Re: [aspell-devel] New dictionary?

[Jose Da Silva]

On June 20, 2005 09:33 am, Richard Shewmaker wrote:
> Hi. I'm not programming-friendly (although not totally uninformed).
> I'm wondering if there is a way for me to create a new dictionary
> from a list of words I assemble. I looked your site over but could
> not find anything which appeared to do what I'd like to do.

Aspell 0.6x is currently a wordlist-based spell checker.
This is an example of an Aspell wordlist:
airplane
algebra
apple

A dictionary needs to have a word with a definition attached which is 
missing information as per Aspell 0.5x and 0.6x wordlists right now.
This is an example of what you need for a dictionary:
airplane, air|plane, (ar-plan) A flying machine......
algebra, al|ge|bra, (al-je-bre) branch of mathematics....
apple, ap|ple, (ap'el) round fruit grown on a tree....

Looking above, there are desirable features in a dictionary, such as 
correct locations for word-breaks and the sounds-like included in the 
dictionary definition which may prove useful for finding additional 
spelling errors plus sounds-like substitutions. Perhaps features like 
that may eventually work their way into Aspell, but not yet at the 
moment with the 0.60 version (read the TODO for Aspell).

> What I'm interested in making is an HTML dictionary for UltraEdit. As
> you most likely know, the dictionaries currently available (to my
> knowledge) do not recognize HTML (as well as CSS) tags, and so spell
> checking coded documents is unrealistic.

You may want to look at gnucide or freedict for direction and 
information since the results are closer to what you are looking for.

> Thank you in advance. I hope I'm not using valuable time!

If you think Aspell could use some dictionary qualities, such as 
word-break information or sounds-like features and are willing to help 
in some manner or other (such as creating improved wordlists, 
whatever?), then the time spent was well worth it.  :-)
Please check if the maintainer is interested, then if you are interested 
in pursuing these, then check these since they would benefit from 
enhanced-type wordlists:

This would benefit from word-breaks in correct location:
http://sourceforge.net/tracker/index.php?func=detail&aid=614305&group_id=245&atid=350245

This may benefit from sounds-like related information:
http://sourceforge.net/tracker/index.php?func=detail&aid=613639&group_id=245&atid=350245