Quote:
Originally Posted by dtanis
If you're successful, could you post the morphems file here?
|
Certainly. It will probably take a while though (a few weeks). There are a lot of grammatical morphemes in Polish and I'm no longer fluent in the language.
Quote:
It's no so difficult. It's just a table for transforming foreign language characters to the language alphabet in uppercase. It also transforms the punctuation marks to nothing (ignoring rule). The binary elements you talked about are just UTF-8 codes for characters which the font in your editor can't display because it doesn't have the correct glyphs. Use a full unicode font and you'll see that more characters will appear.
So for a Polish collate file, take the English version and add entries for the extra polish letters: ĘÓĄŃŚŁŻŹĆ (don't forget to delete these characters, upper and lower case, from the original entries). It's quite easy so I made a Polish one for you. It's attached to this message.
|
Thanks very much for the collates.txt file. I have no excuse not to continue now.
It looks like the first line of business for me is to get my editor working with a full unicode font. I can then start with the most common morphemes and keep adding more as I remember/discover them. The only thing I'm not sure of is what happens in the case of competing rules in the morphems.txt file. Is the first one given priority, or are all of them applied to give multiple answers? I suppose I'll see when I try out the dictionary on the device.