can't search with unicode regular expression

[feichao]

can't search with unicode regular expression

• VSCode Version: 1.18.1 (1.18.1)
• OS Version: macOS 10.12.6

Steps to Reproduce:

1. input a unicode regular expression, e.g. [\u4e00-\u9fa5]
2. some error occur
   

Reproduces without extensions: Yes

[roblourens]

I think you need to use [\x{4e00}-\x{9fa5}]. Does that work how you expect?

[feichao]

Thanks a lot. But it doesn't work... Actually I want to search all the Chinese in my project.

[adelphes]

You can't use unicode escape character ranges because VSCode search uses two incompatible regex engines - one search engine uses the JavaScript regular expression syntax (albeit without the 'u' unicode flag), the other is based on ripgrep which uses Rust regular expression syntax.

If you use the Rust syntax: [\x{4e00}-\x{9fa5}] the JS RegExp parses this as a range between } and x - this is an invalid range because x < {.
If you use the JS syntax [\u4e00-\u9fa5] ripGrep cannot recognise the \uXXXX escape.

Since JS RegExp doesn't support the \x{XXXX} syntax and Rust doesn't support \uXXXX syntax, the
end result is one of the engines will always error.

[roblourens]

Correct - I filed an issue upstream. vscode could also rewrite the \u patterns to \x for ripgrep.

[skyfore]

@roblourens Well, Is there any way to slove the problem ?

[roblourens]

There is activity on the upstream issue but it's not pulled into vscode yet.

[AaronWing]

oh no Chinese no way