Word splitting is not taking CJK into account #1506
Description
Please complete the following tasks
Valid word
HashiCorp
Incorrect correction
HashCorp
Justification
Notes
When Japanese (CJK) characters are adjacent to ASCII characters without whitespace (e.g., 含むHashiCorp製品),
typos treats the entire mixed-script string as a single identifier rather than splitting at the script boundary.
As a result, extend-identifiers entries like HashiCorp = "HashiCorp" do not match because the actual identifier is the longer mixed-script string.
The identifier is then split into subwords,
and Hashi is flagged as a typo for Hash,
which extend-words for HashiCorp also cannot suppress since it only matches at the subword level against the full string HashiCorp.
The expected behavior would be for typos to treat script boundaries (e.g., CJK to Latin) as identifier boundaries,
so that HashiCorp in 含むHashiCorp製品 is recognized as a standalone identifier and correctly matched against extend-identifiers entries.