API Reference

The ansiRegex() function generates a regular expression that matches ANSI escape codes in strings. Use it when you need to detect, strip, or otherwise process ANSI escape sequences — such as color codes or cursor movement commands — from terminal output or log data.

Returns a RegExp object compiled with the global flag (g). The pattern matches all ANSI escape sequences in a target string, including CSI sequences (e.g., color and formatting codes beginning with ESC [ or \u009B) and OSC sequences terminated by BEL (\u0007). Because the regex is created with the g flag, you can use it with methods like String.prototype.match(), String.prototype.replace(), or RegExp.prototype.exec() in a loop to find all occurrences in a given string.

Important: A new RegExp instance is returned each time you call ansiRegex(). Reuse the returned instance within a given operation, but call ansiRegex() again if you need a fresh regex with its lastIndex reset to 0.

Under normal usage ansiRegex() does not throw. No arguments are accepted, so there is no risk of type-related errors from unexpected input.

Check whether a string contains any ANSI escape codes

const ansiRegex = require('ansi-regex');

const str = '\u001B[4mHello, World!\u001B[0m';
console.log(ansiRegex().test(str));
// => true

Extract all ANSI escape sequences from a string

const ansiRegex = require('ansi-regex');

const str = '\u001B[31mError:\u001B[0m Something went wrong';
const matches = str.match(ansiRegex());
console.log(matches);
// => [ '\u001B[31m', '\u001B[0m' ]

Strip ANSI escape codes from a string

const ansiRegex = require('ansi-regex');

const raw = '\u001B[32mSuccess\u001B[0m: Operation complete';
const clean = raw.replace(ansiRegex(), '');
console.log(clean);
// => 'Success: Operation complete'

Iterate over all matches using exec()

const ansiRegex = require('ansi-regex');

const str = '\u001B[1mBold\u001B[0m and \u001B[4mUnderline\u001B[0m';
const regex = ansiRegex();
let match;

while ((match = regex.exec(str)) !== null) {
  console.log(`Found: ${JSON.stringify(match[0])} at index ${match.index}`);
}
// => Found: "\u001B[1m" at index 0
// => Found: "\u001B[0m" at index 8
// => Found: "\u001B[4m" at index 14
// => Found: "\u001B[0m" at index 26

• New instance per call: Every invocation of ansiRegex() constructs and returns a brand-new RegExp object. If you plan to use the regex multiple times in the same logical operation, store the result in a variable rather than calling ansiRegex() repeatedly.
• Global flag and lastIndex: Because the returned regex uses the g flag, its lastIndex property advances with each exec() call on the same instance. If you reuse the same instance across unrelated strings without resetting lastIndex to 0, you may get unexpected results. Calling ansiRegex() again is the safest way to obtain a clean instance.
• Covered sequence types: The regex is designed to handle the two most common families of ANSI escape sequences: C1 control sequences (CSI, started by ESC [ or the 8-bit equivalent \u009B) and OSC sequences terminated by the BEL character (\u0007). Less common or proprietary terminal extensions may not be matched.
• CommonJS module: The source uses module.exports, so you import the function with require('ansi-regex'). If you are working in an ESM context, use a dynamic import() or a bundler that handles CommonJS interop.