Firefox Extension: Search Everywhere

Have you ever been at a web site and wanted to find something on it but couldn’t because the site either lacked search functionality or had a useless one? Did you want a tool to help in your predicament? Here’s one that I’ve found helpful: Search Everywhere by Paul Komarek. It was developed by a colleague of mine, and as a search junky, I got to be one of the beta testers. It’s now officially on the Firefox Add-ons page.

The extension was designed “to search the outgoing links from any webpage”. A use case for that is if you’re on a particular site and wanted to learn more about a particular topic, you would use this extension to search the keyword on Google and it would look up the keyword through the site and other sites the current site linked to, on the premise that the current site would link to relevant/related resources.

When I started using this, there was no way to specifically search for results in the current site, so I requested it and Paul implemented it in the latest version. This one little addition has made the feature much more useful to me. I could have manually done a [site:domain keyword] query but that would have taken more effort.

Of course, the extension can only provide results as long as Google has indexed the site. Also, a heads-up that it might take a few seconds to get results when doing a search involving linked sites… probably due to the Custom Search Engine having to go fetch and process the linked sites. A SiteSearch is as quick as any Google Search so that would seem to confirm that theory.

One thought on “Firefox Extension: Search Everywhere”

  1. Hi Thu, you are correct about the slowness for extracting links. If you search again from that same page, the extracted links will have been cached.

    In short, here’s what happens when you do *not* use site search with my extension:
    1) The extension grabs the url you are viewing and your search query, and sends them to Google Custom Search Engine servers. We’re basically constructing a “linked cse” using the “makecse” convenience tool we provide. See http://google.com/coop/docs/cse/cref.html for docs.

    2) We (custom search) get the request, and run the makecse script on the url you were viewing, with the options your toolbar set. This requires Google to fetch a current-ish version of the url, then we parse it and build a linked cse xml definition.

    3) The custom search infrastructure uses this definition, along with your query, to compute and render the search results you see in your browser.

    The slow parts of this (fetching the url and parsing it, for instance) are the parts we try to cache.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>