Search Engine

A Proof Of Concept demonstrating how to handle Interrogatory Messages (Query in CQRS).

The Command / Query Responsibility Segregation principle explains that Imperative and Interrogatory messages shouldn't be mixed together.

Note: Learn more about the different messaging flavours.

Usually imperative messages are handled using the CommandBus pattern, which leaves us with the following question: how Interrogatory Messages should be handled?

This component tries to explore one of the possible answers: a SearchEngine that would try to return results matching a given criteria.

Caution: this component does not provide actual SearchEngine features, if you're looking for one you should rather have a look at ElasticSearch, Solr, etc.

Installation

Download SearchEngine using Composer:

composer require gnugat/search-engine:^0.3

You'll also need to choose one of the following implementations:

• PommSearchEngine, a Pomm Foundation implementation
• hey you just met this library, and this is crazy, but it has interfaces, so implement them, maybe?

Other possible implementations: PDO, Doctrine DBAL, Doctrine ORM, etc. In the tests directory you'll find an Array implementation example.

More information about implementations

Usage

SearchEngine expects a Criteria object which describes:

• the resource to query
• relations to embed
• filters to apply
• orderings instructions
• pagination parameters

It can be built from query parameters as follow:

// ...
$criteria = $criteriaFactory->fromQueryParameters('blog', [
    // Filters
    'title' => 'IG',
    'author_ids' => '1,3',

    // Pagination
    'page' => '2',
    'per_page' => '3',

    // Ordering
    'sort' => 'author_id,-title',

    // Relation embeding
    'embed' => 'author',
]);
print_r(iterator_to_array(
    $searchEngine
        ->match($criteria)
        ->take(
            $criteria->paginating->offset,
            $criteria->paginating->itemsPerPage
        )
        ->getIterator()
));

In a web context, this $queryParameters array could actually be $_GET, corresponding to the following URL:

/v1/blogs?title=IG&author_ids=1,2&page=1&per_page=3&sort=author_id,-title&embed=author

The result could be the following:

[
    'items' => [
        [
            'id' => 1,
            'title' => 'Big Title',
            'author_id' => 1,
        ],
        [
            'id' => 2,
            'title' => 'Big Header',
            'author_id' => 2,
        ],
    ],
    'page' => [
        'current_page' => 1,
        'per_page' => 3,
        'total_elements' => 2,
        'total_pages' => 1,
    ],
]

Further documentation

You can see the current and past versions using one of the following:

• the git tag command
• the releases page on Github
• the file listing the changes between versions

You can find more documentation at the following links:

• copyright and MIT license
• versioning and branching models
• contribution instructions

Next readings:

• Implementing