Jekyll + indextank = Static websites with powerful full-text search

How to use:

• Place jekyll_indextank.rb inside the _plugins folder in the root of your jekyll site.
• Add 'indextank_index', 'indextank_api_url' and optionally 'indextank_excludes' settings to your _config.yml

All pages' main content will be indexed at indextank, now all you have to do is hook up some Javascript and ajax to a form to query the indextank api.
Note that, while at the time of writing, the indextank api docs do not show that the search api supports jsonp, it does. Just add &callback=mycallback to the api call.

How it works

Everytime you generate your Jekyll site, all pages will be indexed at IndexTank. Now when you query the IndexTank search api, something like this will be returned:

{"matches": 1, "facets": {}, "search_time": "0.002", "results": [{"docid": "/blog/2009/01/04/colored-uiprogressview.html"}]}

The "docid" field contains the absolute url (minus domain name) to the matched resource.

JavaScript usage example

$.getJSON("http://<your_api_key>.api.indextank.com/v1/indexes/<your_index>/search?q=<a query>&callback=?", function(json) {
	alert(json.results); //you might want to approve usability a bit ;-)
});

See examples/example.html for a basic example, showing results in a div.

For a live example see the search on my blog http://pwiddershoven.nl

Requirements

Jekyll_indextank requires the following gems:

• indextank
• nokogiri

For an example of available config settings, see example_config.yml.

Installation

• gem install indextank nokogiri
• rake build
• copy build/jekyll_indextank.rb to your hyde site's _plugins folder