ispell-get-line: Wrong type argument: stringp, ni

ispell-get-line: Wrong type argument: stringp, ni

[Tom Dye]

Aloha all,

I'm installing emacs 23 on my OS X 10.5 system.  The Emacs Wiki helped 
me with a configuration in .emacs to find Cocoaspell and its 
dictionaries.  These load now, but I get this error:

Spell-checking tsd_135_drr.tex using aspell with english dictionary...
Spell-checking region using aspell with english dictionary...done
ispell-get-line: Wrong type argument: stringp, nil

I googled a bit but didn't find anything useful.   If someone can get 
me started in the right direction, I'll be thankful.

All the best,
Tom
-- 
Tom Dye
T. S. Dye & Colleagues, Archaeologists, Inc.
Honolulu, Hawai`i

Re: ispell-get-line: Wrong type argument: stringp, ni

[Tom Dye]

On 2009-12-25 18:23:00 -1000, Tom Dye <tsd@tsdye.com> said:

> Aloha all,
> 
> I'm installing emacs 23 on my OS X 10.5 system.  The Emacs Wiki helped 
> me with a configuration in .emacs to find Cocoaspell and its 
> dictionaries.  These load now, but I get this error:
> 
> Spell-checking tsd_135_drr.tex using aspell with english dictionary...
> Spell-checking region using aspell with english dictionary...done
> ispell-get-line: Wrong type argument: stringp, nil
> 
> I googled a bit but didn't find anything useful.   If someone can get 
> me started in the right direction, I'll be thankful.
> 
> All the best,
> Tom

This problem "went away" without any apparent intervention from me.  
Looking at the cocoAspell web site again, perhaps this was the culprit:

> In some rare case the first time you spell check a document, cocoAspell 
> may fail to come up with a list of suggestions. It happens because the 
> spell server needs some time to load the dictionary and initialize 
> itself. The client application simply times out on the spell checking 
> request and behaves as there were no suggestions. This may not happen 
> at all if you have a fast machine.
Apologies for the noise level.

Tom

Tom Dye
T. S. Dye & Colleagues, Archaeologists, Inc.
Honolulu, Hawai`i

RE: ispell-get-line: Wrong type argument: stringp, ni

[Drew Adams]

> This problem "went away" without any apparent intervention from me.  
> Looking at the cocoAspell web site again, perhaps this was 
> the culprit:
> 
> > In some rare case the first time you spell check a 
> > document, cocoAspell may fail to come up with a list
> > of suggestions. It happens because the spell server
> > needs some time to load the dictionary and initialize 
> > itself. The client application simply times out on the 
> > spell checking request and behaves as there were no
> > suggestions. This may not happen at all if you have a
> > fast machine.
>
> Apologies for the noise level.

My guess is that the text you quoted is irrelevant here.

A wrong-type arg error is a product bug, even if it is often only a minor
problem. Please try to reproduce it, starting with `emacs -Q', and then report
it using `M-x report-emacs-bug'.

Often, such a bug occurs because of a lack of proper initialization of something
(typically a variable that is supposed to be a string is nil by default).

The same code that raises the error often performs initialization, so after the
error has been raised once, initialization has occurred and the error is never
raised again. That can explain why it "went away without any apparent
intervention from you." There is nevertheless a bug.

Re: ispell-get-line: Wrong type argument: stringp, ni

[Thomas S. Dye]

[-- Attachment #1: Type: text/plain, Size: 2303 bytes --]

Hi Drew,

On Dec 26, 2009, at 7:57 AM, Drew Adams wrote:

>> This problem "went away" without any apparent intervention from me.
>> Looking at the cocoAspell web site again, perhaps this was
>> the culprit:
>>
>>> In some rare case the first time you spell check a
>>> document, cocoAspell may fail to come up with a list
>>> of suggestions. It happens because the spell server
>>> needs some time to load the dictionary and initialize
>>> itself. The client application simply times out on the
>>> spell checking request and behaves as there were no
>>> suggestions. This may not happen at all if you have a
>>> fast machine.
>>
>> Apologies for the noise level.
>
> My guess is that the text you quoted is irrelevant here.
>
> A wrong-type arg error is a product bug, even if it is often only a  
> minor
> problem. Please try to reproduce it, starting with `emacs -Q', and  
> then report
> it using `M-x report-emacs-bug'.
>
> Often, such a bug occurs because of a lack of proper initialization  
> of something
> (typically a variable that is supposed to be a string is nil by  
> default).
>
> The same code that raises the error often performs initialization,  
> so after the
> error has been raised once, initialization has occurred and the  
> error is never
> raised again. That can explain why it "went away without any apparent
> intervention from you." There is nevertheless a bug.

I appreciate your perspective here and admire the combination of  
skills, knowledge, and intellect that would guide someone to reproduce  
the state of such a complex system.  That combination doesn't describe  
me, however, and I'm not able to reproduce the error message starting  
with emacs-Q (though I was pleased to find that I could do enough to  
raise two other errors that were familiar to me through my  
configuration trials).  I wish I was able to do more.

In response to your final paragraph, I did observe the error  
repeatedly as I tried the various menu items that launch a spell  
check: the various options under Edit/Spell and the AucTeX one under  
Command/Spell.  I don't know if this observation bears directly on the  
problem, but it came to mind when I read your last paragraph, and I am  
passing it along in the hope that it might prove interesting to you.

All the best,
Tom

[-- Attachment #2: Type: text/html, Size: 3551 bytes --]

RE: ispell-get-line: Wrong type argument: stringp, ni

[Drew Adams]

>> My guess is that the text you quoted is irrelevant here.
>> 
>> A wrong-type arg error is a product bug, even if it is often only a minor
>> problem. Please try to reproduce it, starting with `emacs -Q', and then
report
>> it using `M-x report-emacs-bug'.
>> 
>> Often, such a bug occurs because of a lack of proper initialization of
something
>> (typically a variable that is supposed to be a string is nil by default).
>> 
>> The same code that raises the error often performs initialization, so after
the
>> error has been raised once, initialization has occurred and the error is
never
>> raised again. That can explain why it "went away without any apparent
>> intervention from you." There is nevertheless a bug.
>
> I appreciate your perspective here and admire the combination
> of skills, knowledge, and intellect that would guide someone to
> reproduce the state of such a complex system.  That combination
> doesn't describe me, however, and I'm not able to reproduce the
> error message starting with emacs-Q (though I was pleased to
> find that I could do enough to raise two other errors that were
> familiar to me through my configuration trials).  I wish I was
> able to do more.
>
> In response to your final paragraph, I did observe the error
> repeatedly as I tried the various menu items that launch a spell
> check: the various options under Edit/Spell and the AucTeX one
> under Command/Spell.  I don't know if this observation bears
> directly on the problem, but it came to mind when I read your
> last paragraph, and I am passing it along in the hope that it
> might prove interesting to you.

OK, no problem. I meant only that if you _are_ able to reproduce it, then you
can help by reporting that.

Even a report that doesn't start with `emacs -Q' can often help. What you've
said so far might even be enough for someone to find the problem (if you report
it). The reason that `emacs -Q' is generally more helpful is that it eliminates
your init file as a possible source of the problem; that's all.

HTH.