psi-report

Crawls a website or get URLs from a sitemap.xml or a file, gets PageSpeed Insights data for each page, and exports an HTML report.

---

• Installation
• CLI usage
• Programmatic usage
• Crawler behavior
• Changelog
• License
• Credits

Installation

Install with npm:

$ npm install psi-report --global
# --global isn't required if you plan to use the node module 

CLI usage

$ psi-report [options] <url> <dest_path>

Options:

    -V, --version               output the version number
    --urls-from-sitemap [name]  Get the list of URLs from sitemap.xml (don't crawl)
    --urls-from-file [name]     Get the list of URLs from a file, one url per line (don't crawl)
    -h, --help                  output usage information

Example:

$ psi-report daringfireball.net/projects/markdown /Users/johan/Desktop/report.html

Programmatic usage

// Basic usage
 
var PSIReport = require('psi-report');
var psi_report = new PSIReport({baseurl: 'http://domain.org'}, onComplete);
reporter.start();
 
function onComplete(baseurl, data, html)
{
    console.log('Report for: ' + baseurl);
    console.log(data); // An array of pages with their PSI results
    console.log(html); // The HTML report (as a string)
}
 
// The "fetch_url" and "fetch_psi" events allow to monitor the crawling process
 
psi_report.on('fetch_url', onFetchURL);
function onFetchURL(error, url)
{
    console.log((error ? 'Error with URL: ' : 'Fetched URL: ') + url);
}
 
psi_report.on('fetch_psi', onFetchPSI);
function onFetchPSI(error, url, strategy)
{
    console.log((error ? 'Error with PSI for ' : 'PSI data (' + strategy + ') fetched for ') + url);
}

Crawler behavior

The base URL is used as a root when crawling the pages.

For instance, using the URL https://daringfireball.net/ will crawl the entire website.

However, https://daringfireball.net/projects/markdown/ will crawl only:

• https://daringfireball.net/projects/markdown/
• https://daringfireball.net/projects/markdown/basics
• https://daringfireball.net/projects/markdown/syntax
• https://daringfireball.net/projects/markdown/license
• And so on

This may be useful to crawl only one part of a website: everything starting with /en, for instance.

URLs from a sitemap.xml or a file

Instead of crawling the website, you can set the URL list with a sitemap.xml or a file.

• --urls-from-sitemap https://example.com/sitemap.xml
• --urls-from-file /path/to/urls.txt

Only the URLs inside this file will be processed.

Changelog

This project uses semver.

Version	Date	Notes
2.2.1	2018-01-19	Fix missing source files on NPM (@blaryjp)
2.2.0	2017-11-27	Prepend baseurl if not present, for each urls in file (@blaryjp)
2.1.0	2017-11-19	Add --urls-from-sitemap and --urls-from-file (@blaryjp)
2.0.0	2016-04-02	Deep module rewrite (New module API, updated CLI usage)
1.0.1	2016-01-15	Fix call on obsolete package
1.0.0	2015-12-01	Initial version

License

This project is released under the MIT License.

Credits

• async
• colors
• request
• crawler
• commander
• sitemapper