Filtering non English character emails

You should be able to do this with a regex.

I don't normally post non-Sophos links, but I found this resource which explains Unicode and Regex and the syntax differences between different Regex libraries.

https://www.regular-expressions.info/refflavors.html

I do not know what RegEx library is used on UTM.   Mr Alfson or Mr. Jaydeep, can you answer that?

Recently I tried this Regex in one of my three spam filters.  I think this particular syntax works all regex libraries..

[\p{S}p{C}P{Latin}]

That is supposed to say:

• Any symbol (heart, smiley, etc.)
• Any control character.
• Any non-Latin script

We actually disabled the rule because it was matching too much.   In retrospect, I think the issue was that I was checking both Body and Subject, and the body has line feeds, so the control character match is not appropriate for the Body check.

For your purposes, [P{Latin}] should catch Russian and Chinese text (and Korean and Katakana, etc.)

You should be able to do this with a regex.

I don't normally post non-Sophos links, but I found this resource which explains Unicode and Regex and the syntax differences between different Regex libraries.

https://www.regular-expressions.info/refflavors.html

I do not know what RegEx library is used on UTM.   Mr Alfson or Mr. Jaydeep, can you answer that?

Recently I tried this Regex in one of my three spam filters.  I think this particular syntax works all regex libraries..

[\p{S}p{C}P{Latin}]

That is supposed to say:

• Any symbol (heart, smiley, etc.)
• Any control character.
• Any non-Latin script

We actually disabled the rule because it was matching too much.   In retrospect, I think the issue was that I was checking both Body and Subject, and the body has line feeds, so the control character match is not appropriate for the Body check.

For your purposes, [P{Latin}] should catch Russian and Chinese text (and Korean and Katakana, etc.)