TLD block: Punicode handling + complete list of Chinese, Iranian, and Russian domains

[nihil-admirari]

I want to (block on the server|directly connect the client to) all Chinese, Iranian, and Russian domains.

Here are the lists of top-level domains:

• https://en.wikipedia.org/wiki/List_of_Internet_top-level_domains
• https://en.wikipedia.org/wiki/Country_code_top_level_domain

1. How internationalised TLDs are supposed to be handled: directly

   {
     "domain": ["regexp:.*\\.(中国|ایران|рф)$"]
   }

   by Punicode

   {
     "domain": ["regexp:.*\\.(xn--fiqs8s|xn--mgba3a4f16a|xn--p1ai)$"]
   }

   or both?

2. The TLD list in the references is enormous:

   • Chinese domains in Latin: .ren, .shouji, etc.
   • Chinese domains in Chinese: .在线, .中文网, etc.
   • Russian domains in Cyrillic: .дети, .католик, etc.
   • Russian domains in Latin: .moscow, .tatar, etc.
   • Chinese corporate domains: .alibaba, .baidu etc.
   • American corporate domains in Chinese: .亚马逊, .谷歌 etc.

   Does anyone know a routing rule for all of the above? (I'd rather avoid wading through the lists myself, sigh.)

[nihil-admirari]

Answering my own question:

1. https://github.com/v2fly/domain-list-community uses Punicode exclusively.
2. Some TLDs are in geosite:cn, geosite:category-ir and geosite:category-ru, though some TLDs listed on Wikipedia are missing. .ren for example is not present in geosite:cn despite https://www.iana.org/domains/root/db/ren.html clearly stating that its sponsoring organisation is in Beijing. Probably geosite:tld-cn needs some updating.