Rule Change: no-control-regex should check Unicode code point escapes

[mdjermanovic]

What rule do you want to change?

no-control-regex

What change to do you want to make?

Generate more warnings

How do you think the change should be implemented?

A new default behavior

Example code

/*eslint no-control-regex: "error"*/

/\u001F/;

/\u{1F}/u;

What does the rule currently do for this code?

Reports only /\u001F/.

What will the rule do after it's changed?

Report both /\u001F/ and /\u{1F}/u.

Participation

• I am willing to submit a pull request to implement this change.

Additional comments

I think not reporting /\u{1F}/u is actually a bug caused by forgetting to pass the u flag to regexpp, so \u{1F} is not parsed as an escape sequence. The bug is observable in (rare) patterns that are valid with the u flag but invalid without it. For example, the rule fails to report \x1F in /\u{1111}*\x1F/u because regexpp stops parsing the pattern after \u{1111}* since it is invalid without the u flag.

Here's a demo with examples: Online Demo

[nzakas]

Makes sense to me.

[nzakas]

@btmills ?

[btmills]

Makes sense to me as well as a semver-minor bug fix 👍