glob-parse

Returns a parsed representation of a glob string; does not require Minimatch.

Features

• Works on any string, does not require Minimatch or any other separate glob library.
• Does not perform glob matching: it just parses a glob expression into segments and produces relevant metadata about those segments.
• Pure parsing/tokenization is useful for working with glob expressions. For example:

wildglob uses glob-parse to parse the different segments of the input glob and then combines the string segments to determine where to start glob matching.

glob2base extracts a base path from a glob. It uses Minimatch to do this, but glob-parse (the .basename() function) can also be used to extract the base path from a glob.

API and examples

Basic parsing

1var parse = require('glob-parse');
2console.log(parse('js/*.js'));
3// [ 'js/', '*', '.js' ]
4console.log(parse('js/**/test/*.js'));
5// [ 'js/', '**', '/test/', '*', '.js' ]

.basename()

basename() works like glob2base:

1console.log(parse.basename('js/test{0..9}/*.js'));
2// js/
3console.log(parse.basename('js/t+(wo|est)/*.js'));
4// js/
5console.log(parse.basename('lib/{components,pages}/**/{test,another}/*.txt'));
6// lib/

Full type annotations

Pass { full: true } to return the token type annotations.

1console.log(parse('js/t[a-z]st/*.js', { full: true }));
2// { parts: [ 'js/t', '[a-z]', 'st/', '*', '.js' ],
3//   types: [ 'str', 'set', 'str', '*', 'str' ] }
4
5console.log(parse('js/{src,test}/*.js', { full: true }));
6// { parts: [ 'js/', '{src,test}', '/', '*', '.js' ],
7//   types: [ 'str', 'brace', 'str', '*', 'str' ] }
8
9console.log(parse('test/+(a|b|c)/a{/,bc*}/**', { full: true }));
10// { parts: [ 'test/', '+(a|b|c)', '/a', '{/,bc*}', '/', '**' ],
11//   types: [ 'str', 'ext', 'str', 'brace', 'str', '**' ] }