You are browsing a version that is no longer maintained.
Simple Search Engine
It is very easy to implement a simple keyword search engine with MongoDB. Because of its flexible schema less nature we can store the keywords we want to search through directly on the document. MongoDB is capable of indexing the embedded documents so the results are fast and scalable.
Sample Model: Product
Imagine you had a
Product document and you want to search the products by keywords. You can
setup a document like the following with a
$keywords property that is mapped as a
Working with Keywords
Now, create a product and add some keywords:
1 new Product(); $product->setTitle('Nike Air Jordan 2011'); $product->addKeyword('nike shoes'); $product->addKeyword('jordan shoes'); $product->addKeyword('air jordan'); $product->addKeyword('shoes'); $product->addKeyword('2011'); $dm->persist($product); $dm->flush();$product = 2 3 4 5 6 7 8 9 10 11 12
The above example populates the keywords manually but you could very easily write some code which automatically generates your keywords from a string built by the Product that may include the title, description and other fields. You could also use a tool like the AlchemyAPI if you want to do some more intelligent keyword extraction.
Searching the keywords in the
Product collection is easy! You can run a query like the following
to find documents that have at least one of the keywords:
You can make the query more strict by using the
all() method instead of
The above query would only return products that have both of the keywords!
You can easily build keywords from a user search form by exploding whitespace and passing the results to your query. Here is an example:
If you want to use an embedded document instead of just an array then you can. It will allow you to store additional information with each keyword, like its weight.
You can setup a
Keyword document like the following:
Now you can embed the
Keyword document many times in the
With the new embedded document to add a keyword to a
Product the API is a little different,
you would have to do the following:
This is a very basic search engine example and can work for many small and simple applications. If you need better searching functionality you can look at integrating something like Solr in your project.