup-and-exit(nil)
>   call-interactively(gnus-summary-catchup-and-exit)

Ah.  Point is at a weird character when that happens.  Can you look
where point is in the summary buffer?

(Does anyone know what character is 43041?  Hm.  Ah: ^[$AP!^[(B.)

I guess that point is somewhere in the subject header.  But why did
it get there?  I don't grok that code.
-- 
file-error; Data: (Opening input file no such file or directory ~/.signature)

Re: error when marking messages in Chinese expirable

[Max Froumentin]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset=iso-2022-jp-2, Size: 330 bytes --]

kai.grossjohann@gmx.net (Kai Gro^[.A^[N_johann) wrote:

> Ah.  Point is at a weird character when that happens.  Can you look
> where point is in the summary buffer?
>
> (Does anyone know what character is 43041?  Hm.  Ah: ^[$AP!^[(B.)

Exactly. Point is each time on the apprently offending character in the 
subject header.

Max.

Re: error when marking messages in Chinese expirable

[Kai Großjohann]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset=iso-2022-jp-2, Size: 561 bytes --]

Max Froumentin <mf@w3.org> writes:

> kai.grossjohann@gmx.net (Kai Gro^[.A^[N_johann) wrote:
>
>> Ah.  Point is at a weird character when that happens.  Can you look
>> where point is in the summary buffer?
>>
>> (Does anyone know what character is 43041?  Hm.  Ah: ^[$AP!^[(B.)
>
> Exactly. Point is each time on the apprently offending character in the 
> subject header.

So how did point get there?  I understand that point is supposed to
be on the mark that's being changed.

-- 
file-error; Data: (Opening input file no such file or directory ~/.signature)

Re: error when marking messages in Chinese expirable

[Max Froumentin]

kai.grossjohann@gmx.net (Kai Großjohann) wrote:

> So how did point get there?  I understand that point is supposed to
> be on the mark that's being changed.

When I do catchup ('C') on my spam group, if the summary has an offending
character the error is raised, and the point is on the offending character.
If I press C again, it moves on to the next, until the whole summary
has been marked expirable.

Cheers,

Max.

Re: error when marking messages in Chinese expirable

[Christoph Garbers]

* Max Froumentin <mf@w3.org>:

>> So how did point get there?  I understand that point is supposed to
>> be on the mark that's being changed.
> 
> When I do catchup ('C') on my spam group, if the summary has an offending
> character the error is raised, and the point is on the offending character.
> If I press C again, it moves on to the next, until the whole summary
> has been marked expirable.

Exactly the same here.

Christoph

Re: error when marking messages in Chinese expirable (here's one:

[Lars Magne Ingebrigtsen]

Max Froumentin <mf@w3.org> writes:

> Debugger entered--Lisp error: (error "Characters in subst-char-in-region have different byte-lengths")                                                         
>   subst-char-in-region(1800 1801 43041 32)
>   gnus-summary-update-mark(32 score)

So you can't mark this article as expirable?  

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: error when marking messages in Chinese expirable (here's one:

[Lars Magne Ingebrigtsen]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> So you can't mark this article as expirable?  

Oops.  I tried putting some Chinese text in the subject, but that
didn't work.

Anyway, I seem to be unable to reproduce this bug...

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: error when marking messages in Chinese expirable

[Christoph Garbers]

* kai.grossjohann@gmx.net (Kai Großjohann):

>> Since a few releases ago (Oort 0.13 or 0.14), when I mark an article
>> expirable, if that summary line happens to have non-ASCII characters I
>> get the error:
>>
>> gnus-summary-update-mark: Characters in subst-char-in-region have
>> different byte-lengths
> 
> Can you provide a backtrace?  M-x toggle-debug-on-error RET, then
> repeat the error.

I get the same error when moving an e-mail with Chinese characters from
one nnml-group to another:

Debugger entered--Lisp error: (error "Characters in subst-char-in-region have different byte-lengths")
  subst-char-in-region(899 900 43472 32)
  gnus-summary-update-mark(32 score)
  gnus-summary-update-line(nil)
  gnus-summary-update-mark(71 unread)
  gnus-summary-mark-article(27751 71)
  gnus-summary-move-article(nil)
  call-interactively(gnus-summary-move-article)
 
Christoph