Re: [libreoffice-users] Specialty Dictionaries

I wonder how hard it would be, or how confusing it would be to make a 2 
word reference in the word list file.  If the "space" appears "blank", 
then what would happen if there was some other "blank" character that 
looked like the "space" character.

I think the originally defined system came from some other organization, 
like Hunspell or Myspell.  It may take a lot of work, and make all of 
the current .oxt dictionaries need to be re-done if there was a new 
formatting system that allowed a 2 word listing.

My question is why would one be needed?  It the only reason for it is 
one combination of terms is valid, while a different ending term is 
not?  That would make some real context oriented spelling and worse that 
figuring out grammar issues. You will need to have defines all the 
possible two word term combinations instead of having each single 
word/term checked for proper spelling.

The more I think about it, the more I want to cringe over what might be 
needed to get it working fully and not mess with defined characters for 
non-English fonts or other issues for taking over a pre-defined 
character for some internal function, when that character may be mapped 
for a glyph/letter for some font used by some user.   Then that user[s] 
would have their document messed up in some way.

Still, if someone can figure out all the issues and work out a way that 
they will not cause a set of users problems down the line, then by all 
means create a request for a modification of the spell checker system.  
I really doubt that it will be changed form the current system, but you 
can try.


On 07/07/2012 06:22 PM, Tom Davies wrote:
> Hi :)
> It sounds like interesting and useful functionality that might well be worth adding if it hasn't been done already.  I 
> think a lot of us here have been focussing on work-arounds "to just get the job done".  But i think it might 
> be a good idea to post a bug-report and make it a "Feature request".
> Regards from
> Tom :)
>
>
> --- On Fri, 6/7/12, Simon Cropper <simoncropper@fossworkflowguides.com> wrote:
>
> From: Simon Cropper <simoncropper@fossworkflowguides.com>
> Subject: Re: [libreoffice-users] Specialty Dictionaries
> To: users@global.libreoffice.org
> Date: Friday, 6 July, 2012, 1:19
>
> On 06/07/12 05:26, nvrk wrote:
> > On Thu, Jul 5, 2012 at 3:06 AM, Simon Cropper <
> > simoncropper@fossworkflowguides.com> wrote:
> >
> > > Yeah, I have thought of both these things. Have hacked a standard file
> > > before, particularly in MS Word. Easily done assuming it is a text file and
> > > not a binary file.
> > >
> > > The problem is the binomial. I also thought of the concatenation string
> > > but most of the single characters have been used for have special meanings
> > > in various word processors. Hyphens for example are used in LO as hyphens
> > > and so how would you know when removing the character at the end of your
> > > report is complete, what is a concatenation character and what is a real
> > > hyphen?
> > >
> > > In other situations I have used =!= as a joining string but as stated it
> > > is messy and hard to read.
> > An underscore works well.
> Yeap, but an underscore gets used a lot in technical reports (e.g. in a URL). If you do a global search and replace to 
> remove the character at the end of writing so the report looks "clean and well presented" you neuter the URL 
> or "corrupt" the other text string that uses it.
>
> As an alternative to "free form" typing of jargon or technical terms then running a spell checker, terms 
> could be inserted from a list. This works OK but in the absence of LO integration you can't flag the inserted text as 
> 'hey this is jargon, I just inserted it from a secure source, don't bother spell checking'. This can be done but 
> requires you to manually apply language characteristics to hundreds or thousands of names, or alternatively hit 
> "ignore" the same number of times with the spell checker. :(
>
> Ideally you need a blank concatenation character that is recognizing by LO as linking two words (such as a 
> non-breaking space already available in LO but does not necessarily have to physically bind the words 
> together but would need to be seen by the spell checker as a joining character) AND IS RECOGNIZED by the 
> spell checker, substituted with something like an underscore and compared to the lists in the dic files which 
> would appear as "Eucalyptus_vulgaris". I just need someone in the know to be able to insert this 
> functionality and these problems would be solved.
>
> -- Cheers Simon
>
>     Simon Cropper - Open Content Creator
>
>     Free and Open Source Software Workflow Guides
>     ------------------------------------------------------------
>     Introduction               http://www.fossworkflowguides.com
>     GIS Packages           http://www.fossworkflowguides.com/gis
>     bash / Python    http://www.fossworkflowguides.com/scripting
>
>
>
> -- For unsubscribe instructions e-mail to: users+help@global.libreoffice.org
> Problems? http://www.libreoffice.org/get-help/mailing-lists/how-to-unsubscribe/
> Posting guidelines + more: http://wiki.documentfoundation.org/Netiquette
> List archive: http://listarchives.libreoffice.org/global/users/
> All messages sent to this list will be publicly archived and cannot be deleted



--
For unsubscribe instructions e-mail to: users+help@global.libreoffice.org
Problems? http://www.libreoffice.org/get-help/mailing-lists/how-to-unsubscribe/
Posting guidelines + more: http://wiki.documentfoundation.org/Netiquette
List archive: http://listarchives.libreoffice.org/global/users/
All messages sent to this list will be publicly archived and cannot be deleted