Help me on rules for i18n hangman game

[Frantisek Fuka]

I'd just like to draw your attention to special case in Czech language 
(I don't know if this also applies to other languages): Not only are 
there various letters with various accents but there is also a letter 
"ch", which looks and is written exactly same as "c" and "h" next to 
each other (Czech language also has normal "c" and "h" letters) but it's 
single letter in out alphabet! For example the Czech word "buchta" has 5 
letters (B, U, CH, T and A) and the letter "CH" comes after "H" and 
before "I" in the Czech alphabet. But there can also be Czech words that 
contain "C" and "H" next to each other, as two letters!

Christos TZOTZIOY Georgiou wrote:
> Hello people,
> 
> I have created a hangman game using Python and Tkinter, where the player
> can compete against the program (the program can try and guess a word
> based on knowledge of other words in the language).
> 
> The game is able to learn any language with 8-bit encodings (this is by
> unfortunate initial design, since I started off by recoding a much older
> C text-only program), and the UI text can be customised too, all that
> from a UTF-8 encoded config.ini file.  So far, I have three languages
> (and no cute graphics yet, but the logic and the UI are completely
> functional): greek, english and swedish (with help from a friend).
> 
> The problem I need you to help me with, is the following: in all three
> of the languages so far, any letter of the alphabet is considered
> distinct from all others for a game of hangman with pen and paper,
> counting accented letters too.
> 
> I can find the alphabets of as many languages I want on the Web, but
> *you* know how you played the game when you were younger, esp. in
> relation to accented characters; for example, in Swedish, "latin small
> letter a with ring above" is considered distinct from "latin small
> letter a" (and thus my program needs not do anything to cope for letter
> equivalences).  But playing hangman in French, say, when the player
> guesses 'e', does the program have to show all accented 'e' letters too
> (with grave, acute or circumflex)?
> 
> Any help would be welcome.  If you want to try the program (which will
> be released under LGPL when done, probably after 2.3 final since I use a
> lot of the new features), I can send a URL, but I consider it alpha at
> the moment.


-- 
                                          Frantisek Fuka
(yes, that IS my real name)
(and it's pronounced "Fran-tjee-shek Foo-kah")
----------------------------------------------------
My E-mail:   fuka at fuxoft.cz
My Homepage: http://www.fuxoft.cz
My ICQ:      2745855