Re: clear_session, store_replacement, and support

"Gary Setter" <[email protected]>
Newsgroups gmane.comp.gnu.aspell.devel
Message-ID <001801c595f6$47c1bde0$19144a0c@pavilion>
----- Original Message ----- 
From: "Bill Moseley" <[email protected]>
To: "Gary Setter" <[email protected]>
Cc: <[email protected]>
Sent: Friday, July 29, 2005 9:58 AM
Subject: Re: [aspell-devel] clear_session, store_replacement, and
support


> Thanks for the reply, Gary:
>
> On Fri, Jul 29, 2005 at 08:35:24AM -0500, Gary Setter wrote:
> > ___POINT 1___
> > The function seems to be in error. Take a look at the
> > implementation of store_replacement.
> >
> >   PosibErr<void> SpellerImpl::store_replacement(const String
&
> > mis,
> >                                                 const String
&
> > cor,
> >                                                 bool memory)
> >
> > It returns a void, not a bool. There is no data returned.
>
> Is the API *suppose* to return a value?  And if false should
the
> error number and error message be available with
aspell_speller_error_number
> and aspell_speller_error_message?  That's what I'm not clear
on.

I didn't design aspell, but it sure looks to me it is suppose to
return success/failure status. The c lib functions do not seem
consistent in how they report failure. Since it clearly return -1
if the function call failed, I submitted a patch to make this
function return 0 on success. Aspell seems to work with error
strings, not error numbers. If (after patching, making, and
installing aspell) our favorite function returns not zero, you
can call aspell_speller_error_message() to get the current error.

<snip>
> Again, I'd like to first make sure that I'm using clear_session
> correctly and I understand the Aspell usage correctly.
>

I made a version of one of my programs that uses aspell to
exercise the session functions. I found that they do return true
on success. I found that after I add a 'misspelled word to the
session dictionary that the check function accepts the word.
After the session dictionary is cleared, the check function stops
accepting the word.
Maybe that would be a more straightforward test for Text::Aspell?
Maybe someone else from the mailing list has a suggestion?
<snip>

> PS -- one other question:  I require the en_US dictionary
installed
> for testing.  Of course, not everyone has that dictionary, but
I'm
> not sure how I can test Aspell without knowing a specific
dictionary
> is installed -- I want to make sure I can lookup misspelled
words, so
> seems like I need a specific dictionary to do that.  If you or
anyone
> has a better idea I'd like to hear.  Also, I'm not 100% clear
how to
> verify that the dictionary I need is installed -- but I haven't
> looked at that yet.
>
Just off the top of my head, you already warn the users that the
dictionary is needed. You could include a note on what to look
for to verify that the dictionary is found ( or not found). You
could provide a sample aspell.conf file and suggest how to
configure it to find the dictionary and data files.

Does anyone else have better ideas?

Best regards,
Gary Setter

I had something of an adventure building the Text::Aspell add-in.
The test you provided is failing with a freeze in what appears to
be the Tex test. I'll send you my notes on what I did to get the
add-in built in case it is of any help.
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.