Search for FuelPHP

Search is a FuelPHP package to search through data

• Jaap Rood
• Ronald Mansveld

Slightly advanced, easy to use, search

We find that implementing search into apps and websites can be quite a pain. MySQL doesn't support much, Google Site Search is far from ideal (and often not accepted by clients) and not everyone can run or has the time to work with Apache Lucence. We want search that is easy to use, pretty quick and supports spelling mistakes. To do this, we implemented a variation of the Damerau-Levenshtein algorithm to calculate the similarity of words.

Make no mistake, in the search world this stuff is pretty basic, and the solution is far from ideal. But we believe it's a pretty cool alternative for other solutions and has come quite handy a number of times already.

Usage

Searching through data is pretty easy. All you need is the data, the field(s) you want to look in and the term you want to search for.

$products = Model_Product::find('all');

$found_products = Search::find('shampoo')
	->in($products) // the data to search through
	->by('name', 'slogan') // the fields you wan't to search through
	->execute();

As data, an array with either objects or arrays is accepted. There are some other cool ways to tweak your results

$posts = Model_Post:find('all');

$found_posts = Search::find('foobar')
	->in($posts)
	->by('title', 'subtitle', 'slug')
	->relevance(80) // this means the matched words in the data must be about 80% the same
	->limit(10)	// let's limit the amount of results to 10
	->offset(5) // offset the results, handy for pagination
	->execute();

Known issues and stuff we're working on

This class does not work properly with multiple word search terms. We're aware of this and are working on a solution. Also, because it looks at every single word, it's not that efficient on big text fields, we recommend using it on stuff like titles, names, etc.

Update: we've made improvements to the way multiple word search terms are handled in the master branch. It's still not quite where we'd like it to be, but it's a lot better than before.

Next to improving results we've got heaps of ideas on stuff we could incorporate, like skipping common words, built in caching and searching in nested arrays / objects. We really want to know what you think would be useful! We will love you if you whip it all up and do a pull request :)

We will use Github's issue tracker to keep track of all issues, bugs, new ideas, etc.