CPANAHierarchyBundle - Symfony and Neo4j

I've just uploaded to Github my last project: HierarchyBundle


This Symfony bundle manages a company hierarchy retrieving and stroring information to a Neo4j database. The hierarchy is based on groups: PHP group is part of Software Developement group, which is part of IT, etc. The users are members of these groups having different kind of roles (manager, employee etc).

I've got the idea for this bundle from the Neo4j documentation: http://neo4j.com/docs/stable/examples-user-roles-in-graphs.html . Basically it says that relational database are not really suited for representing hierarchical data :)

HierarchyBundle is using the Neo4jPHP library https://github.com/jadell/neo4jphp . On top of this I've created a class called GroupHierarchyManagement which contains specific functions for this scope, like retrieving the complete hierarchy above a user:

    /**
  * Return the hierarchy of of groups above one user node
  *
  * @param  Everyman\Neo4j\Node   node
  * @return row| null
  */
    public function getGroupHierarchyOfUser($node)
    {
        $id = $node->getId();

        $queryString =  ' MATCH n-[rel:'. $this->defRelTypeUserToGroup .']->group '.
                        ' WHERE id(n)='.$id.' '.
                        ' WITH group '.
                        ' MATCH p=group-[r:PART_OF*..]->d '.
                        ' WHERE ID(d)='.$this->rootGroupId.' '.
                        ' RETURN nodes(p) as nodes';

        $query = new Query($this->client, $queryString);
        $result = $query->getResultSet();

        if ($result->count() != 0) {
            return $result->current()->current();
        } else {
            return;
        }
    }

This class is configured as a service with the Symfony Dependency Injection Container. There are several parameters to be configured in order to work correctly:

#app/config/config.yml

cpana_hierarchy:
    group_hierarchy_manager_neo4j:
        neo4j_user:  'neo4j'
        neo4j_password:  'parola'
        def_rel_type_group_to_group: 'PART_OF'
        def_rel_type_user_to_group: 'MEMBER_OF'
        root_group_id: '69374'
        manager_role_property: 'manager'
        default_property_group:  'name'
        default_property_user: 'name'

Th first two are the user and password in order to login to Neo4j database. Next are the types (names) of the relations between Groups and Groups or between Users and Groups.
We can say that the team "PHP " is "PART_OF" departmenet "Software developement".
Also we can say employee Kenny is "MEMBER_OF" team "PHP".

You need to specify which is the root group node of your hierarchy by prodiving the Neo4j Id of that node in parameter "root_group_id".

The relation "MEMBER_OF" between Users and Groups has a property called "role", you can define which is the value of this property for users which are managers of a certain group: it can be a number, or explicitly  "manager", or "master" :) etc.

A Group or a User node can have many properties, like name, description for Groups, or age, salary for Users, from these we need to configure which to be displayed by default when we talk about that node, the name should be the most obvious value.

With this service called " group_hierarchy_manager_neo4j" I've built an application with a front side where regular users can browse the data and an Admin area to edit  the data.
Below some photos to get the idea:
Front side:

  

Admin side :

Neo4j intro

Neo4j is a graph database management system developed by Neo Technology, Inc. Described by its developers as an ACID-compliant transactional database with native graph storage and processing,Neo4j is the most popular graph database according to db-engines.com ^[wikipedia]

I am interested in Neo4j for modelling hierarchies as the RDBMS are not really suited for this ( http://www.codeproject.com/Articles/22824/A-Model-to-Represent-Directed-Acyclic-Graphs-DAG-o)

 Installing Neo4j Community Edition on Windows is a piece of cake: http://neo4j.com/download/
Also there is plenty of documentation and many answered questions on StackOverflow :)

A good place to start: https://www.airpair.com/neo4j/posts/getting-started-with-neo4j-and-cypher

Some useful queries:
--------------------------------
Match Node by id:

MATCH (n)
WHERE id(n) = 14
RETURN n;
---------------------------------
Match relationship by Id

MATCH ()-[r]-() 
WHERE ID(r)=1 
RETURN r 

------------------------------
Find path between 2 nodes, replace "REPORTS_TO" with your relation

START a=node(11), d=node(12)
MATCH p=a-[r:REPORTS_TO*..]-d
RETURN p;
----------------------------
Delete node and all its relationships

MATCH (n { name:'Andres' })
DETACH DELETE n
-----------------------------------------
DELETE all relationships

MATCH  ()-[r:REPORTS_TO]->() DELETE  r;

Importing data from CSV

Official docs: http://neo4j.com/docs/stable/query-load-csv.html

I copied the files to be imported next to the database file because I had issues with other paths. My path looks like: C:\Users\my_user\Documents\Neo4j\file_to_import.csv

Loading nodes:

LOAD CSV FROM 'file:///C:/Users/my_user/Documents/Neo4j/nodes.csv' AS line
CREATE (:User { some_id: line[0], user_name: line[1], link: line[2] })

Add an index on some_id property:

CREATE CONSTRAINT ON (n:User) ASSERT n.some_id IS UNIQUE

Load relations from CSV file, replace PART_OF with your relation:

LOAD CSV  FROM "file:///C:/Users/my_user/Documents/Neo4j/relations.csv" AS line
MATCH (p:User { some_id:line[0] })
MATCH (f:Group { group_id:line[1] })
CREATE (f)-[:PART_OF { some_property: line[2] } ]->(p);