ncp - Asynchronous recursive file & directory copying

Build Status

Think cp -r, but pure node, and asynchronous. ncp can be used both as a CLI tool and programmatically.

Command Line usage

Usage is simple: ncp [source] [dest] [--limit=concurrency limit] [--filter=filter] --stopOnErr

The ‘filter’ is a Regular Expression - matched files will be copied.

The ‘concurrency limit’ is an integer that represents how many pending file system requests ncp has at a time.

‘stopOnErr’ is a boolean flag that will tell ncp to stop immediately if any errors arise, rather than attempting to continue while logging errors.

If there are no errors, ncp will output done. when complete. If there are errors, the error messages will be logged to stdout and to ./ncp-debug.log, and the copy operation will attempt to continue.

Programmatic usage

Programmatic usage of ncp is just as simple. The only argument to the completion callback is a possible error.

var ncp = require('ncp').ncp;

ncp.limit = 16;

ncp(source, destination, function (err) {
 if (err) {
   return console.error(err);
 }
 console.log('done!');
});

You can also call ncp like ncp(source, destination, options, callback). options should be a dictionary. Currently, such options are available:

  • options.filter - a RegExp instance, against which each file name is tested to determine whether to copy it or not, or a function taking single parameter: copied file name, returning true or false, determining whether to copy file or not.

  • options.transform - a function: function (read, write) { read.pipe(write) } used to apply streaming transforms while copying.

  • options.clobber - boolean=true. if set to false, ncp will not overwrite destination files that already exist.

Please open an issue if any bugs arise. As always, I accept (working) pull requests, and refunds are available at /dev/null.