README

This package creates unified API for some NoSQL databases (at the time only Elasticsearch, MongoDB and Cassandra need to be finished).

Installation

The recommended way is via Composer:

composer require hovjacky/nosql

It requires PHP version 7.0 and higher.

Usage

Connection parameters (host, port, index) have to be passed as an array to the constructor of ElasticsearchClient.

The following methods are supported:

• insert - array $data
• bulkInsert - array $data
• get - int $id
• update - int $id, array $data
• delete - int $id
• deleteAll
• findBy - array $params - details described below.

All have table as first parameter (in Elasticsearch it is the type name).

####Method findBy Method findBy can be used for search. It supports quite a wide range of parameters (you can find comparison with SQL in parentheses):

• fields - array of fields user wants to retrieve (SELECT id, name FROM)
• limit - max. number of results (LIMIT 10)
• offset - can only be used with limit (LIMIT 10, 10)
• count - only get number of results
• orderBy - array of fields to order by, e.g. ['id', 'name'], descendant order is marked by lowercase desc - ['id desc', 'name'] (ORDER BY id DESC, name)
• groupBy - array of fields to group by, e.g. ['id', 'name'] (GROUP BY id, name)
  • Elasticsearch only:
    • groupByScript - instead of a field name, groupBy can contain the word script and groupByScript can then contain an elasticsearch script (https://www.elastic.co/guide/en/elasticsearch/reference/current/search-aggregations-bucket-terms-aggregation.html)
    • groupInternalOrderBy - array of fields to group by content inside each bucket (no SQL equivalent)
• aggregation - an associative array, keys are aggregation functions like max, min, avg, etc. and values are names of fields the aggregations should apply to, e.g. ['min' => ['id', 'age'], 'avg' => ['age']] (SELECT MIN(id), MIN(age), AVG(age) FROM)
• where - an associative array, keys are conditions with placeholders (?) and values are values. Conditions accept operators =, !=, >, <, >=, <=, LIKE, IS NULL, IS NOT NULL, CROSS FIELDS.
  • LIKE - the value should contain % (the same as in SQL)
  • = - the value can be an array, e.g. ['id = ?' => [1, 3, 7]] (id IN (1, 3, 7))
  • IS NULL, IS NOT NULL - Elasticsearch only
  • CROSS FIELDS - searching string in multiple fields, e.g. name John Smith in fields firstname and surname. Syntax is "firstname,surname CROSS FIELDS ?" => "John Smith"