[quote user="fojtik"]

Some time before, I have created special program for harvesting internet and create such a list of words. So I could attempt to reuse this program and create a valid list of czech words anew.

This was a main reason why I has been asking for possibility to add Czech dictionary. I can even help with creating license free word list for other languages.

If I send Email to David Harris, I am afraid that he redirects all Emails to /dev/null.

[/quote]

I contact David to see with him what we can do, the tools he uses need to adapt the data files, because there is no accentuated characters in English...

Regards

Dears,

 I am interested in a Czech spell dictionary. It is not available, so I could create it (and if I succeed I could provide it for others). Is there any way, how to create pmd file?

Do you have any tool that could help me with it?

thank you very much

   J.F.
 

Is it top secret the spell checker structure and intention to have only English spell checker?

The Spell Checker in Pegasus Mail is proprietary and the license material is not in the public domain. So building additional dictionaries is out of the question for now. I have a demo of a SpellChecker that supports Czech word checking. It can be used to check words in the clipboard and correct them. It is based on ISpell

Martin

Dear Martin,

Thank you very much for your response. It is not convenient to use external application. If it is necessary, I can use OpenOffice and copy & paste block of text.

Anyway, it would be a good idea to be able to add spell checker for different language than English and German.

It if of course up to David whether he allows foreign language spell checkers.

 Jara

The French translation provides a spell checker for Pegasus Mail.

We could create it thanks to David Harris tools but the most important problem is how to get the datas: you need 2 lists of words in the chosen language, one as complete as possible of the words used in that language and the second containing about 5000-7000 words which are the most used in the chosen language (this one is the most difficult to find or to create), it is not necessary for the second one to be included in the first one but if it is it doesn't hurt. I think that if you can send them to David he'll be able to create the dictionary, if he hasn't the time to do it the French team can help.

I forgot to say that the lists have to be free of use.

Regards

They can also use one of the various spell checkers like As-U-Type http://www.asutype.com/dictionaries.html?src=index&p=5 to make the corrections as you enter the text.  This is the one I use and there are a number of dictionaries.  Tiny Spell is another with a lot of dictionaries.  http://tinyspell.numerit.com/

As for the Czech dictionary you might ask Lukas Gebauer  <gebylist@MLP.CZ > what he uses for a spell checker in Pegasus Mail.  He might be like David though who does not ever use a spell checker. 

Some time before, I have created special program for harvesting internet and create such a list of words. So I could attempt to reuse this program and create a valid list of czech words anew.

This was a main reason why I has been asking for possibility to add Czech dictionary. I can even help with creating license free word list for other languages.

If I send Email to David Harris, I am afraid that he redirects all Emails to /dev/null.

There are two distinct ways of going about this. Neither of which work totally as needed.

1, Monitor keyboard strokes and report. TinySpell or As-U-Type do this.

2. Copy/Paste content to be checked, and use a third party spell checker such as ISpell or its derivatives

For 1. the problem is that the base dictionaries are 1998 or earlier and that omit several major changes in a number of languages. There are a number of places where wordlists can be downloaded from. Problem is that all derivative words, like for example putting an "s", as a plural suffix, have to be a separate entry. 

2. These dictionaries are up to at least 2005, and include general, technical, medical and other words. Problem here is that the application cannot be invoked inside Pegasus Mail as TER is used already.  Derivative words are held as alternative prologs or epilogs in a special encoding scheme that  are very labor intensive to maintain. See  http://www3.telus.net/public/irelam/man5.pdf

In summary, the dictionaries for ISpell are public domain, and could be added into Pegasus Mail's dictionary flat file. I think this would need to be done selectively for performance and capacity reasons.

Anyone game?

Martin