Filter option for actions.prompt

[passy]

I would like to propose a feature I found while digging around in grunt-init that did not have an equivalent in Yeoman.

The prompt should accept a filter option that allows to sanitize, validate or manipulate the value provided by the user.

Example Usage

var prompts = [{
  name: 'title',
  message: 'Project Title:',
  filter: function (value, cb) {
    // Callback expects (err, newValue)
    cb(null, value.replace(/jquery/i, 'jQuery'));
  }
}];

Benefit

While sanitizing data is already possible after the prompt is finished in the callback, this way makes it clearer to see where the manipulation is happening. Also, this allows to abort early if a value fails to validate.

[passy]

/ping @sindresorhus

[sindresorhus]

Should also have a pattern to quickly validate a prompt with regex.

But when thinking about this some more, why are we reinventing the wheel instead of using an existing prompt module? Like https://github.com/flatiron/prompt

[passy]

Yes by all means. I'm on my phone and not sure where I mentioned this, but
Addy and I were wondering if reusing an existing prompt library in our
kickoff anyway. The API for that one looks pretty good.
On Apr 6, 2013 2:42 PM, "Sindre Sorhus" [email protected] wrote:

Should also have a pattern to quickly validate a prompt with regex.

But when thinking about this some more, why are we reinventing the wheel
instead of using an existing prompt module? Like
https://github.com/flatiron/prompt

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/198#issuecomment-15995568
.