eliot-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Eliot-dev] Catalan language file and dictionary


From: Olivier Teuliere
Subject: Re: [Eliot-dev] Catalan language file and dictionary
Date: Sun, 21 Jun 2009 15:06:50 +0200

Hi Joan,

On Sun, Jun 14, 2009 at 10:04 PM, Joan Montané<address@hidden> wrote:
> In [1], you have an utf8 Catalan wordlist file. It's not oficial, but
> it's the best Catalan wordlist publicy avaiable. Note: this file has
> words with Ç,L·L,NY and QU.

Thanks, it is now saved in the repository (direct link: [1]).
I have also added the compiled dictionary for download on the website
(note: I compressed it without the substitution for Ç).

> About input methods, I think the best aproach for "special chars" is
> allow two input strings, if needed: ascii and unicode. ascii single
> char ensures an easy input (on foreing keyboard). Unicose char (or
> string) allow a natural input for users with the good keyboard and
> it's the information for the good output.
>
> For example, in Catalan case:
> K  10 1 0 1 Ç
> W 10  1 0 1 L·L
> Y  10  1 0 1 NY
> Q  8  1 0 1 QU

Actually, the dictionary compression algorithm requires a single char
for each tile, so this will definitely stay. I don't know if
restricting the char to ASCII characters is so useful: if users really
can't type the chars, they have several solutions (copy-paste,
changing the keyboard layout, ... or even compiling their own version
of the Eliot dictionary).

To be really complete, I think we also need another type of
information: the display char (or string, rather), shown on the board.
So here is the data we would needed for each tile:
1) Single char, as used in the compressed dictionary. Only used at
compression time
2) Points of the tile
3) Frequency of the tile
4) Display string, i.e. what to display on the tiles. It could be an
optional field (if not specified, we would use field 1)
5) List of input strings, each of which could be accepted as user
input. This could be optional as well, the default value being the
value of field 4.

Example for Catalan:
Ç  10 1 0 1
W 10  1 0 1 L·L [L·L L.L L-L ĿL W]
Y  10  1 0 1 NY [NY Y]
Q  8  1 0 1 QU [QU Q]

It looks like this solves all the issues, tell me if I missed something.
I plan to implement these changes, but I might be quite busy during
the next 2 months, so... don't hold your breath :-)

Regards,
-- 
Olivier
[1] 
http://cvs.savannah.gnu.org/viewvc/*checkout*/dic/danosc/danosc.txt?root=eliot




reply via email to

[Prev in Thread] Current Thread [Next in Thread]