How to properly support Arabic in manticore?

Question

Sphinx

I'm trying to transfer the sphinx config to manticore.
The lines are in different languages, I want the most universal config.

What was in the Sphinx:

# chinese lang setting
ngram_len = 1
ngram_chars = U+3000..U+2FA1F
        
# ignore arabic chars
ignore_chars = U+0640, U+064B..U+065F,U+06D6..U+06DC,U+06DF..U+06E8,U+06EA..U+06ED

charset_table = /много-много кодов из разных языков/

Sphinxsearch.com/wiki/doku.php?id=charset_tables#arabic says :

Its necessary to add ignore_chars to ignore vowels, black space and other Arabic signs

In the manticore documentation, the recommended config is more compact:

ngram_len = 1
ngram_chars = cjk
charset_table = non_cjk

Question: Will the absence of ignore_chars break Arabic searches?

Answer 1

May break. If the quality of the search in Arabic texts is extremely important, then it is better to leave it as in the sphinx.