Glob to Regex

Common errors

• Error: Cannot find module 'glob-to-regex'

  cause Attempting to import the package using the old name without the '.js' suffix.
  fix Change the import path to 'glob-to-regex.js'.

• TypeError: Invalid regular expression: /.../: Invalid character class

  cause The glob pattern contains a malformed or unescaped character class that JavaScript's `RegExp` engine cannot parse.
  fix Ensure character classes in glob patterns (e.g., `[a-z]`, `[!0-9]`) are syntactically correct and don't conflict with JavaScript regex rules. The library copies them directly.

• My glob pattern `*.js` isn't matching `src/file.js`.

  cause The `toRegex` function produces anchored regular expressions (`^...$`), meaning `*.js` only matches files directly in the root, not in subdirectories.
  fix Use the `**` pattern for recursive matching, e.g., `**/*.js` to match files in any subdirectory.

Warnings

• gotcha The package name was changed in v1.0.1 from 'glob-to-regex' to 'glob-to-regex.js'. Ensure your import paths reflect the `.js` suffix, especially for ESM environments, to avoid module resolution errors.
  Fix: Update import statements: `import { ... } from 'glob-to-regex.js';` or `require('glob-to-regex.js')`.
• gotcha The `toMatcher` function does not specially parse 'negative patterns' (e.g., `!**/*.d.ts`). Such patterns are treated as standard glob patterns, meaning `toMatcher(['!foo.ts'])` will match 'foo.ts' if no other positive pattern prevents it.
  Fix: For negative pattern logic, combine `toMatcher` with your own conditional logic or use `toRegex` with custom RegExp for exclusions.
• gotcha Generated regular expressions are always implicitly anchored at the start and end (`^...$`). This means `toRegex('*.js')` will match `file.js` but not `path/to/file.js`.
  Fix: Design glob patterns to match full paths (e.g., `**/*.js`) or wrap the generated regex in your own pattern if partial matching is desired, though this is generally not the library's intended use case.
• gotcha Brace groups (e.g., `{a,b,c}`) are not nestable. If nested braces are encountered, they will likely be treated as literal characters or result in unexpected regex syntax.
  Fix: Avoid nesting brace groups in your glob patterns. Flatten complex alternations or use multiple `toRegex` calls if necessary.
• gotcha Support for extended globbing features (like `?(pattern-list)`, `*(pattern-list)`) was added in v1.2.0 but requires the `extglob: true` option to be explicitly set. They are not enabled by default.
  Fix: Pass `{ extglob: true }` as the second argument to `toRegex` or `toMatcher` to enable extended globbing.

Install

• npm install glob-to-regex.js npm
• yarn add glob-to-regex.js yarn
• pnpm add glob-to-regex.js pnpm

Imports

• toRegex
  wrong:

  import { toRegex } from 'glob-to-regex';
  const toRegex = require('glob-to-regex.js').toRegex;

  correct:

  import { toRegex } from 'glob-to-regex.js';

  The package name in import paths requires the '.js' suffix for ESM. CommonJS usage needs `.toRegex` property access.
• toMatcher
  wrong:

  import { toMatcher } from 'glob-to-regex';
  const toMatcher = require('glob-to-regex.js').toMatcher;

  correct:

  import { toMatcher } from 'glob-to-regex.js';

  The package name in import paths requires the '.js' suffix for ESM. CommonJS usage needs `.toMatcher` property access.
• glob-to-regex.js (module)
  wrong:

  const GlobToRegex = require('glob-to-regex');

  correct:

  import * as GlobToRegex from 'glob-to-regex.js';

  When importing the entire module, remember the '.js' suffix for ESM. CommonJS users must use the full package name.

Quickstart

This example demonstrates converting simple and complex glob patterns to regular expressions, using the matcher utility, and highlights key behaviors like extended glob support and implicit anchoring.

import {toRegex, toMatcher} from 'glob-to-regex.js';

// Build a RegExp from a glob
const re = toRegex('src/**/test.ts');
console.log(`'src/a/b/test.ts' matches 'src/**/test.ts': ${re.test('src/a/b/test.ts')}`); // true
console.log(`'src/test.ts' matches 'src/**/test.ts': ${re.test('src/test.ts')}`);     // true
console.log(`'src/test.tsx' matches 'src/**/test.ts': ${re.test('src/test.tsx')}`);    // false

// Build a predicate function from a pattern or an array of patterns
const match = toMatcher(['**/*.ts', '!**/*.d.ts']); // negative patterns are not special; use a RegExp if needed
console.log(`'index.ts' matches ['**/*.ts', '!**/*.d.ts']: ${match('index.ts')}`);    // true
console.log(`'types.d.ts' matches ['**/*.ts', '!**/*.d.ts']: ${match('types.d.ts')}`);  // true (negation is not parsed specially)

// Demonstrate extended globs (requires option)
const reExt = toRegex('file.@(jpg|png|gif)', {extglob: true});
console.log(`'file.jpg' matches 'file.@(jpg|png|gif)' with extglob: ${reExt.test('file.jpg')}`); // true
console.log(`'file.txt' matches 'file.@(jpg|png|gif)' with extglob: ${reExt.test('file.txt')}`); // false

// Demonstrate anchored regex behavior
const reAnchored = toRegex('*.js');
console.log(`'my-file.js' matches '*.js': ${reAnchored.test('my-file.js')}`); // true
console.log(`'path/to/my-file.js' matches '*.js': ${reAnchored.test('path/to/my-file.js')}`); // false (due to implicit anchoring)