glob
Source: lib/glob.js (pure JS, runs under node, bun and deno too) — exports Glob (also the default) and globToRegExp
Glob patterns with the rules of Bun.Glob, which is this class. match() compares strings only; dot files and
the file system matter only to scanSync() and scan(). The matching rules were checked against bun 1.4.2 on 21
patterns times 27 paths.
import { Glob } from 'glob';
new Glob('src/**/*.ts').match('src/a/b.ts'); // true
new Glob('!*.js').match('a.ts'); // true
[...new Glob('**/*.js').scanSync({ cwd: 'lib' })]; // relative paths, '/' separated
for await (const p of new Glob('*.md').scan('.')) {}| Pattern | Matches |
|---|---|
* | any characters except / |
? | one character except / |
[ab], [a-c], [!a], [^a] | one character of the set, or not in it |
{a,b} | either alternative, nests |
** | any number of whole path segments, also none: a/**/b matches a/b; src/** does not match src |
!p | at the start: everything p does not match |
\x | x literally |
scanSync(options | cwd) and scan(options | cwd) (an async iterator) take:
| Option | Default | Meaning |
|---|---|---|
cwd | '.' | where to start |
dot | false | also names starting with a dot (on by itself when the pattern has a segment starting with one) |
absolute | false | absolute paths instead of relative ones |
onlyFiles | true | leave directories out |
followSymlinks | false | walk into symlinked directories |
globToRegExp(pattern) returns { re, negated }, the anchored RegExp and whether the pattern started with !.
Export list
Generated by utilities/doc-api.js from the glob dump of utilities/api-surface.sh: 11 members. Edit the text above this block, not the block.
Functions: globToRegExp()
Glob
class, constructible
- methods:
match(),scan(),scanSync()
default
class, constructible
- methods:
match(),scan(),scanSync()