[Request] Option to Exclude Latin/English characters from check(to check only non-ASCII languages)

[Heavygrass]

Hi!

It would be great to have an option to exclude all Latin/English characters/words from check, so that one could spell check directly non-ASCII languages/characters.

For example check only Cyrillic words/characters or Chinese etc.

I tried excluding every single dictionary installed and specifying a Cyrillic language, but Aspell still checks for Latin/English characters and reports them as misspelled (not a bug of course):

echo "helo keboard" | aspell --ignore-case --master=ru --lang=ru list --rem-extra-dicts=en --rem-extra-dicts=en-variant_0 --rem-extra-dicts=en-variant_2 --rem-extra-dicts=en-variant_1 --rem-extra-dicts=en-w_accents --rem-extra-dicts=en-wo_accents --rem-extra-dicts=en_AU --rem-extra-dicts=en_AU-variant_0 --rem-extra-dicts=en_AU-variant_1 --rem-extra-dicts=en_AU-w_accents en_AU-wo_accents --rem-extra-dicts=en_CA --rem-extra-dicts=en_CA-variant_0 --rem-extra-dicts=en_CA-variant_1 --rem-extra-dicts=en_CA-w_accents --rem-extra-dicts=en_CA-wo_accents --rem-extra-dicts=en_GB --rem-extra-dicts=en_GB-ise
Output:
helo

keboard

Please let me know if such an option already exists, I couldn't find it.
Thank you!

[kevina]

Sorry for they delay.

There is no option for this, but it is possible by modifying the language data file. The location of this file can normally be found by running aspell config data-dir. You should see a file ru.dat in that directory that contains the line:

charset koi8-r

change that to add -nl to the charset name, i.e.:

charset koi8-r-nl

and add these two files (koi8-r-nl.zip) to the same directry you found 'ru.dat' in.

If the charset is not koi8-r than you will need different files. You can create those files by downloading the aspell-lang package available at https://ftp.gnu.org/gnu/aspell/aspell-lang-20101122.tar.gz and then running ./pre <lang> <charset>-nl. The results will be in the directory <lang>. I can also create those files for you.

Latin characters are accepted as part of a word for mostly historical reasons. It can also help correct words where a similar looking Latin character is used instead of the Cyrillic one.

I agree though, that an option would be far better than having to hack the language data file so I will keep this issue open.