Add --runs option to phantomas

[macbre]

• merge run-multiple.js with main phantomas "binary"
• add --runs option

Waiting for #152

[rupl]

When this is integrated, it would be awesome if the multi-run can support all of the arguments that a single run accepts. I have been making custom modifications to the run-multiple.js file, for instance to pass user-agent and viewport args into the multi-run. I just followed suit and added each one individually. I'm not sure if there's a more elegant way. Example:

if (params.viewport !== 'undefined') {
        cmd.push(' --viewport=' + params.viewport);
}
if (params['user-agent'] !== 'undefined') {
        cmd.push(' --user-agent=\'' + params['user-agent'] + '\'');
}

[stefanjudis]

@macbre any chance to get that in v1.0? I need to do major changes on my side with v1.0 already. Would be nice to not have more major changes at grunt-phantomas in v1.1.

If not, no problem. ;)

[macbre]

@stefanjudis, I keep adding more and more features to v1.0 :) Merging run-multiple helper into main "binary" actually fits into v1.0 milestone.

[stefanjudis]

Perfect!!! Thank you. 😄

[macbre]

Actually, here comes the question: how to provide the results of each run? Currently npm module is provided with a JSON with single run results. How should multiple results be provided / aggregated?

[stefanjudis]

Hmm.
I would prefer drop single run results completely. So that output is in all cases an object for given metric.

metric : {
  average : ...,
  median : ...
  ...
}

That's a breaking change for the current implementation and its dependants, but well it's a major release. It's allowed. ;)

[macbre]

Use:

• https://www.npmjs.org/package/ascii-table
• https://github.com/caolan/async#seriestasks-callback

[stefanjudis]

No problem. :)