tallnerds/howfast

A simple CLI for generating more accurate data from Lighthouse reports.

howfast

A simple CLI for generating more accurate data from Lighthouse reports.

Installation

The easiest way to install is to globally install via npm.

npm install -g @tallnerds/howfast

Usage

To see up to date usage, you can always run

howfast -h

Here is a typical example of hitting a site 5 times

howfast -s https://google.com -t 5

Here is an example of hitting a site that requires an Authorization header

howfast -s https://google.com -a 'Basic aG93ZmFzdDpyb2Nrcw=='

By default, this will also generate a json file with all of the metrics you would need to do more detailed analysis. Upon completing the runs, it will open this file automatically for you. By default, this file is stored in /tmp, but can be modified with the -f flag.

howfast -s https://google.com -t 5 -f ~/Downloads

Lastly, if you are creating a PR and want to quickly generate some markup, you can change the report type to html using the --html flag. This will give you a generated html file (rather than json) with the results displayed in different ways (json, markdown, etc).

howfast -s https://google.com -t 5 -f ~/Downloads --html

Reports

There are 2 types of reports that can be generated from this tool: json and html.

Json

The default report output is json. Here is the generate structure of the json file:

{
  "metricName": {
    "label": "<[string] label>",
    "scorePath": "<[string] path to score in original lighthouse data>",
    "metricPath": "<[string] [optional] path to metric in original lighthouse data>",
    "average": {
      "score": "[number]",
      "numericValue": "[number]"
    },
    "median": {
      "score": "[number]",
      "numericValue": "[number]"
    },
    "actual": {
      "scores": "[array<number>]",
      "numericValues": "[array<number>]"
    }
  }
}

Here is a nice visual representation

Contributing

Re-render Usage Gif

We use a really neat-o tool for generating the console gif above. To re-render, simply re-record

terminalizer record terminal-demo

tweak the settings at ./terminal-demo.yml

and record

terminalizer render terminal-demo