Arabic: Support Harakat

[Jason3S]

As mentioned the link to improve the dictionary in this link

we have a problem in Harakat in Arabic. They are considered complete letters. How we can ignore them?

Originally posted by @WatheqAlshowaiter in #941 (comment)

[Jason3S]

@WatheqAlshowaiter,

Great example. Would you please add (in text) some example word pairs?

It is possible words with Harakat are not in the dictionary, you might need to follow up with @linuxscout to see if they are there. I'll regenerate the dictionary to a greater depth see if those words are added. But he might have a better idea on how Harakat is handled in other spell checkers.

[linuxscout]

ٍHi @Jason3S,
To resolve this issue, we must remove diacritics before spellchecking,
we add remove_diacritics(word).

The word to compare against the dictionary is a non diacritized word (word without diacritics).
This case is handled by Hunspell like this, we added a feature name "IGNORE" to ignore a list of diacritics in the word. In the code we remove diacritics before spelling, but we keep it on display.

[linuxscout]

@Jason3S No need to rebuild the dictionary

[Jason3S]

@linuxscout,

Thank you.

Found it:

I'll see if I can add support for IGNORE to cspell.

[Jason3S]

I have added support to CSpell to ignore accents if specified in the dictionary definition.

I'll update the VS Code extension tomorrow.