sspanak / tt9

A T9 keyboard for Android devices with a hardware keypad.
Apache License 2.0
257 stars 44 forks source link

Add support for Catalan. #594

Closed Roconx closed 3 months ago

Roconx commented 3 months ago

Mostly fixes #592.

Missing support for l·l, but I don't think it is that important.

A way to continue recommending characters when words have - or ' would be nice, would also fix the issue with l·l.

sspanak commented 3 months ago

@Roconx, I have re-added the large wordlist and cleaned it up as much as possible. After that I did a quick typing test by using 6 Wikipedia articles and it was mostly fine, but I am under the impression many words are missing. This is surprising, given that the dictionary now contains more than 900k words. Or maybe I picked too specialized articles, I don't know.

Please, try the latest commit from this branch and let me know if there is anything weird.

Also, please let me know which of the following contractions/combinations are possible in Catalan:

'en
'hi
'ho
'l
'ls
'm
'n
'ns
's
't
-d'
-de-
-el
-els
-em
-en
-en-
-ens
-hi
-ho
-i-
-in
-l
-la
-la-
-le-
-les
-li
-lo
-los
-m
-me
-n
-ne
-ne-
-nos
-out
-s
-se
-t
-te
-us
-vos
-vu

Some of them come from French words, but I wasn't sure if they are also used in Catalan, so I took a conservative approach and kept them.

If you don't have any comments, I'll just merge the branch as it is.

Roconx commented 3 months ago

Thank you very much for doing this! I wanted to do it but have been very busy lately. I'll take a look tomorrow and I'll let you know!

Roconx commented 3 months ago

Wow, you were pretty much spot on! I managed to find a pretty detailed source here and I think that this is the final list (feel free to check in case I missed any):

'hi
'ho
l'
'ls
'ls-
'm
'n
n'
'ns
'ns-
's
-t'
d'
m'
t'
s'
-el
-els
-em
-en
-ens
-hi
-ho
-l
-la
-les
-li
-li'
-li-
-lo
-los
-los-
-m
-me
me'
-me'
-me-
-nos
-nos-
-se
-se'
-t
-te
-te'
-te-
-us
-us-
-vos
-vos-
-m'
sspanak commented 3 months ago

Alright, I think we sorted everything out. I am merging this PR and will include Catalan in the next version. If you notice any misspelled words, feel free to open an issue or a PR to correct them.