regexUnnecessaryAssertions

Reports assertions in regular expressions that always reject.

✅ This rule is included in the ts logical presets.

Reports assertions in regular expressions that always reject. These patterns will never match because the assertion contradicts its surrounding context.

Examples

Word Boundary Between Same Character Types

A word boundary (\b) always rejects when placed between two word characters or two non-word characters.

❌ Incorrect

const pattern = /a\bb/;
const pattern2 = /-\b-/;

✅ Correct

const pattern = /a\b-/;
const pattern2 = /-\ba/;

Negated Word Boundary at Transition

A negated word boundary (\B) always rejects when there is a word/non-word transition.

❌ Incorrect

const pattern = /a\B-/;

✅ Correct

const pattern = /a\Ba/;

Anchors Not at Pattern Boundaries

Start (^) and end ($) anchors always reject when not at pattern boundaries (without the multiline flag).

❌ Incorrect

const pattern = /a^b/;
const pattern2 = /a$b/;

✅ Correct

const pattern = /^ab/;
const pattern2 = /ab$/;
const pattern3 = /a^b/m;

RegExp Constructor

The rule also checks regex patterns in RegExp constructor calls.

❌ Incorrect

const pattern = new RegExp("a\\bb");

✅ Correct

const pattern = new RegExp("a\\b-");

Options

This rule is not configurable.

When Not To Use It

If you are working with dynamically constructed regex patterns where the simplified detection may cause false positives, you may disable this rule.

Further Reading

• MDN: Assertions
• MDN: Word boundary assertion

Equivalents in Other Linters

• ESLint: regexp/no-useless-assertions